User talk:Robert Ullmann/2007b

Re-reverted ja-template

lots of #ifs are a very bad idea. If we wanted to do that, there is a much easier way. Why not just ask on the talk page? But you would still be missing some cases (for example on and kanon both specified; which does occur.) Robert Ullmann 13:39, 6 April 2007 (UTC)

Please read my code carefully. If you can do it with less #if s, then please, by all means, fix it. However I had considered the case you mentioned, and your objection is not correct:

[1]

Reverting my entries

Robert, please look at what you are reverting before you do so. I was removing finished words, NOT adding anything. Furthermore, if there is a word I wish to add, as long as I do not add a huge amount of words, that is fine. — This unsigned comment was added by sewnmouthsecret (talk • contribs).

Hebrew Characters

I posted a question on Wiktionary:Grease pit#Hebrew Characters concerning Hebrew template writing. I figured there was a good chance that you might have some insight on this issue, and so am requesting that you take a look. Thanks very much. Atelaes 22:10, 7 March 2007 (UTC)

Template:third-person singular of

(You came recommended as the local template expert so I'm asking you.)

Is it reasonable to alter Template:third-person singular of so if lang is not specified it adds a category for category:third-person singular with unspecified lang? So ones without lang (because I didn't know it needed it) can be fixed. Will that take effect when the template is changed or does the article have to be edited first? RJFJR 17:31, 8 March 2007 (UTC)

It is easy to do: {{#if:{{{lang|}}}||[[Category:...]]}}. But as to reasonable, note the issues about lang= in the plural of template, on WT:BP and User talk:BD2412, this needs to be sorted. Robert Ullmann 16:19, 11 March 2007 (UTC)

Answer to the other part of your question: the page doesn't have to be edited, the template change is sufficient. However, the updates are done in the background job queue, so they can take a while. Special:Statistics shows the job queue; when it reaches 0, the updates should be complete. Robert Ullmann 16:25, 11 March 2007 (UTC)

ja-noun template

Hi Robert, I slightly modified ja-noun template as the conditional branching for putting a comma after the hiragana part if param kata or rom exists was not functioning in the case when kata is defined but empty, e.g. in the case with 井戸 or 下着. I think I've tested it exhaustively, but please correct it again if needed. If it's ok, please let me know, and I'll apply the same change to other POS templates. --Tohru 04:56, 9 March 2007 (UTC)

Good, thanks. Is an annoying case in the template syntax. Necessary, because one wants to specify (blank) to override the default. Testing for whether something is defined is much worse: #ifeq:{{{par|+}}}|{{{par|-}}} .... Where else is this not done correctly? (the other ja- templates? Probably, they started as copies of ja-noun ;-) Robert Ullmann 16:15, 11 March 2007 (UTC)

I've modified all the derivations ja-adv, ja-verb and ja-pos I found in Category:Japanese inflection templates, in the same way. Thank you for the review! --Tohru 12:55, 12 March 2007 (UTC)

ISO codes

A while ago Muke suggested using the general code followed by a specifier (in the case of Samoan Plantation Pidgin, cpe-samoanpidgin or something I guess). I don't think it really matters too much what we pick though, so long as we're consistent. --Ptcamn 14:16, 15 March 2007 (UTC)

(I'm going to reply on your talk page, and try to keep it together; if you'd prefer, you can move the whole thing back here ;-) Robert Ullmann 14:19, 15 March 2007 (UTC)

User:[redacted]

Thanks for helping to nab this miscreant. I've reverted his other edits and banned him. — Paul G 14:25, 16 March 2007 (UTC)

Edited the section name (anyone curious can look at page history). Note this one could have been instantly blocked infinite on username alone, which both Connel and I did ... Robert Ullmann 14:33, 16 March 2007 (UTC)

Re: רבי

Thank you, Robert. I'm new to editing Wiktionary so I was unaware of when it was appropriate to use "Translingual." I didn't like the idea of using it anyway, I prefer the way you described. Thanks again. --334a 01:38, 17 March 2007 (UTC)

I would appreciate some help with the "Zlomiti" article

Hello and thank you for your fast reply to my plea for help. I understand now why the article got the template, but I haven't quite figured out how to fix it. I tried something, but it doesn't look quite right. Would you please care to take a look at it and tell me how it should be done? Here are the meanings of zlomiti in Slovene (from the break page:

• (intransitive) To end up in two or more pieces, which can't easily be reassembled.
• (intransitive) (medicine) Of a bone, to crack or fracture through a sudden physical strain, such as a collision.
• (transitive) To cause to end up in two pieces.
• (transitive) (medicine) (ergative) Of a bone, to cause to crack under physical strain.
• (transitive) (medicine) (ergative) Of a bone, to fracture accidentally.
• (transitive) To cause a person or animal to lose his/her/its will, usually obtained by means of torture.

Thank you very much in advance for your help. I believe it would be most helpful if you showed me what to do in the article itself (editing it acordingly), or just explaining it to me here or on my talk page. Thank you again, --Burek 08:14, 17 March 2007 (UTC)

Bot block

Sorry about that - I guess you are still experimenting. I've undone that block and will just let you go at it. (Didn't check ownership before the bad block.) --Connel MacKenzie 17:43, 17 March 2007 (UTC)

Was going to post a heads up on WT:BP presently, once I had just a bit put together. Just playing a bit first. Thanks! Robert Ullmann 17:45, 17 March 2007 (UTC)

BTW, fantastic. I can't wait to see it in high-volume. --Connel MacKenzie 04:18, 21 March 2007 (UTC)

L2 headers

Would you mind updating your analysis of valid and invalid L2 headers? It would help me a great deal, since I have done quite a lot of work adding ISO templates and such since the last run. I keep re-examining the same links because I can't remember which ones I've attended to! Thanks. --EncycloPetey 03:53, 18 March 2007 (UTC)

Need the next XML dump. The current one is 25/26 February. We are about due for another run, but the WMF elves haven't started another pass. You can look at the index to the current dumps and prod me if I don't notice. Robert Ullmann 13:49, 18 March 2007 (UTC)

I'm pretty sure I got a more recent one, that hasn't updated the links properly. (download.wm.org is down at the moment.) --Connel MacKenzie 04:17, 21 March 2007 (UTC) Ooops, no, I regenerated some stuff based on Dvortygirl uploads. That is the most recent one. (And I hope the devs give priority to the new toolserver server, instead of the next round of XML dumps.) --Connel MacKenzie 00:50, 23 March 2007 (UTC)

Since toolserver is essentially caught up (for us) right now, it might be worth checking Patrick Stridvall's toolserver page. Robert, do you have a toolserver account (yet)? Are you comfortable in PHP? --Connel MacKenzie 06:59, 23 March 2007 (UTC)

No, I don't have a toolserver account. I haven't used PHP very much, but it'll only be the 42nd or so computer language for me ... The XML dumps work very well for me because the net here sometimes is flaky, usually slow, and sometimes goes away for many hours, and I can just run some things offline. Robert Ullmann 11:07, 23 March 2007 (UTC)

OK. Irealized last night that might be a problem. It does seem to be the case since I know there have been a lot of edits on the pages with invalid L2 headers that were quite certainly wrong. --EncycloPetey 13:59, 19 March 2007 (UTC)

I've been adding to the program, to catch the codes that are used as extensions within WM for languages that aren't coded; minor languages only coded as groups in ISO 639: Template:fiu-vro and codes that were used even though a perfectly good ISO 639-3 code existed: Template:zh-min-nan. We will probably end up with a lot in the first category. Robert Ullmann 14:05, 19 March 2007 (UTC)

nan-cn nan-tw

Hi Robert, what is {nan-can} and {nan-tw} in Wiktionary? And what is the difference. --Keene 01:07, 24 March 2007 (UTC)

I'm a complete layman with regards to Min Nan. :)--Keene 01:08, 24 March 2007 (UTC)

If you don't know that much about Min Nan, why are you mucking about with Template:nan-adj? I put it back.

nan-cn is Min Nan in simplified characters, nan-tw is traditional, nan (in this case) means both. These cat names are non standard, but we haven't figured out what to do with them yet. When we do, we'll fix all the templates. (nan- and cmn-) Robert Ullmann 11:26, 24 March 2007 (UTC)

líu or liú (or both)

Are líu or liú different syllables? Or is one an error (I would think liú is correct, as every other i-u combination has the diacritic over the u)? Cheers! bd2412 T 04:55, 25 March 2007 (UTC)

líu is incorrect (presuming we are talking about Mandarin Hanyu Pinyin ;-) the tone diacritic is on the u. Rule: a or e gets the tone; if ou, o gets it; else last vowel. Equivalently: if first is i, u, or ü, second gets the tone, else first. We can delete the redirect, since you fixed the remaining incorrect reference ;-) Robert Ullmann 11:27, 25 March 2007 (UTC)

Thanks - done. bd2412 T 14:51, 25 March 2007 (UTC)

Merge

Could you please merge the contents of User talk:AutoFormat here to this page, and leave a redirect over there, to here? --Connel MacKenzie 22:03, 25 March 2007 (UTC)

Um, I think it is much better there. Especially since I am not necessarily the permanent bot-runner. Watching both is fine? Robert Ullmann 22:14, 25 March 2007 (UTC)

In particular, anyone interested in the history of AutoFormat is much better served by not having the other stuff here mixed in. Robert Ullmann 22:53, 25 March 2007 (UTC)

AutoFormat headers

You might wish to add Pronunciation 1 (2, 3) to the list. Also, I *think* we eventually decided that Transitive verb and Intransitive verb should simply be "verb" with the {{transitive}} or {{intransitive}} template inserted at the head of each definition. This is a trickier issue than I'd want a bot to handle, but it would be great if your bot tagged these in some way and categorized them for people who like to do that sort of repair editing. --EncycloPetey 22:57, 25 March 2007 (UTC)

Yes, Transitive/Intransitive verb should be fixed, but not by a bot I don't think. If we had a category, we should tag all of them with the category. (fairly easy to do).

Since when are Pronunciation N entries legit? They should be ety entries I should think?

Not in cases where the etymology is the same for two or more pronunciations; each of which pertains to a different part of speech. I've just edited abject to show you what I mean. Compare the format of the pronunciation section(s) before and after my edit. This sort of situation is relatively uncommon overall, but common enough that I encounter it on a regular basis. --EncycloPetey 23:10, 25 March 2007 (UTC)

moved to User talk:AutoFormat Robert Ullmann 23:07, 25 March 2007 (UTC)

Answered there. Robert Ullmann 23:21, 25 March 2007 (UTC)

Online hanja encyclopedia available

Very few people seems to know about this but the korean search engine www.naver.com has online enyclopedias and dictionaries, including a Hanja dictionary that contains more information on hanja than en.wikitionary. My believe is that this new source could be added as an additional reference for han characters found in hanja. And since all the pages use the address hanja.naver.com/hanja.naver?where=brow_hanja&id={5-digit max charater number} it would be an easy job for a bot to find the corresponding page number and add the link to the current wiktionary page.

The easy way to find the hanja's page is, starting at the main page http://hanja.naver.com/, do a search with a hanja character and a link to the hanja's page comes up as the first search result.

Using 姉 as an example; some of the informations given on naver for 姉 are:

• 훈음  : 손윗누이 자 - the eumhun (caracter's sound and name), here the name comes first, followed by the sound
• 사성음  : zǐ - the pronouciation in pinyin
• 부수 : 女 - the radical character
• 획수 : 8 (부수획수:3) - number of stokes (here: 8) and number of stroke in the radical (here: 3)
• 난이도 - degree of difficulty, but I don't fully understand how the qualification works
• 뜻풀이 - list of the possible definitions

• 회의문자 - ethymology and description of the character
• 획순 보기 - stroke order (as used in Korea)
• 활용 단어 - examples of applications of the character

While some entries here give the full eumhun, such as 字 or 有, alot don't, such as 云 or 短.

I'm saying all this because I want to propose an improved model for {{ko-hanja}} that will use some elements of {{ja-kanji}} and {{ja-readings}}. I don't know anything about bot programming but I think a bot may be able to gather and fill in with the new information (except maybe romanization).

My proposition for the changes in {{ko-hanja}} and in the Korean section would look like this (for 姉):

Hanja

姉

Readings

Eumhun:

• Sound: 자 (revised ja, McCune-Reischauer cha)
• Name: 손윗누이 (revised sonwitnui, McCune-Reischauer sonwitnui)

Compounds

(the compounds in this example are written in hanja but linking to the words in hangeul)

• 姉妹
• 姉妹結緣
• 姉妹校
• 姉妹機關
• 姉妹船
• 姉妹新聞
• 姉妹語
• 姉妹篇
• 姉妹學校
• 姉妹會社

Reference

• http://hanja.naver.com/hanja.naver?where=brow_hanja&id=11063

I know you're responsible for alot of the work in the han character pages. I hope this would allow to bring some improvement. - Luccas 05:53, 26 March 2007 (UTC)

XML dump in progress

2007-03-28 16:31:22 enwiktionary: Dump in progress --Connel MacKenzie 16:33, 28 March 2007 (UTC) http://download.wikimedia.org/enwiktionary/20070328/enwiktionary-20070328-pages-meta-current.xml.bz2 is done. --Connel MacKenzie 17:00, 28 March 2007 (UTC)

Good! Collecting it now. Thanks Robert Ullmann 17:21, 28 March 2007 (UTC)

unblock

Please unblock Dodde...he was creating the 1st draft of the language templates as I had requested he do...

--Connel MacKenzie 02:06, 30 March 2007 (UTC)

Sorry, it was 3:17 AM and I could barely keep my eyes open; why I added a note to GP. Robert Ullmann 13:03, 30 March 2007 (UTC)

{{t}}

Please add some docu on your change to {{t}}. I think it is a good solution, everybody should be able to live with that, although, it makes a lot of my effort of the last weeks more or less in vain. But there’s worse.

I added some documentation: Template_talk:t#Section_references, please have a look at it. H. (talk) 14:32, 1 April 2007 (UTC)

I think the template can be simplified, but I am not sure enough about my template coding skills, so I’d like to discuss it with you first.

1. One check on ls should do: in the beginning. Alternatively, just use {{{ls|English}}} always.
2. If you define an identity template (this probably does exist somewhere), you can do the same trick for sc: {{{sc|id-temp}}}.
3. There is a totally redundant check for {{{1}}} in there.
4. I don’t see the use for the switch on {{{3}}}: just always use {{{{{3}}}}}, um no, that doesn’t work if there is no {{{3}}}, well, why not just an if. Similar for {{{4}}}.

What do you think? H. (talk) 21:16, 31 March 2007 (UTC)

Let me just acknowledge your note for the moment; it is 4:57 AM here; I will answer better later. Work sometimes doesn't want to follow time zones; especially the stuff I do. Robert Ullmann 01:58, 1 April 2007 (UTC)

1. The check on ls is repeated because it needs to do two different cases for sc (and remember that it is parsing it as text, not precompiling it; so only one of the two #if:ls calls is fully parsed and executed. I don't want to default to English always because the result is visible in the URL the user is sent to. So if there is a #section reference it needs to be right.
2. The parser conditional costs much less than an identity template. (Something several other people don't seem to get at all; a parser function call is a thousand or so machine instructions all-in, maybe one ten-millionth of a second. A sub-template call causes an entry fetch, which results in an SQL query, that goes over the internal net to the backend database server and then may cause disk accesses. Many orders of magnitude more load! Mind you if it is useful, all good; and then the results are cached. But see what the trade-off is?)

   ACK.

3. I don't see any redundant check for {1}? The #if has to be outside the wiki-brackets, the #switch (to handle the zh- cases) has to be inside.

   Yes, but why is there an #if at all? You use {2} before that, so there must be a {1} at that point. Do you want to design something like Dodde has in mind, which does not require the language code?

   Not true, {2} can exist without {1}. I think this is one reason why Dodde is going off into lots of cruft, and DAVilla likewise; they know enough to be dangerous, and not enough to write decent code. (Why have "xx"? you just leave it blank.) Like this: {{t||word}} see? It doesn't require the language code now. Robert Ullmann 15:38, 1 April 2007 (UTC)

   Ah yes, I tend to forget that. I guess my comments have been void then. H. (talk) 09:42, 5 April 2007 (UTC)

4. The switch is effectively just a fancy if testing set membership. In python, I'd write: if p3 in ('f','m','mf','n','c'), that is what the switch is doing in the template syntax. Keeps people from getting interesting results if they use a non-standard value.

   Yes, but not particularly nice. An idiom one has to get used to, I guess. Backdraw is that the template is to be adapted if other similar templates are created (which might be the case for Bantu languages sometime).

   Not too likely, but yes you could add them; I wouldn't put noun classes in for Swahili or Kinyarwanda (the ones I know something of, I live in Nairobi), any more than we would want to put verb classes in for, say, Finnish. Anything more than m/f pl really just belongs in the entry. Robert Ullmann 15:38, 1 April 2007 (UTC)

   I remember a discussion with someone (was it you or Stephen), about noun classes in translation sections: I removed them, the other one put them back. Who knows, one day they’ll be expected. H. (talk) 09:42, 5 April 2007 (UTC)

I don't see any particular tweak right now that would improve it; but always willing to Robert Ullmann 11:53, 1 April 2007 (UTC)

I was going to make a bot for handling {t} in translations sections on svwikt and suggested Connel I could run the service on enwikt too... He told me this would be very much appreciated, and mainly because of this I have participated in discussions regarding the syntax. When you blocked me and entered the discussion, it seems that everything of the discussions that far got wiped off by you with degrading posts about rocket science and reinventing the wheel etc and you substituted everything with a non-flexible solution set by you based on rules you set up. Maybe this is how it works on enwikt. Well, in that case I won't intervene. What I wonder though is if Doddebot is still in the plan taking care of the service, or if you are going to make the bot yourself. In that case I concentrate running the service only for svwikt, and leave my "watch" over the discussions. ~ Dodde 11:06, 3 April 2007 (UTC)

First: I did not block you. I blocked DoddeBot because it was doing something that had not even been mentioned on GP. Blocking a bot that is doing something wrong or simply not understood is routine. It is not blocking a user.

At to the rest of it: I was pointing out considerations that were being ignored in the rush to do something new and whizzy. THey aren't "my rules", they are an observation of reality: t will be used hundreds of times per entry, and 10's of millions of times in the en.wikt (say 200,000 words/senses with fairly complete tables, by 500 languages). It simply can't use sub-template calls. Just can't. The performance hit is impossible to accept. If having that reality intrude is upsetting, well, I understand, but there it is.

As to bots: yes, I have a bot that will run as of an XML dump. I believe you were thinking of something that does some lookups on the current wikts and updates? just explain on GP what you would do and see what people think. Robert Ullmann 12:33, 4 April 2007 (UTC)

Fair enough. I hope you agree that the discussion has been held long enough to come to a result. I have made a list on GP which I hope you and all others in most can come together accepting. If the list is accepted as it is, I wish that you would be willing to use parts of {t} {t8}, {t8+}, {t7} {t7+} and {t7-} to create server effective coding in those templates that will support all points. Regarding the bot I will reply later... ~ Dodde 07:01, 5 April 2007 (UTC)

I agree that it is a big pity that the ls is not added automatically anymore, but I think your bot is handling that nicely. It is something we have to live with, I guess. Until #locallanguage: gets implemented... H. (talk) 09:42, 5 April 2007 (UTC)

Ullman can you verify DAVilla's suggestions are (summarized in point 2 and 4 here [2]) not causing too much server load for {t+} {t-} and {t!}? ~ Dodde 07:18, 8 April 2007 (UTC)

I haven't had time for a few days to go through the discussion again; but on those: having 2 or 3 or 4 t* templates isn't a problem, using the t-xx code templates is. The parameters really need to be the same for all of the templates; making one "user friendly" and the others "server friendly" or something is a bad UI design mistake. And unneeded; users will just use the first 2-4 unnamed parameters (+ lang in some cases), and the bot will play with the rest. More when I get a chance (I still have to look at more things on BP ;-) Robert Ullmann 07:28, 8 April 2007 (UTC)

top/mid/bottom --> trans-top/trans-mid/trans-bottom

In the main namespace, can you please add this conversion to the "todo" list for AutoFormat? I suppose it can skip the transformation, if there is no bold line preceding the {{top}}. --04:17, 2 April 2007 (UTC) (Connel)

If there isn't any gloss, it just generates trans-top with no parameters. Done, and it seems to find a number of them. Robert Ullmann 21:52, 2 April 2007 (UTC)

So I'm on your black list...

Okay, so you don't like what I'm doing with the code, whether it be {{context}}, {{t}}, {{langcode}} or what have you. And experimenting with the language templates was a last straw for you, especially when I didn't realize my error. You've asked me not to complicate things any more. I can live with that. I've made a bunch of reversions and posted a lot of messages on talk pages, some apologies and suggestions and requests, but not actually toying with anything since you've bound my hands. I should say that I plan to gut {{context}} as you've asked and hopefully not break anything, although I'm not sure how to do categorization without {{language}}. But then I saw that you reverted my changes to Template talk:t. Why? The changes were cleaning up the documentation. You can make all the accusations you like against me and I'll try to accept them and learn from you. But this change really offends me. Do you just distrust absolutely everything I do now? DAVilla 07:33, 3 April 2007 (UTC)

No you aren't on a blacklist ... I must apologise for the rant ... I took Miriam (SO) to the hospital last night in severe pain and came home after they admitted her; tried to do some wiktwork and found I had to undo things to get to the starting point I thought I was at. From a grammatical POV, that sentence should be shot.

I reverted the talk page because I had just done three other things, and wanted to start where I left off. (I can compose edits off-line in my head like when I was sitting at the hospital; but if the text has changed when I go to key it, I have to start over again ;-) Sorry, I'll go look at it. One thing was the the examples were intentionally verbatim from butterfly; if anything the variant pronunciations/transliterations for the Arabic could just be left out, it is just explaining {t}. Robert Ullmann 10:45, 3 April 2007 (UTC)

Sorry to hear.

Don't worry about Template talk:t. I will make an {{example}} template to make the cleanup easier to do. DAVilla 12:48, 3 April 2007 (UTC)

This is proving to be of dubious utility. DAVilla 13:50, 8 April 2007 (UTC)

WOTD woes

Could you please take a look at the latest bug I've noted in the Grease Pit (assuming you haven't already noticed it). The bug is affecting the Main Page and making us look bad. --EncycloPetey 15:38, 6 April 2007 (UTC)

quotations

Thank you. Is this the way it should look then? [3] BTW, I have a profile with some 'clues' on Wikipedia if you care to know - but surely learning new words is more important. It's a shame I didn't know about this layout for quotes sooner - I have been adding quotes for a while now...zigzig20s 23:44, 6 April 2007 (UTC)

BP discussions

Robert, these two discussions should have your feedback:

1. Wiktionary:Beer_parlour#Components_of_Chinese_characters
2. Wiktionary:Beer_parlour#Amoy

A-cai 13:39, 7 April 2007 (UTC)

Confirmation: translation templates

For sake of clarity, could you please confirm that your code as now present in {{t-}} is "server friendly"? (Dodde had a question about sc= and alt= and elsewhere I saw a comment by Connel about #switch statements that seems hyperbolic.) Thank you! DAVilla 13:26, 8 April 2007 (UTC)

As far as it can be I think. There is some overhead in #if and #switch I don't understand yet. But the current version is the friendliest we can get to. t- et al can be simplified by changing the #if #language to unconditional; the code should always be present. I'm sorry I haven't time yet to get back to the whole conversation; need to. Robert Ullmann 13:31, 8 April 2007 (UTC)

I was wondering, incidentally, if it's possible to move {{|{{{sc}}}|}} inside of the link. This would simplify the code.

The conclusion of the conversation between Dodde and myself is at Time to settle this thing #2 and there's a new discussion about gender and number. DAVilla 13:50, 8 April 2007 (UTC)

Yes, I've read part of it; need to read through the whole thing again.

sc inside, good idea; it didn't matter much before the lang= conditional as well. Done. Robert Ullmann 14:07, 8 April 2007 (UTC)

Meaningless syllables

See my reply. --KYPark 12:55, 9 April 2007 (UTC)

A few things

First, I agree that translingual should be the overarching L2 header for graphemes, and will try and argue that point, but cannot promise any results, as there are others who disagree with me. Secondly, do you know if there is any way to prduce a numbered category? What I mean by that is, to place a bunch of pages in some category, but with the format [[Category:Random Category|#]], and then have the category sort the pages within it by those numbers? I've looked through all the Meta literature on the subject and can't find anything. I'm starting to think that it simply cannot be done, but I thought I'd at least throw it by you, and see if you knew of anything. Note, this is completely unrelated to the graphemes issue. Thanks. Atelaes 17:45, 12 April 2007 (UTC)

The sort key is a string, the category will sort by the entire string, but will always use only the first character as the index letter shown for each section. What are you really trying to do? Robert Ullmann 17:48, 12 April 2007 (UTC)

To be honest, I would like to use such a category for a number of things. First, as I had mentioned a while back, I would like to use this for a Strong's index. Certainly, we can simply hand-write such an index, and then we'll be able to see what's red and what's blue. But, I really want a dynamic list, which only lists words that we have (and have properly tagged). Secondly, I want this for the books of the Bible. As there are many different bibles, (and I'd want them sorted by biblical order, not alphabetical order if possible), I'd have to hand-write each index, which would be a pain in the ass. I've thought of a number of different projects where a category which can be ordered however I see fit would be terribly useful, but can't think of them all off the top of my head. Atelaes 17:56, 12 April 2007 (UTC)

If the set of entries all use some template (like the one for Strong's), and it puts all the entries in a cat (in default order), it isn't to hard to read the cat and write an index page by minor magic. Do you know how to run pywikipediabot? I can give you a framework module and you can tweak it as you like? It isn't a general solution, but with Python code writing a page you can do whatever you want. Robert Ullmann 18:04, 12 April 2007 (UTC)

I suppose that might be the best way to do it. And no, I don't know how to run a bot. It is my intention to learn a bit of python (I have the book and everything). God only knows when that will actually happen. But, once I do know enough to safely handle a bot, I'll definitely take you up on that offer. Thanks. Atelaes 18:10, 12 April 2007 (UTC)

Tempalte:wikipediadab

Do you prefer like that? 16@r 16:28, 13 April 2007 (UTC)

Kazuo Ishiguro

Why the sigh, if you don't mind me asking?zigzig20s 23:24, 15 April 2007 (UTC)

Sigh because incredibly good writers I have not spent enough time reading! Robert Ullmann 23:28, 15 April 2007 (UTC)

Well, it goes to show that I'm incredibly uncultured, since I'm only reading some of those books now...zigzig20s 23:32, 15 April 2007 (UTC)

Thanks for your suggestions.zigzig20s 23:53, 15 April 2007 (UTC)

Question marks are a title?zigzig20s 00:07, 16 April 2007 (UTC)

Re:Mainpage/otherlang

Hi, it may be so but it is included in the mainpage and on en.wikt you seem to have protected the last with cascading turned on => otherlang is also protected.

Regards --birdy _{(:> )=|} 14:58, 16 April 2007 (UTC)

Idiom template

Hey, the template: {{idiomatic}} needs to have the language qualifier jazz added to it, so that γλαῦκ’ εἰς Ἀθήνας goes under Category:grc:Idioms instead of Category:Idioms. I'm not particularily confident in my knowledge of how all that works, and I'm generally hesitant about editing templates that a lot of pages use, and so I figured I'd just play it safe and ask you to do it. Would you have the time to do so? Thanks. Atelaes 23:59, 16 April 2007 (UTC)

It already has it, it uses context/tag. Robert Ullmann 10:22, 17 April 2007 (UTC)

Ok, I'm really sorry for my ineptness, but I can't figure it out for the life of me. How do I enter the language qualifier? I'm assuming some parameter in the template is required, but none of my attempts work. I tried putting the parameter as grc, as lang=grc, as lang={{grc}}, but none of these work. What am I doing wrong? Atelaes 23:56, 18 April 2007 (UTC)

Ah, the category has to exist. Create it first ;-) Robert Ullmann 23:59, 18 April 2007 (UTC)

Ok, the category exists, but it's not being pulled up. What now? Atelaes 00:02, 19 April 2007 (UTC)

Also, this template requires the full language name, not the ISO code, so the category must be Category:Ancient Greek idioms. --EncycloPetey 00:02, 19 April 2007 (UTC)

Yes, that would be the problem, this is a POS cat. Robert Ullmann 00:04, 19 April 2007 (UTC)

But it's still not working! could the space in the language name be a problem? Is Ancient Greek recognized as a language by the appropriate templates? --EncycloPetey 00:06, 19 April 2007 (UTC)

Really? I guess I thought that "Idiom" as a part of speech was sort of deprecated, along with verbal phrase, noun phrase and the like, and that we were simply using it here because we couldn't figure out the proper POS. Atelaes 00:07, 19 April 2007 (UTC)

No, it's still used. See WT:POS. We just try to use other POS headers first. --EncycloPetey 00:10, 19 April 2007 (UTC)

No, Idiom is standard, it can't be reduced to Phrase or something. But this ought to be working now; and I've purged the cache on the page. The {{grc}} template: Template:grc is okay. Dunno. Robert Ullmann 00:11, 19 April 2007 (UTC)

DAVilla make this dependent on a "Language:" pseudo-namespace entry in some way. Don't worry about it, it will get fixed when that goes away. I suppose. Robert Ullmann 00:16, 19 April 2007 (UTC)

Sorry for getting into this discussion so late. I changed {{language}} because lang templates that evaluate to a link, as Ancient Greek in this case, do not work with the categorization. I only filled out a few languages in this experiment, not even the top 40, because I wasn't sure if "Language:" would be agreed to as the best place to put the information.

When I wrote the evil {{language}} template I had used the reserved two- and three-letter templates not knowing what they were for, initially, and it had seemed like a secondary purpose when I did learn. But it's clear now that the language information needs to be duplicated somewhere else. Or we need to find a way of getting rid of {{language}}, which I don't know how to do since it seems like the information is required for some context templates like {{idiomatic}}.

Note that {{grc}} is not top 40 and would not have worked even under the old version, so rewriting {{language}} had little to do with {{idiomatic}} not working properly. It wouldn't have worked anyway. The story should be different with Language:grc and Language:Ancient Greek up.

There is also a problem with {{context}} in that the categories have to exist even when they are named explicitly as with {{idiomatic}}. It's not that this wasn't thought through, just the conclusion was a bit off. The new version (momentarily at {{context2}}) already fixes this. DAVilla 18:34, 19 April 2007 (UTC)

I changed it to the old-fashioned way of designating POS cats. This works better anyway, because it allows for the proper sort name (i.e. no diacritics). Atelaes 00:19, 19 April 2007 (UTC)

Interesting. I did a preview with {context|idiomatic|lang=Ancient Greek}, but that puts the term in both Category:Idioms and Category:Ancient Greek idioms. --EncycloPetey 00:23, 19 April 2007 (UTC)

Yes, very interesting. Robert Ullmann 00:25, 19 April 2007 (UTC)

get

Hey,

Could you take a look at get? Your bot recently tagged it, and I think I've fixed the problem, but am not sure (there were a lot of problems there, and I wasn't 100% clear on what exactly it was complaining about). Thanks in advance!

—Ruakh_TALK 16:02, 22 April 2007 (UTC)

Looks pretty good to me. The bot could have fixed some, but the odd headers would have been a problem. It is a bit odd to have "other derived terms" as one of the tables under "Related terms". (Specifically the complain was that Related terms at level 4 as a sub-section of "Quotations" is wrong somehow. That it would have fixed. Robert Ullmann 18:11, 22 April 2007 (UTC)

nappy

I have a question, please. Why do you keep removing references to afro textured hair from nappy? Is it a question of legitimacy or what? I think that it is needed, even if it is a US colloquial phrase... 67.99.175.237 15:05, 23 April 2007 (UTC)

Because it isn't specific to African/Afro/African-American, and the definition is there. (there is also a serious POV problem, which we are avoiding for the present) Robert Ullmann 15:10, 23 April 2007 (UTC)

Message on rw.wiktionary

Hello Responding here. Koavf 20:40, 26 April 2007 (UTC)

Hi, wassup? Robert Ullmann 22:03, 26 April 2007 (UTC)

wine

I think the reason the Pronunciation section was above the Etymologies is because it applies to both Etymology sections. Isn't that the way we do it? Or are we supposed to be duplicating the information? Widsith 12:34, 27 April 2007 (UTC)

But it wasn't, it was after the first Etymology 1 header. If it applies to both it should indeed be before. Does it apply to Ety 2? Robert Ullmann 12:40, 27 April 2007 (UTC)

Ah, you were editing at the same time I was, I was already working from the edit before. (why this didn't cause an edit conflict is an odd WM bug. Robert Ullmann 12:43, 27 April 2007 (UTC)

Why does wine have cat Greek derivations? Robert Ullmann 13:00, 27 April 2007 (UTC)

Hmm, shouldn't be – I'll take that out. Widsith 14:54, 27 April 2007 (UTC)

Footnotes

We don't use footnotes. That's too bad since they can be quite useful. Okay, I'll list my templates for deletion. Jimp 16:20, 1 May 2007 (UTC)

No worries, I didn't know about the no-footnote rule, thanks for keeping me informed. I am thinking of merging w:List of English irregular verbs to Appendix:Irregular verbs:English. In the table on the Wikipedia article there's a column of notes about the form of the verb (week, strong class 1, etc.) each with a link to the Wikipedia article/section. What I had in mind was reproducing this column in the Wiktionary appendix but using footnotes instead of repeated links so that we'd only linkto the article/section once. I guess I can get by without the footnotes. Any ideas? Meanwhile on with the deletion of the footnote templates (there are others besides mine). Jimp 16:47, 1 May 2007 (UTC)

Japanese verb forms

So, for some reason I've decided to work my way through "I Am a Cat" and turn all these links blue. There are lots of verb and other inflections here, and I suppose since we include every little word form in other languages that we'll be including Japanese ones too.

Do you have any suggestions on how to use {{form of}}-type templates to keep these organized? My first thought would be to create a "Template:ja-past of verb" patterned after "Template:simple past of", and similarly for the other inflections. Do you have a better idea?

Perhaps an all-in-one template?

Cynewulf 17:35, 1 May 2007 (UTC)

I'm replying here because I'm interested in the CFI for Japanese articles. Looking at WT:CFI and WT:AJA, I couldn't find a clear rule. Of course CFI requires attestation and idiomaticity. Attestation is obviously not an issue here, but idiomaticity is unclear.

The basic rule "An expression is “idiomatic” if its full meaning cannot be easily derived from the meaning of its separate components." is quite clear, but there's an exception for inflections: "Although it is not forbidden, there is no particular need to include completely regular inflections such as cameras or asked." But WT:AJA does not clearly state what is and what is not an inflection, so it is not clear if this exception applies or not.

There are a few possibilities given the text of WT:AJA:

1. Inflection is whatever the most popular way to define it among Japanese linguists is.
2. Inflection is whatever is taught in schools (modern 国語 texts) to be so.
3. Inflection is whatever is mentioned in the template (e.g. Template:ja-suru) for each inflecting term.
4. Inflection is some subset of 3., e.g. the ones marked as "Stem Forms".
5. Something completely different, as WT:AJA doesn't say clearly.

What little I have heard of Japanese grammar suggests to me that 4. and 2. are the same thing. E.g. if you look up 活用 in 大辞林, it says that when 書く combines with the auxiliary verb ない it inflects to the form 書か, and when it becomes the imperative form it inflects to 書け. 大辞林 has entries for such auxiliary verbs as e.g. れる, させる, and いる. The situation with these would be analogous to English must, was, not, etc. (and there aren't entries such as must not eat, even if eaten exists)

Also, Japanese Wikipedia (how much do you trust it?) says in its 活用 article that in the school grammar the 活用形 (inflection forms) are 未然形, 連用形, 終止形, 連体形, 仮定形 and 命令形. These are the stems given in the templates (except that in 五段活用 article there is additional 未然形 form -o and additional 連用形 form -i).

Your example follows either 1. or 5. as つかぬ (and いられぬ) can't be formed solely using the forms given in the templates, and would (as far as I understand) under interpretations 2.-4. be split as つか・ぬ (and いられぬ as either い・られ・ぬ or いられ・ぬ).

What kind of rule do you think would be best? I like the school grammar approach as I don't see any more point in entries such as みない than I see in ones such as does not look, but then I don't see the usefulness of cars (as opposed to just car) either, so I'm biased. In any case some kind of line between attested entries that should be included and ones that shouldn't should be drawn, right? Whatever is chosen it'd be nice if WT:AJA would then state the rule clearly. -- Coffee2theorems 23:56, 4 May 2007 (UTC)

The rule for idiomaticity only applies to multi-word forms, which seems to be interpreted by everybody else to mean "things that have spaces in them". The assumption is that we can't assume people know to decompose weeknight into week night and not wee knight, and for languages other than English this is even more relevant. So, we have 見ない because people looking on the English wikt for such a thing probably aren't able to instantly realize that the dictionary form is 見る. (And, by the way, 見ない shouldn't be a hard redirect.) For an inflection in another language, see e.g. comeríamos (which is what, "if we eat"?). Cynewulf 15:07, 6 May 2007 (UTC)

The rule "things that have spaces in them" works unfortunately badly in Japanese, as you well know - they are rarely used even in full sentences :-) Spaces are convenient in that they usually mark some kind of word boundaries in languages, but as Japanese doesn't use them, some other means are needed. I suggested using school grammar for deciding what's an inflection (eating) and what is an auxiliary (must eat). I take it that you disagree with that, and you may be right in that something else would be more suitable for English Wiktionary (I'm not convinced though). Could you explain what the exact rules you use for deciding whether something consists of multiple words or not are?

Personally I don't see how, if people can't split 見ない to 見 (imperfective form of 見る) and ない (auxiliary verb), they would then be able to do that for e.g. 言われなくても or させられたくなかったけれど. Based on your example text, you would have an entry for 見ない but not for 見ていない (instead splitting it to 見て and いない) - why? How about 見ながら, 見るまい, 見たり, 見られる, 見れる, 見ず, 見ぬ, 見ます and the endless combinations such as 見られませんでした? I think that as only limited amount of forms can ever be included, it would be good to pick and choose them so that the set makes some sort of sense as a whole, instead of being a random subset of a much larger set. The latter will very easily happen because there are so many ways to combine auxiliary words. -- Coffee2theorems 18:47, 6 May 2007 (UTC)

I've gone and created a prototype at 生れた, with a modification to {{ja-pos}} and a new template Template:ja-past of verb. Please let me know what you think. Cynewulf 15:07, 6 May 2007 (UTC)

I've been busy with several things, will get back here soon. There is a much larger issue re formatting of inflections I think needs to be looked at ;-) Robert Ullmann 22:50, 6 May 2007 (UTC)

If you're looking at inflections, you could modify the godan stems in the template to match that at Wikipedia while you're at it (there's a table at the bottom - what's missing is the modern written forms of certain uses of 未然形 and 連用形, i.e. 書いて instead of 書きて, 切って instead of 切りて, 読もう instead of 読まう, 切ろう instead of 切らう). -- Coffee2theorems 01:40, 7 May 2007 (UTC)

Han character template

Hi! I take it that the "Robert" A-cai referred to is you. I made a concrete proposal for the Han character template at Wiktionary:Beer_parlour#Components_of_Chinese_characters. A-cai made a very similar proposal (mine is more detailed and differs in linking and use of IDS), and I think that the differences between the proposals do not affect the template at all. Could you respond to the discussion and possibly modify the template if you agree with it? -- Coffee2theorems 20:59, 1 May 2007 (UTC)

Sorry, lots of things to look at; A-cai did ask me to look at this.

I added ids= to the template, as at least a starting point. Robert Ullmann 12:33, 2 May 2007 (UTC)

Interwiki

Ok Robert Ullmann, No problem Thank you for every event. I nee my bot for add interwiki from arabic Wiktionary to english Wiktionary.--OsamaK 13:42, 4 May 2007 (UTC)

Kanji entry structure

Continuing on the topic of kanji entries, I think that their structure could be generally improved. Off the top of my head (some of this applies only to Japanese parts of the entries):

1. Each kanji reading could have more information about it. Their uses are often very different. In some cases they could be marked as archaic, rare or "used only in names". The information of 唐音/漢音/呉音 is also missing. There's also okurigana, e.g in the verb 光(ひか)る the kanji is read as "ひか" and in the name 光(ひかる) as "ひかる", but 光 just says "ひかる" without explaining any further. Perhaps okurigana should be listed as e.g. ひか・る like some kanji dictionaries do (and list all attested ones)? Or as e.g. 光(ひか)る?
2. The translingual section lists meanings. I don't think that's a good idea. It's not consistent with other parts of Wiktionary, e.g. raison d'être doesn't have a translingual section even though the meaning is the same for each language. Editing such a section is difficult as it requires knowledge of many languages, and if you're just interested in one language (as is likely) you have to look in two places. Also, examples for each meaning and the status of each one (is it used anymore? has it ever been used? is it rare?) are likely to differ.
3. The compounds lists can be long. Is there a way to hide them behind a "Show" button?
4. The kanji articles have too much visual clutter. (though that is common elsewhere in Wiktionary too, e.g. finding the meanings from see takes a lot more effort than finding them from 見る - and the conjugation table in the latter would also be better hidden behind a show button) How about moving the non-clickable references to other dictionaries to the end of the article as a references section? If you then move the meanings from the translingual section to the language-specific kanji sections, and the han character template and stroke diagrams to the top (where "see also" is), you could get rid of the translingual section altogether. The unihan link could be added to the han char template, but with shorter link text (e.g. just "unihan" or "U+DEADBEEF").

Is there an example for what a good entry should look like? What do you think of the above proposals? One entry could be chosen and worked on as an example of what an ideal one looks like, and others could then be slowly improved with that as a model. -- Coffee2theorems 18:34, 5 May 2007 (UTC)

more soon, there are good answers to your questions Robert Ullmann 22:51, 6 May 2007 (UTC)

Now that I am sitting here with my coffee ;-)

1. Yes, we could do with more notes on the readings. What we have now is what we have ... I added the kanon= goon=, etc params so that someone could add the more specific information, but for most entries we don't have it. We also have params nadzuke=, nanori= but likewise, most entries don't have the information yet.
2. The "common meaning" in the Translingual section is problematic as you say; the definitions clearly belong in the language sections, under Hanzi or preferably under the correct POS (where applicable). But again, in most of the 20,000 entries, the "common meaning" is all we have right now. (Getting people to stop adding "such and such (Cantonese only)" to the common meaning and put it under Cantonese where it belongs has been a problem ...)
3. Using {{rel-top4|Compounds}} for compounds is a fine idea.
4. The references (like most things) have to be inside a language section or Translingual, so that they are associated with a language or clearly apply to all. (In the case of Han characters it is clear, but in general in the wikt a L2 References section at the end of an article would be impossible to parse ;-) There is a set of things that have to be in Translingual (Canjie input, Four corner, radical and stroke info, etc. Even IDS...) Trying to squirrel them all away in different places isn't a good idea. Where would you put the Etymology for the character? In each of 15+ Chinese language sections, and Japanese and Korean and Vietnamese? Has to be in Translingual. Robert Ullmann 10:19, 7 May 2007 (UTC)

I didn't notice the extra parameters. Looking at Template:ja-readings there's both 名付け and 名乗り - I thought these were the same thing? I.e. readings that are not used in words, only in names. Also shouldn't the romaji for なづけ be nazuke (I don't know much about romaji but I thought they didn't make a distinction between づず)? Otherwise looks nice. But how about okurigana? E.g. ひかる is properly a nanori/nazuke of 光, its kunyomi are ひかり and ひか(continued with the stem forms る, れ, ...). 光った would be read ひかった, not ひかるった.

Then there's more detailed information one could add, e.g. that the 手(しゅう) reading is not used in modern Japanese, examples for each reading etc., but for those one could use a separate bulleted list, right? Same for irregular readings, e.g. お母(かあ)さん, 蜻蛉(とんぼ), 詩歌(しいか). These are usually listed as 難読 if at all, but this is the kind of point where Wiktionary often differs from traditional dictionaries. お母さん is said to come from 母(かか)-->お母(かか)さん-->お母(かあ)さん. Similarly for 父(とと) and お父さん. とと and かか are little children's words. It may be that かか is not written 母 (little children rarely write in kanji! some dictionaries give kanji forms for the word though), so I can see that it's not given as a reading by kanji dictionaries. But お母(かあ)さん is common - should かあ then be listed as a reading, even though traditional dictionaries do not? It is not a stand-alone word as kunyomi usually are. But then again neither are 御(お) and 殿(どの). Whether 御(お) is listed at all, as kun, or as nazuke varies by dictionary. Also one might want to list しい as a 慣用音 of 詩, and then say that it's only used in 詩歌.

Is it OK to duplicate the meanings from the translingual sections to the language sections (where they apply)? Where exactly in the language section each meaning should be listed is an interesting question, too. E.g. I'm not sure what combination of "prefix", "suffix" and "meaning of kanji" the "magazine" meaning of 誌 (which comes from 雑誌) is. I'm not even sure if distinguishing such is meaningful or just an exercise in the fine art of hair-splitting.

Now that you mention it, I'm not so sure about the translinguality of stroke counts (similarly to stroke orders). E.g. 葬 lists stroke count as 15 (also 15 in unihan db), but in modern Japanese at least it is clearly 12. Perhaps they write the three-stroke 艹 part as a six-stroke 艸 in Chinese? In that case, the IDS might also better be ⿳艸死廾 for chinese rather than ⿳艹死廾. I don't know about the radical, maybe that's fine as 艸艹 are just different forms of the same thing. In general there are alternative forms with different stroke counts which were unified in Unicode, e.g. everything with ⻌ in it can also be written with 辶, adding one stroke. In the new characters in Japan the former one is usual, but there might be 表外字 usually printed with 辶 (see wikipedia:Asahi_characters for some information on 新字体/旧字体 of 表外字).

By the way, in Beer parlour you mentioned that only non- compatibility characters, that is Han Unified + Ext A + Ext B, are used here. wikipedia:Han_unification lists other non- compatibility blocks: CJK Radicals Supplement and CJK Strokes. IDS itself allows those code points that have the "Unified_Ideograph" or "Radical" property (literally "Unified_CJK_Ideograph" and "CJK_Radical" in the 4.0 standard text available as PDF, but there are no such properties in either property or property alias lists) and states that there are also some compatibility characters in the set "because the shape differences involved may be relevant to description of the forms of unencoded ideographs". Examples of various cases: 辶 is from the unified ideographs block and ⻌ radicals supplement block. 﨤 is a compatibility character with the Unified_Ideograph property and ⼅, ⼓ compatibility characters with the Radical property.

I agree that etymology, cangjie, radical and references are translingual. The idea about moving the references was just to make the information the reader is probably looking for most prominent. It may be impossible, as Wiktionary tries to be so many things all at once.

Asking you to answer all of these may be a bit much, especially as you're busy :-) It's just that when you ask one question you come up with ten others, and when you're already on the subject then you put them in too.. -- Coffee2theorems 13:54, 7 May 2007 (UTC)

One more idea: If Template:ja-kanjitab allowed readings for each kanji (say, 手(て)紙(がみ<かみ)), you could build complete indices such as "all words with 手(て)" in it, and it could be shown in the template as furigana. I don't know how much is possible with templates, though. -- Coffee2theorems 14:04, 7 May 2007 (UTC)

Addition: 浅 has two forms, one for China and another for Japan. It also says that stroke counts differ, but gives only one of them. -- Coffee2theorems 14:28, 7 May 2007 (UTC)

Fixed; will reply to other things later ;-) Robert Ullmann 14:36, 7 May 2007 (UTC)

Nice to see that such stroke information can be added! :-) Should've looked at the template source.. -- Coffee2theorems 23:27, 7 May 2007 (UTC)

See 級 and 盟 for some cases where I've tried to add some of this information. Cynewulf 14:44, 7 May 2007 (UTC)

Interesting. By the way, I used the words "rare", "obsolete" and "archaic" rather freely until I read Wiktionary:Obsolete and archaic terms. It's not an official policy, but surely consistent usage is a good thing. Now I'm confused which one to use when I know that something is rare now but have no idea about the past :-) I guess "rare" with a date, but I don't know such dates. So just "rare", like you used? (or do you know that those you marked have always been rare?) For readings it's probably best not to use the templates, or we'll have just about every character listed in the relevant categories.. -- Coffee2theorems 23:27, 7 May 2007 (UTC)

The model we have used is 字. (What else? ;-) Robert Ullmann 10:19, 7 May 2007 (UTC)

That entry is a bit simple in many ways at the moment :-) The compound list for Mandarin in 光 is an example of a long list, I tested the template on it using time's "Derived terms" section as a model. What do you think? -- Coffee2theorems 10:41, 7 May 2007 (UTC)

Um, you need to use the -mid4 and -bottom templates to close the table ... see 光. Robert Ullmann 10:51, 7 May 2007 (UTC)

Duh.. Much better. By the way, are the compound lists intended to eventually list everything possible, or why is that one so long? -- Coffee2theorems 23:27, 7 May 2007 (UTC)

They are intended to list everything useful, in the style of Chinese dictionaries that are indexed by characters, with lists of the compounds together with their page references. Some characters/languages have fairly comprehensive lists, most have none. Your definition of "useful" may vary ... Robert Ullmann 13:03, 9 May 2007 (UTC)

The Japanese character dictionaries I've seen (I don't have any Chinese ones, as I don't know Chinese) usually list a few representative ones, nothing as overwhelming as the 光 list. Wouldn't such comprehensive lists be better implemented as a categories? -- Coffee2theorems 14:24, 9 May 2007 (UTC)

About Japanese/Transliteration

You wrote:

• "Protected "Wiktionary:About Japanese/Transliteration": don't want edit war with IP user making non-concensus changes [edit=autoconfirmed:move=autoconfirmed]"
• "this is a policy page; log in, get actual agreement"

Fine, have it your way. There have been numerous comments in opposition to "ī" and diaeresis usage. There is absolutely no opposition to the comments in the past or present. Two edits over a two week period hardly makes an w:WP:EW. If you are so opposed to improving Wiktionary, then so be it. I do not need to waste my time here. Just for reference, you may like to read concensus and consensus. 122.29.162.129 17:43, 8 May 2007 (UTC)

This is not wikipedia. Edits by IP-anon users are usually very suspect, policy page edits essentially always bad news. The few comments were made by unknown IP-anons; this user has refused to create an account and log in. Robert Ullmann 12:42, 9 May 2007 (UTC)

Some strange ones

There are a bunch of transliterations in your User:Robert Ullmann/Mandarin Pinyin/Table that don't appear in Index:Chinese Pinyin - particularly diōu, gèm, iǒng, lüán, lǜn, niōu, shàp, suèi, tóan, zhuěn, and a handful of others. Some of these seem like rather odd constructions for Pinyin - are they non-standard (i.e. slang)? bd2412 T 03:17, 9 May 2007 (UTC)

They are odd indeed. Look at [4] and note that Nanshu's original import of these from the Unihan DB included SHAP4 for 烠. But it certainly isn't in the current Unihan file. (I wish I had a copy of the Unihan file from then, but I haven't found one. Um, maybe the Wayback Machine? I should look ;-)

Part of the impetus behind that table is to highlight these things; scattered throughout 21,300+ entries we'd never find them. I suspect from the general quality of the Nanshu imports that they are simply bogosities. Robert Ullmann 12:59, 9 May 2007 (UTC)

Repeated tags

Regarding your comment on my talk-page: Yeah, I figured as much. :-) —Ruakh_TALK 15:33, 9 May 2007 (UTC)

Oh, good. —Ruakh_TALK 15:43, 9 May 2007 (UTC)

acacia

My apologies, I didn't see those two there; I'll have to be more careful in the future. Thanks for catching my mistake. On a slightly related note, I came up with a question about header levels today, and you seem like a reasonable person to ask on this. If an entry has multiple etymologies, how should the "References" header be placed (assuming the same references were used for the entire entry)? Take a look at οὖρος for example. Atelaes 02:17, 10 May 2007 (UTC)

Appreciation

Hi, Robert. Just a quick note of appreciation. You've done a great job with templates, WT:AJ et al. I especially admire the motivation you expressed on Template talk:infl:

Just about all of the complexity is to make it simpler for the user, all of the parameters are optional in one way or another. A basic principle of computer programming is that modules should hide complexity. Most, if not all, simple forms will work the way you would expect

Anyway, sorry I haven't been around much to help, but thanks. Cheers! Rod (A. Smith) 16:48, 10 May 2007 (UTC)

Category:Persian derviations

Thanks for the information, I didn't realise that was the case. There are some words to which I can add an etymology, but generally speaking I don't feel confident enough to write etymologies. Pistachio 12:54, 12 May 2007 (UTC)

he.wikt

It is an extraordinarily serious problem: the architecture of the various wikts calls for the Hebrew definitions of English, etc. words to go in the he.wikt, just as the English definition of a Hebrew word goes in the en.wikt. The original explanation was, IIRC, that they didn't have enough contributors to do more than Hebrew/Hebrew, as reflected in the quote above. This is—as you observe—self defeating. The people who want to work just on Hebrew should not be getting in the way of people working on other languages; there is no conflict. At some point the WM foundation (meta, stewards etc.) will need to simply over-ride this "policy". Robert Ullmann 15:03, 15 May 2007 (UTC) I should also point out that the Bureaucrat's position as a bureaucrat is way out of line, and you might mention that to him; being a bureaucrat does not entitle one to dictate policy; like sysop, it carries no more authority than any other community member; it is simply that the user is trusted to set sysop and bot flags according to policy. Robert Ullmann 15:03, 15 May 2007 (UTC)

He doesn't dictate, the policy is decided "democratically" by the community. The problem is of course, that the only active members of the community are Heb/Heb editors; FL/Heb editors are not there to vote since they have been driven away according the same policy which they would vote to change had they not been driven away :)

Yet more ;-) point out that the primary reason that they have only 4000 entries for Hebrew after 2 years is that they are driving people away. If people could add FL words, they would want the definition lines to have blue links? Ya think? Robert Ullmann 15:08, 15 May 2007 (UTC)

This argument has already come up, but it doesn't seem to have left any impression on them.

Do you think there is a chance that wikimedia will allow the creation of a parallel he.wikt for people who want Heb/Heb and FL/Heb? Shai 18:47, 15 May 2007 (UTC)

A "parallel" wikt is counter-productive. The situation leaves me rather incredulous: do the people there have any comprehension that driving away contributors is the cause of the utterly pathetic number of entries in the he.wikt? (Is it really only four thousand? That is insane! ONE serious contributor can do that in a few months!) IMHO, needs to be just overridden. (Holy shit! They really have only 4,281 entries!) Robert Ullmann 22:56, 15 May 2007 (UTC)

Almost 4,300 in three years and two weeks; that is almost 4 entries per day :) But that's not fair, most of the time he.wikt had one or two people working on it. Well, maybe one day I'll be able to convince them :) Shai 20:19, 16 May 2007 (UTC)

Japanese templates

Some miscellaneous stuff about Japanese templates:

• I created Template:ja-okurigana and used it at 終る. Does it look OK?
• I think the template Template:ja-verbconj would be better rendered the following way (except with the redundant <big>Conjugation of</big> removed). Would it be OK to make it like this?

    Conjugation of 言う

Conjugation of "RobertUllmann/2007う" (See Appendix:Japanese verbs.)

Katsuyōkei ("stem forms")
Mizenkei ("imperfective")	RobertUllmann/2007わ	言わ	言wa
Ren’yōkei ("continuative")	RobertUllmann/2007い	言い	言i
Shūshikei ("terminal")	RobertUllmann/2007う	言う	言u
Rentaikei ("attributive")	RobertUllmann/2007う	言う	言u
Kateikei ("hypothetical")	RobertUllmann/2007え	言え	言e
Meireikei ("imperative")	RobertUllmann/2007え	言え	言e
Key constructions
Passive	RobertUllmann/2007われる	言われる	言wareru
Causative	RobertUllmann/2007わせる RobertUllmann/2007わす	言わせる 言わす	言waseru 言wasu
Potential	RobertUllmann/2007える	言える	言eru
Volitional	RobertUllmann/2007おう	言おう	言ō
Negative	RobertUllmann/2007わない	言わない	言wanai
Negative continuative	RobertUllmann/2007わず	言わず	言wazu
Formal	RobertUllmann/2007います	言います	言imasu
Perfective	RobertUllmann/2007った	言った	言tta
Conjunctive	RobertUllmann/2007って	言って	言tte
Hypothetical conditional	RobertUllmann/2007えば	言えば	言eba

• Template:ja-readings is missing a line for jōyō readings. Look at e.g. the jōyō table between アーイ: it prescribes not only allowed characters, but readings too. I think these would be best given on a separate line completely independently of the other lines. These are in a processable format at [5], so they could be added by a bot. IANAL, but [6] says that "The following are works which do not form the subject matter of copyright because of their nature to be widely and freely used by the public : (a) constitution and other laws/regulations ; (b) notifications, instructions, circular notices, etc. issued by the national or local authorities". I don't know how to make sure of that, though. (update: there's an incomplete list of the readings at Wikipedia, and on the talk page there are some old posts bemoaning Wiktionary's lack of them :-) (update 2: one more thing, the 付表 stuff should be included too, perhaps on a separate line of "additional jōyō compounds but they could go into their own jōyō compounds section") (update 3: Wikisource has one similar 告示 (for romaji), so I guess it really is kosher to use them)
• The readings template is also missing parameters for old orthography. Japanese kanji dictionaries give these in parentheses, e.g. コウ(カウ) for 校. Perhaps the same would work here too (instead of using parameters)? Parentheses are already used for romaji, though. What do you think is best? (these are only applicable to on readings)

There are some other things too, but those I mentioned in the mammoth post here in the section "Kanji entry structure", which you haven't yet had time to read :-) These are mostly simpler stuff. -- Coffee2theorems 19:57, 18 May 2007 (UTC)

Append mod to pagefromfile.py

Hi Robert. I created a clone with your replacement text. It did no harm but the result doesn't seem quite right. See SemperBlottoBotTest. SemperBlotto 21:40, 19 May 2007 (UTC)

Because your test file didn't end with a newline; you could stick + "\n\n" + in there between old_text and contents, any extra AF will lose. Robert Ullmann 22:38, 19 May 2007 (UTC)

There was a line of code that removed trailing newlines (dor redirects). Removed it and it worked fine. SemperBlotto 07:11, 20 May 2007 (UTC)

Oops, no it didn't - added blank line to top of all new articles. Reverted change and added your mod. SemperBlotto 07:48, 20 May 2007 (UTC)

re Parts of speech in an entry - order of sequence?

Hello Robert Ullmann -- Thank you for replying to my question at the Information Desk. I'm not entirely satisfied with the "alphabetical order" answer and have responded and moved the discussion to Wiktionary:Beer_parlour#Parts of speech in an entry - order of sequence?. If you have any further comments concerning the points in my rejoinder, I'd be glad to hear them. Respectfully -- WikiPedant 04:09, 20 May 2007 (UTC)

"Root" header

Thanks. :-)   —Ruakh_TALK 17:33, 24 May 2007 (UTC)

Contextual labels

Thanks for changing back the πρόβλημα's maths context/category - I put it there in the first place and then removed it, since it seemed out of place amid English terms. I now know that these templates have a lang argument.

In the course of investigating this I came across {{context}} template is there a preference between using {{maths}} and {{context|mathematics}}? —Saltmarsh 06:22, 26 May 2007 (UTC)

We use context if there is more than one label, or for things that have modifiers. {{context|mainly|UK}} for example, yields (mainly UK) and the category. Some people just use context all the time. The only difference between maths and mathematics is a redirect; there are quite a few of those, so that similar names can be used, and redirected to the preferred name (colour/color, railways/railroading/rail transport, non-standard/nonstandard, etc.) and the preferred name can be changed. Robert Ullmann 13:13, 26 May 2007 (UTC)

Thanks —Saltmarsh 13:24, 26 May 2007 (UTC)

one

What happened is that in this edit you changed top to trans-top for Pronoun/accusative, but not the mid or bottom templates. So the rest of the page was collapsed into that table. (trans-bottom closes two div tags that bottom doesn't) Fixed. Robert Ullmann 16:23, 26 May 2007 (UTC)

Ohh! Thank you for fixing it. I couldn't figure out what did it — and that was the first page I looked at today, so I thought at first that Wiktionary might have implemented Hippietrail's language separation software. — Beobach972 17:17, 26 May 2007 (UTC)

ko sort order

I'm pretty sure, given my recall of what we were doing in Unicode with Hangeul blocks, that ㄷ더블린 will give the same net sort order (if used for all entries) as ㄷㅏㅂㅡㄹㄹㅣㄴ. So we could just use hidx=ㄷ and add {{PAGENAME}} in the template? The only visible effect either way is the single letter keys? Would make life easier. Or is there something else we want to do with the hidx key? Robert Ullmann 16:29, 15 May 2007 (UTC)

Ah. Much easier. Rod (A. Smith) 17:13, 15 May 2007 (UTC)

You're right that changing from a sort key of "ㄷㅏㅂㅡㄹㄹㅣㄴ" to a sort key of "ㄷ더블린" does not cause any additional problems. You may, however, be interested to know that MediaWiki sorts neither scheme properly. See Wiktionary talk:About Korean#Category sorting. Rod (A. Smith) 23:57, 28 May 2007 (UTC)

WT:ID

I'm not sure you addressed that question properly. Spoken languages are a thorny issue; the spelling of any given word is debatable, and documentation of them is all either original research, or (generally speaking) a copyvio. Extreme caution is needed when discussing the topic. I understand the desire to be supportive and encouraging, but that must be counterbalanced with caution. --Connel MacKenzie 20:04, 29 May 2007 (UTC)

Yes, I agree. However, this is someone who has gotten nothing but negativity, and can be very valuable. I did some research; she lives there, with the people. It is, IMHO, a good idea to be very supportive, and then watch those issues. Robert Ullmann 22:18, 29 May 2007 (UTC)

239 languages

Where did you get that figure? By my last count, we had about 389 languages with entries here. It even says so on the Main Page. :) Although, once we've gotten all the valid/invalid language headers sorted out, that number could be quite a bit higher. --EncycloPetey 23:46, 29 May 2007 (UTC)

WT:STATS says 223. My L2 report says 323. Not sure which I got 239 from recently. Do we count languages with one entry? ;-) Robert Ullmann 00:08, 30 May 2007 (UTC)

I do. I'm not sure how WT:STATS does its counts. I have the nagging suspicion that it doesn't count languages with a space in the name. I counted from the categories that have entries. --EncycloPetey 00:13, 30 May 2007 (UTC)

It does count them, but doesn't count languages with less than 10 entries. (Which seems reasonable if you are going to blare about the number of languages...) Robert Ullmann 00:16, 30 May 2007 (UTC)

Mopán

Actually it's a bad idea to assume that the butterfly entries are correct. I'm constantly finding errors. Probably, this is because this was a pet project of WF; he apparently scoured the internet for every translation he could find without checking even to see whether the language name was valid. In some cases, they weren't.

Yes, and along the way I found out his real name and address ;-) Just using it as an example, I think the questioner will probably tell us? Robert Ullmann 00:08, 30 May 2007 (UTC)

Also, would it be better to call the language Mopán or Mopán Maya? Ethnologue and Wikipedia use the latter. --EncycloPetey 00:03, 30 May 2007 (UTC)

That is easy to fix Robert Ullmann 00:08, 30 May 2007 (UTC)

Yes, but it makes a difference when I set up {{mop}} so that the language header will register as "Valid". --EncycloPetey 00:14, 30 May 2007 (UTC)

ety levels, defaultsort

See my reply. Rod (A. Smith) 00:27, 2 June 2007 (UTC)

What is is

I have replied on my talk page. Atelaes 05:53, 4 June 2007 (UTC)

Likewise, I have laid out a reply to your question on my talk page. --EncycloPetey 21:35, 5 June 2007 (UTC)

A request for your template skills

A new user recently routed all of the Ancient Greek noun templates through a central blank template (something I hadn't thought of doing myself). I am thinking that this presents a solution to a problem which I have been contemplating for some time. The problem is that Ancient Greek proper nouns generally don't have duals and plurals, only singulars (with few exceptions). Thus far, I've been entering all of them manually, but this is unecessary work, as the forms they do have follow all the same rules as the regular nouns. So, what I'd like to do is enter some code into {{grc-decl-blank}} (the blank template which all others are routed through), so that an editor can add a parameter so that it only includes the singular forms. So, what I was thinking is that perhaps we could put the code from {{grc-decl-blank-sing}} into {{grc-decl-blank}}, and add a switch or something. Another route which might also work is to put a switch into all the noun templates, telling them to access {{grc-decl-blank-sing}} instead of {{grc-decl-blank}} when the parameter is entered. Perhaps the parameter could be something like LIM=SO (limitation is singular only)? What do you think? Atelaes 03:20, 10 June 2007 (UTC)

Sorry for delay. It doesn't help to try to merge the two tables into one template with a conditional; you end up creating two sub-templates anyway! This is because tables and conditionals don't play nicely with each other, as both use | as separators. If the WM parser was straight one-pass recursive descent (which IMHO, it should be ;-) and there was a way to escape | one level, it would work better. But it isn't; it is a series of passes doing different things. (For example, the TOC is built so late in the process that it is parsing the HTML!)

I'd suggest moving grc-decl-blank to grc-decl-dual and grc-decl-blank-sing to grc-decl-sing. (leaving the redirects during the process of course) Then in each of the calling templates, use {{grc-decl-{{{form|dual}}}|...}}. Then the entries that need the singular-only table can use form=sing; form=dual is allowed but not required, and if you come up with another table format, you just add another value. You can use whatever parameter name you want instead of "form" of course, and maybe plural instead of dual. Whatever ;-) Robert Ullmann 13:39, 16 June 2007 (UTC)

Re: "This is because tables and conditionals don't play nicely with each other, as both use | as separators": I've dealt with this problem before. The MediaWiki software does support normal HTML-style table markup, a fact which I've used to advantage in {{he-noun-inflection}} (which only includes a "dual" line if an indefinite dual form has been supplied). —Ruakh_TALK 19:13, 16 June 2007 (UTC)

Okay, I've just read these few comments over, and I must admit that I still don't understand how these switch features work in Wiki markup, as I'd never even encountered them before until several days ago. If I can avoid making lots of extra templates, that would be great. But I would appreciate some help. For now, I'll place the extra templates and categories I've already made on speedy deletion. - Gilgamesh 01:57, 19 June 2007 (UTC)

I moved grc-decl-blank to grc-decl-blank-dual. Now look at {{grc-decl-1st-M-eta}}. It calls the -dual form by default, but if you call it with form=sing it will generate just the singular forms, and if you call it with form=plur it will generate just the plural forms. Look at the change, easy to add to the other specific decl templates as needed, and you're done. Robert Ullmann 02:10, 19 June 2007 (UTC)

Okay, I will try. Though we got our edits crossed because I briefly and accidentally over-edited grc-decl-blank-dual to reflect the rarer dual number. Ancient Greek has three numbers—singular, dual and plural. - Gilgamesh 02:14, 19 June 2007 (UTC)

Hey, should I change template:grc-decl-blank-plur to only use the higher argument numbers that your new template:grc-decl-blank-dual is using? I'm still a tad confused. - Gilgamesh 02:16, 19 June 2007 (UTC)

Yes, to generate just the -plur forms you want to use just the arg numbers in the 3rd column of the full sing/dual/plur form. Exactly right. (I'd say don't try to call the -num one you created, just use the 3 with the tables. All this presumes that we want to generate either sing only, sing/dual/plur, or plural only. Is that right? I think you've got the pattern now; your question is, as I said, exactly right. (sorry it is after 5 AM in nairobi, and I must sleep!) Robert Ullmann 02:21, 19 June 2007 (UTC)

I see you have -full. You've got it! Robert Ullmann 02:24, 19 June 2007 (UTC)

Participle

No, probably not. (Not until we have all words in all languages that ever were, are, or shall be). :) --EncycloPetey 17:34, 13 June 2007 (UTC)

{{trans11}}

{{trans11}} now works fine and can use upto 11 items. Thank you a lot . --Mac 15:15, 16 June 2007 (UTC)

Appeared a problem. See it at http://test.wikipedia.org/wiki/Template_talk:Trans11 . When you give less than 11 parameters, appears blank headers and there is no space between trans-bottom and the next trans-top. --Mac 07:38, 17 June 2007 (UTC)

Fixed, used magic trick to test the existence of parameters, and nested the ifs to get the blank lines right. Robert Ullmann 12:19, 18 June 2007 (UTC)

Missing Spanish Words

Hello, Could you run a process to update a-d, e-m, n-s and t-z

Thanks Bearingbreaker92 16:59, 19 June 2007 (UTC)

okay, but it still as of May 25; WM has had trouble with the dumps. Robert Ullmann 23:40, 19 June 2007 (UTC)

Style on Template:IPA

See Template talk:IPA#Style. Hairy Dude 15:13, 26 June 2007 (UTC)

Re: iwiki links

Oh, sorry. We do this all the time at the Vietnamese Wiktionary, since it's heavily "templatized". (runs and hides) – Minh Nguyễn (talk, contribs) 17:11, 30 June 2007 (UTC)

Your thought

Nicely done. † Raifʻhār Doremítzwr 14:46, 1 July 2007 (UTC)

It does look good. Would be better not centered. Note that I did it by subbing in |}</div></div> which is what rel-bottom does, and then adding the note in between. If we like it, it has to be added into the template, not hacked that way in an entry; just did that to see what it looked like. Robert Ullmann 14:50, 1 July 2007 (UTC)

I’d vote for it. And I agree, the centering is its only shortcoming. † Raifʻhār Doremítzwr 15:07, 1 July 2007 (UTC)

Ah, perfect — no longer centred! Looks to me like the template doesn’t need to be changed then — right? † Raifʻhār Doremítzwr 15:13, 1 July 2007 (UTC)

No, it means I changed it ;-) I added a note on WT:GP asking what other people thought. Only works fully with rel-top4, if you use note= on rel-bottom with (e.g.) rel-top, it will still be centered. Notice I used : at the start of the note= parameter. Robert Ullmann 15:16, 1 July 2007 (UTC)

Noted. What exactly is the purpose of that opening colon? –I’ve seen it in category names as well (where I infer that it prevents the page on which it is written from being added to the said category). † Raifʻhār Doremítzwr 15:32, 1 July 2007 (UTC)

Indent. Just like this paragraph. Try taking it out and preview and you see it won't look as nice. Robert Ullmann 15:35, 1 July 2007 (UTC)

Yes, you’re right, although it would be better justified and with about half its present indentation. † Raifʻhār Doremítzwr 15:39, 1 July 2007 (UTC)

Latin/Roman spelling

Hey Robert. The entries that have "Latin/Roman spelling" should be changed to "Roman spelling". The Latin/Roman spelling was placed there by me when I wasn't sure which word to use for the Roman script. In Serbo-Croatian the script is called "Latin", however in order to avoid confusion with "Latin, the language" (pointed out by Stephen), we have been using the word "Roman". I haven't had the time to go back and fix all of the entries that contain "Latin/Roman spelling". --Dijan 04:04, 2 July 2007 (UTC)

Good. There are 190 of them; better to have AF fix them as it finds them rather than tagging them all. Thanks. Robert Ullmann 11:24, 2 July 2007 (UTC)

A cleanup page?

Do you have a sub-page somewhere that uses DynamicPageList to list the oldest ten entries of each category associated with each cleanup tag? Something like User:Robert Ullmann/Entries needing manual cleanup or something? --Connel MacKenzie 17:02, 2 July 2007 (UTC)

No, I just put them on the categories. Is it useful to present them together? I could steal your table code and put it on AF's user page? Robert Ullmann 17:09, 2 July 2007 (UTC)

Well, there is server penalty for using it, as a result, it sometimes makes "Popups" get goofy, so I think a sub-page would be better. But it would be nice to have a "You can help clean up" page somewhere, that covered all these new cleanup categories. Possibly with an explanation of what needs cleanup for each. --Connel MacKenzie 17:12, 2 July 2007 (UTC)

And yes, I think using the ones I have, is a good starting point. But mine are pretty general categories. I find that having them grouped together is useful. --Connel MacKenzie 17:14, 2 July 2007 (UTC)

Okay; I was thinking of something along this line. I'll think for a while on it. Robert Ullmann 17:16, 2 July 2007 (UTC)

jué

Could you please take a look at definition 85 of jué? I'm wondering if there is there any reason that it's all majuscule. Your robot created it that way, but I'm asuming that's simply because User:NanshuBot created this page (覐) with that information, in majuscule in this edit. I am making the additional assumption that you'll be able to tell whether the information is correct or not and whether it belongs in minuscule letters or not, but if not, we can post this to WT:RFC. PS- I do find it interesting that this entire trail seems to be bot-created... — Beobach972 21:30, 2 July 2007 (UTC)

Sort of seems to be only that entry? NanshuBot was converting the meaning in the Unihan DB to lower case, but not for this one? And it wasn't at the beginning of the run, seems to have been in the middle somewhere. It looks correct, should just be changed to lc. (and edited a bit). A very large number of these entries have only ever been bot edited! Robert Ullmann 17:51, 4 July 2007 (UTC)

Was this meant for me?

Hi. You put this in my talk page, but I'm not sure why:- How about take down? Which isn't described at down/Adverb. We do have the noun takedown. Algrif 16:33, 3 July 2007 (UTC)

যাওয়া

Hi, Can you check if this is the right format for putting derivative terms? thanks --Countincr ^(কথা) 19:34, 3 July 2007 (UTC)

AutoFormat

What's the use of putting horizontal lines between languages with AutoFormat? All languages are heading level 2, so they already have horizontal lines under them. Putting one more line above the heading, in different distance than the line below, creates a big mess. In my opinion, multi-lingual entries having short descriptions of other languages, like film, look bad. There are more lines than text there. --Derbeth ^talk 10:23, 10 July 2007 (UTC)

This is our standard format, AutoFormat is just following it. I don't know what the original reason was, (from before I was paying much attention to the wikt). You can go to WT:PREFS and turn the display off. Robert Ullmann 10:29, 10 July 2007 (UTC)

template:en-noun and regular possessive forms of modern English nouns

There is currently an active vote at [[7]] regarding whether regular possessive forms of modern English nouns should have their own entries or not. As part of this it has been suggested that the {{en-noun}} template might be modified to show the possessive forms in the inflection line of modern English noun entries (irrespective of the outcome of the vote). Your comments and/or votes are welcome until the end of the vote on 5th August 2007. You are receiving this note as you have edited template:en-noun and/or template talk:en-noun Thryduulf 17:23, 11 July 2007 (UTC)

Welsh mutations

Changing the header to Mutation does seem like a good fit with conjugation. I have a feeling I've seen this suggested somewhere else before (perhaps on the grease pit thread I started about the templates - which I will sort!) as well.

Do we need to propose this on the Beer Parlour to see if there are any objections or do we just go for it? If we do agree to change the name, it seems like a good fit to get user:AutoFormat to change any it happens across. Thinking along these lines, I've always put the mutations as a header one below the level of the noun/proper noun - is this standard already? If not should it be? Thryduulf 16:24, 15 July 2007 (UTC)

potentially cool modification to Template:Han ref

• As I was fixing the entry for 㬂, I decided to check in the online Kangxi dictionary (http://www.kangxizidian.com), to see what it says. Then I had a brainstorm, what if I could link directly to the Kangxi page from the Wiktionary entry? I tried modifying your Template:Han ref so that it would link to the correct page at http://www.kangxizidian.com, but was redirected to the index page (I used the php call, ex. http://www.kangxizidian.com/org/orgpage.php?page=498). I thought of another way which might work which is to modify Template:Han ref so that it links directly to the gif image: ex. http://www.kangxizidian.com/kangxi/0498.gif. The problem is that this would require some more ifs and mods in your Template since the number of zeros would change depending on the page number (presumably 0001, 0015, 0397, 1569 etc). Questions:

1. Do you like the idea?
2. Do you think you could modify your script to make it work?

If the answer to one or both of these is no, feel free to revert back to your original version of Template:Han ref. Thanks. -- A-cai 11:29, 17 July 2007 (UTC)

rfc-xphrase

Are you finding a lot of useful things to do other than change the heading? Connel is suggesting that AF just fix the heading. And there are other ways of chasing down all the entries that aren't templated or in the POS cats; not just the ones that have/had phrase in the header. I'm going to experiment a bit with just fixing some of them. Robert Ullmann 20:53, 17 July 2007 (UTC)

For noun phrases I'm also converting to the en-noun template. There are occasionally other changes too. RJFJR 13:56, 18 July 2007 (UTC)

Thanks for answering.

Ok, I now know. Sorry if I caused any troubles. 194.144.87.74 15:15, 18 July 2007 (UTC)

New account.

I don't think it's such a good idea. I don't come here very often for edits, most of the times to just look up some words, so the account would be no use. 194.144.87.74 15:33, 18 July 2007 (UTC)

Well, okay, but it doesn't cost anything ;-). And we do treat edits from registered user differently from IP-anon (in practice, an awful lot of IP-anon edits are crap!) I have a 'pedia account I seldom use, but I log in to it when I edit something. Robert Ullmann 15:38, 18 July 2007 (UTC)

Done. TheBlazikenMaster 15:43, 18 July 2007 (UTC)

Thanks for the vigilance with AutoFormat

Thanks, I always see you fixing format things with that bot, you're doing great things here at Wiktionary. It's great to have bots working on things like indentation and such. By the way, I updated About:Japanese to include the "Reciprocal" header for verbs. So there is no more need for AutoFormat to label those verbs with "rfc-header|reciprocal" or whatever. Hope this helps. I took the rfc-header off those verbs on my watchlist, but I'm not sure if I got them all. Language Lover 03:19, 19 July 2007 (UTC)

Uncountable

Maybe you know. Is Coulomb's law uncountable or proper? RJFJR 14:34, 19 July 2007 (UTC)

I think I have it figured out now: proper, particular law. RJFJR 16:25, 19 July 2007 (UTC)

Searching Text

I've noticed that you are pretty good with bots and automation and such, so I hope I wont bother you with this question.

Suppose that I have a document, "Web Page A", and a database file, "Database". What I want to do is use the list of words in "Database" to search "Web Page A", and then have it return the words that are in "Web Page A", that are not in "Database".

Does that explanation make sense? And how would you reccommend going about that? Any programs or anything?

Thanks Bearingbreaker92 23:40, 22 July 2007 (UTC)

Mandarin modal particles

This seems to solve some of the Pinyin problems - it seems there are some Mandarin onomatopoeia. bd2412 T 01:10, 23 July 2007 (UTC)

link from veil fetish

ok i shan't remove it again. Pistachio 08:46, 24 July 2007 (UTC)

i should have checked it, but i removed it cuz i thought it linked 2 a wierd fetish site. anyway it's ridiculous that that user also put this in category:Islam. Pistachio 08:52, 24 July 2007 (UTC)

Icelandic cardinal numbers

Please stop adding the "no definition line" template to Icelandic cardinal numbers like sjö and átta. They have a definition. --BiT 19:22, 24 July 2007 (UTC)

The problem is the template. The # has to be outside the template, so that the line is identifiable as a definition. So the line would start # {{is-num... Robert Ullmann 05:37, 26 July 2007 (UTC)

Pinyin problems

Can you re-run this page? I think I cleaned up a good portion of these in my last run-through. Cheers! bd2412 T 05:05, 25 July 2007 (UTC)

Yes, many are fixed ;-) Robert Ullmann 06:55, 26 July 2007 (UTC)

Why do the fixed problems still show up on the page? Not a big deal, but I'd like to see the actual scope of the task ahead. Cheers! bd2412 T 05:47, 1 August 2007 (UTC)

Language templates

Templates and coding are not among my strong suits. However, I know that DAVilla was working on templates that had to call and identify languages in much the same way. You might look at DAVilla's edits to the {{context}} and related termplates, or ask DAVilla directly. --EncycloPetey 08:54, 24 July 2007 (UTC)

The {language} template is an atrocity. We need to get it out of {context}, not use it more. It is insane that we can't use the language code templates. Robert Ullmann 06:44, 26 July 2007 (UTC)

Thanks for you reply Robert, I'm a little confused too why we can't work out some way to use the plain language templates. The discussion on WT:BP is going to be moved to to WT:GP so your input would be very much appreciated.--Williamsayers79 07:50, 26 July 2007 (UTC)

{{notred-pt}}

I don't understand. This used for Icelandic language ({{notred-is}}) and for Latin ({{notred-la}}). Why not for Portuguese? LipeFontoura 16:31, 27 July 2007 (UTC)

those are about to get cleaned up too! Robert Ullmann 16:34, 27 July 2007 (UTC)

About to.. it's still being discussed -__- --BiT 22:04, 28 July 2007 (UTC)

Languages used in translations

Using your list as a guide I checked languages Fininsh through Finnsih and Suomi, made sure they were used in connection with Finnish words and corrected the entries accordingly. Similarly I corrected the languages Swedsh through Swesish and Svenska to Swedish. The only Swedish/Norwegian entry I divided into two. I'll do a similar operation with Spanish one of these days, if you don't disagree.

There seems to be something strange with the entry Nicholas. It produces a line in your list for every language used in the translations, although they are correctly spelled.

Is the list complete and reasonably up-to-date? Is it correct to say that there are more translations into Finnish than into any other language? Hekaheka 20:54, 28 July 2007 (UTC)

The list is as of 20 July 2007, as it says at the top. If you look at the entry for Nicholas as of then you'll see what was wrong, and since fixed. By all means sort out Spanish or whatever else you like. And yes, there are more translations for Finnish than anything else, even French! Robert Ullmann 07:55, 29 July 2007 (UTC)

Spanish and Estonian done. Hekaheka 16:54, 6 August 2007 (UTC)

"I can't tell you, I'm waiting for the verb!"

Very funny! I hadn't heard that one. Are you familiar with Tom Weller? It is the single funniest book I've ever read, and sadly it is out of print. You comment brought to mind one of the "projects" in the book. There is a cut-out-and-assemble-it-yourself coffee table edition of Gibbon's Decline and Fall of the Roman Empire. It is a two page spread to be assembled over cardoard with false page edges to make it look like a real book. The idea is that no one (on realizing what book it is) would bother to pick it up and notice that it only contained the two visible pages. The reason tyour comment brought this to mind is that the content of the two-page spread of "Gibbon's" book is part of one ponderously long sentence that stretches across both pages. It is a trube tribute to Gibbon's painfully dull writing and run-on-and-on sentences. --EncycloPetey 05:00, 1 August 2007 (UTC)

Cardinal numbers

My bad. sewnmouthsecret 19:53, 1 August 2007 (UTC)

yes it was

Probably pushing my luck typing this late/early; thanks for catching that. (Did you notice the comment on the next edit? Time for bed I think.) :-) ArielGlenn 13:47, 2 August 2007 (UTC)

Automation

You know, I think the /w/index.php stuff may be coming from a simple bad external link from somewhere. Like a goofy blog somewhere that has "enter topic: ______" (max 30 characters), "enter question: ______" (max 30 characters) then has a bad href= that points here instead of wherever it is supposed to go. (Or, an intentionally misleading/snide: "ask the Wiktionarians" page, or something.) --Connel MacKenzie 04:32, 4 August 2007 (UTC)

I'm fairly sure not; it hits all of the wikts, pedias, etc. And discussion forums all over the net. Note how it often says "Hello from J*anek M*akowski" (less the *'s); there are 10's of thousands of discussion forums that have been hit. Robert Ullmann 11:55, 5 August 2007 (UTC)

Plastic cat food can cover

Yes thanks, that's a better example of the problem. As Groucho Marx says, "Time flies like an arrow; fruit flies like a banana."  ;) --EncycloPetey 18:15, 7 August 2007 (UTC)

Template:fr-verb

Just noting what you are playing with, a few things:

1. it would be better to call it fr-form-of (or whatever), fr-verb looks like what one would use on the inflection line
2. the # have to be outside the template, in the page wikitext, else it doesn't look like a definition line to most tools
3. wouldn't it be easier to call {{form of}} for each case? that's what it is there for, so you don't have to do all the span tag stuff ;-)

Robert Ullmann 15:28, 13 August 2007 (UTC)

I don't think I can place the # outside the template. That would defeat the idea to make it cover multiple homographic forms, such as the 3 listed at gouge, which is why I created it in the first place.

I will see about adding {{form of}} and using a better name once I'm done with the basic implementation (I think I can only implement reasonably well the first and second regular conjugation. My conjugation books analyses ca. 80 different conjugation sets o_O). Circeus 15:56, 13 August 2007 (UTC)

Not putting the # outside the template is not an option. The # MUST be in the page wikitext, not in the template. Robert Ullmann 15:59, 13 August 2007 (UTC)

Oh, silly me. I should be able to just remove the first one when there are multiples. I'll test that when I'm done with -ir verbs.

Oh, and I had no idea about {{form of}}. I've been copy-pasting from what was added tomy forms in gouge. If I'd known, I would have used them (I love a good template, although imbrication sometimes drive me batty). Circeus 16:01, 13 August 2007 (UTC)

Yup, works that way. Now to implement {{form of}}... Circeus 16:11, 13 August 2007 (UTC)

Aaaand done. It's at {{fr-def-verbform}} now. Is it better? I'm testing it on the verb forms of gouger right now and it works superbly so far. Circeus 17:02, 13 August 2007 (UTC)

The form-of templates are not language specific, as far as I am aware. They should not be prefixed with the language code. Also, you will need a separate template for each form. Strike: While I believe this to be true, it may not apply to your template anyway. DAVilla 05:18, 16 August 2007 (UTC)

I have no idea what is at issue with the template. However, I am very new to Wiktionary, so I might just be missing the point entirely. Could you review the discussion at DAVilla's talk page and try to make it clearer for me? Circeus 03:14, 16 August 2007 (UTC)

CheckUser

Hi, I was wondering if you'd consider accepting a nomination for CheckUser. Thanks! DAVilla 10:04, 15 August 2007 (UTC)

I would, except for several considerations:

1. I do have the technical knowledge (I wrote some of the IP specs ;-), but this also means that I can often trace a user without the WM tools.
2. Using IRC is not so easy from where I am; I could hack it, but I haven't.
3. User Rodasmith is a good candidate, has the technical knowledge, has accepted, and I would definitely support him.

pleased to be asked Robert Ullmann 22:53, 15 August 2007 (UTC)

Okay, thanks for considering. You're already more useful to me as a mentor anyways. DAVilla 02:07, 16 August 2007 (UTC)

E flat major

Thank you Robert, that means my references towards en. pages on nl. will follow that convention. nl:Gebruiker:Jcwf

Hi Robert, FYI: there are also errors in the German notation of chords and such here. I asked in the German tea room. The correct orth. is E-Dur etc. (capitalhyphencapital) for major and e-Moll (lowercasehyphencapital) for minor keys.nl:Gebruiker:Jcwf

152.1.193.137 13:49, 16 August 2007 (UTC)

Ahh

Okay, thanks Robert for that. I'll try it. I'm pretty new here, I keep learning new things.

I appreciated the help. Jakeybean 14:42, 15 August 2007 (UTC)

That's useful to know! Wiktionary can be a confusing little fiend at times. ;) Jakeybean 12:05, 17 August 2007 (UTC)

Wait!

I didn't realize that "See also", "References", and "External links" fell under a POS. Does your vote propose that they should? DAVilla 15:14, 17 August 2007 (UTC)

How many bloody, fracking, [many other imaginative expletives deleted] times does the vote have to say that it is NOT, NOT, NOT saying those things should appear at L4, only the order if they do?!!

Besides: the present version of WT:ELE only shows them at L4 under a POS, not at L3. Robert Ullmann 15:18, 17 August 2007 (UTC)

To answer your question, I threw up premature because I wanted to suggest alternative wording. Given the recently posted vote, I confused the idea of what level the headers should be at collectively versus what any particular header should be at. But I understand now and agree with the revert. Just one more minor change, if you don't mind. DAVilla 15:30, 17 August 2007 (UTC)

Okay. I had dropped the text I added back by accident, not intentionally. Fine now. Robert Ullmann 15:33, 17 August 2007 (UTC)

Revoked votes

As you have revoked the votes of several people with this edit (which I am making no comment on), it would be polite for you to leave a note on the talk page of the users concerned so that they are made explicitly aware of this. Thryduulf 23:34, 17 August 2007 (UTC)

Doing that now. They cannot be considered valid because DAVilla changed the text substantively since your (their) vote. Why he can't understand this I don't know. The "vote" on basic header levels is completely invalidated. I had hoped to try one single thing and have it SET AND UNDERSTOOD AND WRITTEN CLEARLY AND NOT CHANGED AND VOTED ON. Robert Ullmann 23:39, 17 August 2007 (UTC)

I told you above that I was going to make another change. "Just one more minor change, if you don't mind." That was completed as of 15:36, just six minutes later. Perhaps I wasn't clear? The only vote it may have invalidated was yours, which is why I told you. I would not, and did not, make changes after any others had voted. If you didn't want me to invalidate your vote, then you shouldn't have rushed to remove {{premature}} from the page. As I told you, I added it for a reason.

In fact, the only person who completely invalidated everybody's vote was you, reverting my changes after others had already voted. If you really prefer your version to mine, then hey, no hard feelings. After all, you went through the process necessary to change the text, by informing all the voters. It was an unnecessary alteration, in my opinion, but entirely your decision.

And I did not "change the text substantially". It went from this to this:

• By "level 4" we mean the level one greater than the POS-type header; if/when a POS-type header occurs at level 4, this sequence refers to level 5.
• By "level 4" we mean the level under the POS-type header. For instance, on some pages this sequence could actually be at level 5.

In retrospect, not entirely perfect, but it fixes a hole in your wording, now reinstated, which is why I might have to oppose it. DAVilla 05:02, 18 August 2007 (UTC)

This is a substantial change. "under" does not have the same meaning as "one greater". You used "under" in the basic headings vote: "The relevant etymology is the etymology section under which the POS heading falls" by which you meant "comes after in sequence". So your change is a serious modification, not some minor rewording.

But the sequence of Etymology–POS is precisely one of dependence. If a POS heading falls under an Etymology section then it is dependent to it, even if that isn't reflected in the hierarchy. Maybe it would have been better to say the headings dependent to the POS or something. I admit "under" isn't perfect wording, but it's not ambiguous to me. I don't see the hole. "By 'level 4' we mean the level", that is, the single level, in which this sequnce falls. For the current practice of having some terms at either levels 3 or 4, "the level under the POS" would be 4. It's not talking about all headers under the POS, only the headers at the level under the POS. Am I missing something? If anything is unclear, it's calling them "level 4 headers" when you don't mean level 4 in the first place.

As clearly evidenced by your decision that you must now oppose the original wording

Well you got me there. I guess I meant it wasn't substantial in that I had preserved your original intent, just releasing one of the assumptions. I know now you don't see it that way, and it was foolish of me not to alert you again, to those changes specifically, those few minutes later. The assumption you made, that the level is one deeper, is currently a point of clarification in the vote. I do agree with the vote in principle. And I'm sorry for making this such a headache. I know for me it's a nightmare I'd rather just learn from and move on.

What are the principles to draw from this? Live votes are dangerous I understand. Votes are not wikis, so let the original vote writer make any corrections to the text, perhaps? Still, I wish people had been more bold on some of my votes. EncycloPetey felt trapped by one of my votes, and thought to correct it only by protesting, and loudly. Other than minor corrections like spelling, no one has ever edited my text, even before the vote started. Do they think I own the vote? Even suggestions for clarification are rare beforehand (only one for me that I can think of). Usually the holes are discovered after the vote has started, and that applies broadly. DAVilla 11:20, 18 August 2007 (UTC)

It is written in terms of existing practice and WT:POS. Your text introduced a hole, not "fixed it". In current practice, these headers occur at two levels (3/4 or 4/5), and the text must specify that it is referring to the headers at POS+1, not the "this sequence" "under the POS-type header". Robert Ullmann 06:59, 18 August 2007 (UTC)

Your message on User talk:DAVilla:

But then you went on making substantive edits and invalidating all of the votes? What are you thinking? Votes are on exact text. Change a punctuation mark and you start over. That's why I want you to write a coherent description of the pronunciation and etymology headers you think we should use. When you "add" an option, you invalidate the whole process. Robert Ullmann 00:17, 18 August 2007 (UTC)

Adding an option... Is this in response to Level of basic headings? The only person who had voted when I added that option was Thryduulf, and I informed him of the change. I really hate having so many options, but the vote would be worse off for not having current practice, and I had mistaken option 5 for that. Technically, by the way, you made a change after he'd voted too. It was a minor one that I wouldn't consider questioning you on, but it does contradict the standard you've accused me of breaking. DAVilla 05:02, 18 August 2007 (UTC)

Oh yes, let's have level 3 heading "Etyomology". And that doesn't "contradict the standard ...". Fixing the spelling doesn't change the meaning of the sentence. Moving/changing a punctuation mark can be, and often is, significant. Robert Ullmann 06:59, 18 August 2007 (UTC)

Okay, I wasn't questioning you on that change. But "meaning" isn't the same as "exact text". Regardless, you do see now that the changes I made were much earlier than you had thought? DAVilla 11:20, 18 August 2007 (UTC)

Kroatian

Sorry, I kind of knew but didn't pay enough attention. I popped into this line Bosnian/Kroatian/Serbian: something, which I divided into three separate lines. Whilst I'm here, I would like to ask a question about your previous comment about "complicated templates":

You wrote: The templating with italbrac or context is important for the consistant display style and user ability to set preferences on it. In undoing it you are creating an inconsistent style. The "of a ..." phrase is a context, not part of the definition sentence of phrase itself. Robert Ullmann 15:27, 8 August 2007 (UTC)

My question: It seems that I may have misunderstood what context -template does. I believed it is equal to categorizing, because writing {{context|nautical|lang=fi}} in the beginning of a line adds the term into fi:Nautical -category. I just thought it would not make sense to create, for example, a fi:Of a difficult situation or state of affairs -category. Did I worry in vain? Or would it be correct in this case to use italbrac instead of context? Hekaheka 22:25, 8 August 2007 (UTC)

see reply at User talk:Hekaheka

Bible templates

Hi. Could you give me a nod when you have any new or modified Bible templates, please. I have an interest in making entries with these categories. Thx. -- Algrif 18:22, 19 August 2007 (UTC)

Re: {{poker slang}}

Thanks for the tweak on {{poker slang}}. I agree that it is better with the space than the comma. I made the changes to {{poker slang}} after dealing with {{vulgar slang}}, which I left with a comma because it originally had one. The main thing I wanted to do with the poker template was preserve the difference between {{poker}} and {{poker slang}} since there is a big difference in my mind between things like draw, straight, flush, and stud on the one hand and dog, trips, quads, and bullets on the other. I'd actually lean toward a separate topic category for Category:Poker slang that is a subcat of Category:Poker and Category:Slang, but poker is not really my area of interest. Mike Dillon 19:17, 19 August 2007 (UTC)

A similar need for distinct categories exists for nearly all slang terms. In the future I think this might be resolved by browsing a unision of categories, e.g. US & informal. On the other hand, that would pull up terms that were specific to the US for one definition and informal in another. DAVilla 03:57, 22 August 2007 (UTC)

misdiagnosises

Oh really? The plural of [mis]diagnosis is [mis]diagnoses. Yes, spelled the same way as the verb form, pronounced differently. Robert Ullmann 16:19, 20 August 2007 (UTC)

Sorry. unCommon misspelling...except by me. I've tagged it for deletion (not common enough to keep, under 500 googles) and corrected misdiagnosis. RJFJR 16:31, 20 August 2007 (UTC)

fake entry

Hi. I notice you are patrolling. This is a "shoot on sight" entry. -- Algrif 18:10, 20 August 2007 (UTC)

I would probably have let it float around for some longer, random amount of time. I have no interest in participating in some anon-IP's "tests". Robert Ullmann 11:22, 21 August 2007 (UTC)

Template:nominative

Is {{nominative}} really a context label? From what I've seen, this template and the {{nom}} redirect are used for consistent linking of nominative case headings in inflection tables, not as context labels. The change to make it a context label have resulted in weird parentheses showing up in a number of inflection tables (e.g. Template:Icelandic interrogative pronouns, Template:la-decl-3rd) and in the addition of a bunch of templates that aren't context labels being added to Category:Context labels. Not sure what the right fix is... Maybe nom/nominative and all the other case template pairs should be split. This should probably including renaming the three letter templates like {{nom}} and {{gen}} to end with a period or something to avoid using a three-letter code. Mike Dillon 15:26, 21 August 2007 (UTC)

I'm not sure why I was adding context to it; it was used as a context label somewhere? I've reverted it. Robert Ullmann 15:42, 21 August 2007 (UTC)

Still not sure; I was working on a list of things that still linked to context/tag; how I got to that one I don't know? Robert Ullmann 15:50, 21 August 2007 (UTC)

There are a number of grammar labels that have that ambiguity. Like the countable/uncountable distinction that is sometimes made in the inflection line and sometimes at the start of the definition, you could make the claim that nearly any grammar label could be used in a similar fashion. DAVilla 03:51, 22 August 2007 (UTC)

As far as Latin is concerned, we've discussed (and by we, I really mean myself and EncycloPetey) phasing out the case templates. I seem to recall one of those templates being used as a context label, but I can't recall which one or where. Medellia 10:48, 24 August 2007 (UTC)

English from Plato

Hello,

it's not from Stephanus translation, it's from Lesley Brown translation, 2005. How should I put it then? 2005, Plato, Sophist. Translation by Lesley Brown. 229a?

Marcot 10:38, 25 August 2007 (UTC)

Yes. Robert Ullmann 10:45, 25 August 2007 (UTC)

Template talk:context x

I'm not particularly fond of putting "of ..." in italbrac, such as with {{context|of a star}}. If this is the standard then so be it, but if not then we shouldn't be promoting it there without first clarifying the point.

The "of a star" is part of the context, not part of the definition. That sort of thing has "always" been written # (''of a star'') .... In that sort of case {italbrac} would certainly work, but {context} should/must handle it as it does, because we don't want (atronomony) (of a star) when there is also a label template.

Regardless, if {{context x}} is not to be moved (seeing so many pages that include it), can we redirect its talk page to the main one? DAVilla 22:10, 25 August 2007 (UTC)

{{context x}} is to be a redirect to {context}. The step after that is to drop the |x in the templates (whenever), they can just say {{context {{{sub|}}}.... It takes a careful sequence of steps over time to make sure at each point we can back up if there is a problem, but at the same time not create double redirects. Robert Ullmann 12:00, 26 August 2007 (UTC)

Chu nom

See w:Chữ Nôm for an explanation of this header. I had a conversation with the individual using it, and I agree with him that we need something, but I'm not sure there's a real English equivalent. I think w:Hán Tự (or an equivalent) works best. --EncycloPetey 18:40, 26 August 2007 (UTC)

I know what it means. But the only English is "Han character" or "Chinese character". For words, the header is the POS of course. For the individual characters, we have been using "Han character" which is already used as L3 header for Translingual. Robert Ullmann 18:44, 26 August 2007 (UTC)

The template used on the repeater line is {{vi-hantu}}. But Han tu (in whatever form) is just not English, not like kanji and hanzi and hanja which are used routinely in English text. (Probably because the Vietnamese usage is entirely archaic.)Robert Ullmann 18:48, 26 August 2007 (UTC)

Although the template might be better named -chunom; the language section is about the use of the characters to write Vietnamese, not the use in Vietnam or by Vietnamese scholars writing Chinese ;-) Robert Ullmann 18:51, 26 August 2007 (UTC)

Is it correct to refer to a native Vietnamese character as a "Han character" or "Chinese character"? —Ruakh_TALK 18:57, 26 August 2007 (UTC)

User Accounts

Hello Robert. I just wanted to clarify with you that this is my sole Wiktionary account. Wikipedia account is w:User:Leitzan and was at one point w:User:MistressBlaed. --Neskaya ^talk 04:30, 27 August 2007 (UTC)

Also you can see the list I have here. Be aware that my email might not have delivered every message that you sent to me, as I only got one from you. You might want to try that again, you might not. --Neskaya ^talk 04:33, 27 August 2007 (UTC)

Block durations

This isn't California...we don't have a "three strikes" law.

That said, I recognize the first two blocks were too short (my fault.) User:Eclecticology set the precedent with Primetime, that the blocks should start out as one day, then go up incrementally by the defaults given in the block selection. (That has changed a couple times, but right now seems to be 1 day, 3 days, 1 week...) Primetime got to skip a few steps when it was shown he was sockpuppetting on open proxies and all his edits were BLATANT copyvios. (It was so cute; he'd argue about formatting needing to be consistent with whatever he happened to be copying from at the time.)

Ruakh though. He's had quite a bad day, methinks. Wow!

--Connel MacKenzie 06:56, 27 August 2007 (UTC)

domestic cats

Hi, I will clean them up over the next couple of days. I have eye-strain issues due to poor vision and heavy computer usage at work -:). Thank you for your input ! WritersCramp 23:52, 27 August 2007 (UTC)

Yes, I am planning to set them up like Siamese cat. WritersCramp 00:40, 28 August 2007 (UTC)

Re: でんせん

Oops. Thanks for your gentle remark. --Izumi5 13:41, 28 August 2007 (UTC)

sospetto

To let you know that I've answered to your question about this word. Please go on my talk page. Cheers. --Barmar 07:15, 2 September 2007 (UTC)

{{t}} again and again

{{t}} does strange things now, e.g. French, when using it with hi or gu. Can you have a look at it? Am I supposed to create the templates {{lang:hi}} and {{lang:gu}}? H. (talk) 14:56, 2 September 2007 (UTC)

Yes, sorry. The lang: templates are needed if the primary template is wikilinked. I was going to check this (and related things) next, but we've just had a 32 hour network outage. (When it breaks on Saturday night, it isn't likely to get fixed until Monday morning. ;-) Robert Ullmann 08:39, 3 September 2007 (UTC)

I thought lang: was deprecated in favor of unlinking. It would be more direct to just unlink all of the language templates so that lang: is taken completely out of the loop. DAVilla 11:13, 4 September 2007 (UTC)

Maybe soon ... I don't think that was completely decided. Perhaps we should ask if anyone (still) objects to un-linking them all? IIRC, Connel objected to re-purposing the templates, and EP (?) was saying that they still get subst'd and it was therefore good to have them linked. In the meantime, the uses of lang:xx should all be inside {language}. Robert Ullmann 11:22, 4 September 2007 (UTC)

Hang on!

I was in the process of trying to execute my first transwiki to Wikitionary (Military reserve) when you seem to have deleted the talk page. This is a complicated procedure and I really don't like being cut off in the middle of it. Can you tell me what I've done wrong? (Buckshot06 at en:Wiki) 90.22.183.189 16:32, 9 September 2007 (UTC)

See your talk page on the pedia; there is a process that preserves the page history. Robert Ullmann 16:39, 9 September 2007 (UTC)

Recent entry re Russian Mountains

Hi. I think you pasted into the wrong place, or something :-) -- Algrif 13:08, 11 September 2007 (UTC)

No, I was commenting on what categories are considered POS or POS like, and named, e.g. "Russian ..." rather than "ru:" or somethings else. Robert Ullmann 13:11, 11 September 2007 (UTC)

Oh OK. I see. :-P Algrif 13:15, 11 September 2007 (UTC)

Language templates

I created a bit of code to sort out where we are; in part because I didn't want to rewrite Wiktionary:Index to templates/languages by hand. See User:Robert Ullmann/Language templates. I'll be updating it again in a day or so when we get the new XML. It tries to catch them all, but if a template isn't in the cat, and there aren't any L2 sections in the wikt, it can't identify it as a language name. (It doesn't look at the trans sections.) Robert Ullmann 10:40, 11 September 2007 (UTC)

Cool. Are there specific ways that I can help you on this? I won't have time until Friday to help much, though. --EncycloPetey 14:33, 11 September 2007 (UTC)

Hello there, great work getting this automated, this has been a pain in the arse for ages! BTW how often will the list be updated by the bot? Often I create new language code templates on my formatting rampages and would like to avoid having to constantly update the list in future.--Williamsayers79 12:27, 12 September 2007 (UTC)

After each XML dump. The interval varies, but we got one 13 August, one 29 August, and the next will probably be tomorrow or so. Robert Ullmann 12:31, 12 September 2007 (UTC)

Note: this does mean that new templates you've added (e.g. 5 Sept) will now have disappeared from the index, but will re-appear presently :-). I'm working on hunting down the templates that (1) aren't in the cat, and (2) don't have any L2 entries, so the bot can't tell that they are language templates! A few were in the index list, have now been dropped; I'm going to compare the lists. Robert Ullmann 12:59, 12 September 2007 (UTC)

Wow. I'd missed 86 of them, not in cats and not in entries. Picked them all up automagically by reading the old code list as well. Robert Ullmann 14:00, 12 September 2007 (UTC)

Things to do: what it says in the last column.

• add lang: templates where needed or unlink the primary (but someone adding a template that needs the lang: will see that, so lower priority)
• delete the un-needed lang: templates
• delete the UC redirects, these have to be checked for links, I've found some in Ety sections
• check some langs that seem to have two 3-letter codes, keep the correct one
• add the cat
• look for the other templates that haven't been found yet and put them in the cat (look in Special:Uncategorizedtemplates ;-)

I've been doing a little bit; it will all get reflected in a new report tomorrow. Robert Ullmann 15:18, 12 September 2007 (UTC)

New report generated as of 14 September ;-) Robert Ullmann 05:51, 15 September 2007 (UTC)

Slovio

Thanks for that. Guess I'll go around and remove the other slovio trans table entries while I'm at it. ArielGlenn 17:16, 12 September 2007 (UTC)

Template:es-verb form of

Hi Robert,

Editing of -slash- arguing over Template:es-verb form of seems to have calmed down now; I don't suppose you'd be willing take a look at its guts at some point and let me know what you think? From a technical standpoint, do you think that's a reasonable way to do things? If not, can you think of a better way?

Thanks in advance!

—Ruakh_TALK 19:57, 12 September 2007 (UTC)

I've looked at it a little bit, and it is pretty awful. Having multiple ways to specify a parameter is very poor design, simply leads to a huge amount of cleanup that will have to be done later. All the named parameters seem excessive; {{es-verb form of|(verb)|first|present|indicative ...}} would seem to be sufficient. The layers of template calls seem un-needed; one can conditionally generate the various words in sequence, and then uc/lc the result, wrapping it in the CSS spans. On a first look, the layers of calls are only because of the (poor design) attempt to allow lots of parameter forms? I'd have to look at it a lot more. Robert Ullmann 14:32, 13 September 2007 (UTC)

Peanut gallery comment: As everyone involved knows, MediaWiki imposes rather severe restrictions on template design. One such restriction, as I'm told, is a reduced level of encapsulation/modularity because of efficiency concerns with respect to nested template calls. Although nesting templates is excellent design from a traditional software development standpoint, I have been told repeatedly that nested templates are inefficient for MediaWiki's parser. Rod (A. Smith) 17:29, 13 September 2007 (UTC)

See the "schools brief" on the GP; nesting can be useful if you do the calls with an understanding of how it works; and the es-form template does a bit of that. But overall es-form is not very good. Robert Ullmann 18:55, 13 September 2007 (UTC)

O.K., well, that's unfortunate. But, fortunately for me, I've just agreed not to deal with those templates anymore, so it's someone else's problem now. ;-)   (In my own defense, the many-alternatives thing was at another user's insistence, and the subtemplate thing was the only way I could see to do that and have the template remain at all readable, since otherwise there would be fifty million copies of every switch statement expression and whatnot. The using-named-parameters-at-all thing, however, was entirely me, and I'd be interested in your reasoning about why it's a bad idea. As for conditionally generating various words in sequence, yeah, I probably should have. The reasoning for this approach was that it allowed the template to be all smart and add cleanup tags if appropriate, but that didn't work out like I'd hoped.) —Ruakh_TALK 20:20, 13 September 2007 (UTC)

nullar and singulative

I knew of nullar number beforehand, but not of singulative. That was interesting to learn — thanks. By the way, regarding how quadral-number words ought to be tagged (if and when such an issue comes up): from what I’ve read (which I shan’t pretend is very much), such words are traditionally considered as quadrals, but are actually used far more loosely, as you explained, as “greater paucals” (in contrast with the trials, which are usually used as de facto “lesser paucals”); however, the uses of both trials and quadrals are apparently restricted to groups of at least three and groups of at least four, respectively. Considering that, and the fact that in the case of the Sursurunga personal pronouns, “[t]he dual, trial, and quadral forms each contain incorporated numeral morphemes (‘two’, ‘three’, and ‘four’, respectively)”^{ref}, I’d say the best solution would be to tag such words as quadral, and then, if necessary, add a usage note explaining the broader usage and/or create a grammatical appendix for the language which explains its use of the quadral number. What do you think? † Raifʻhār Doremítzwr 15:52, 16 September 2007 (UTC) REF: p139 of Analyzing Grammar: An Introduction by Paul R. Kroeger (2005)

Certainly use quadral if the literature does. This is rare enough that it isn't going to be a large issue. Usage notes are always good. Robert Ullmann 14:02, 17 September 2007 (UTC)

That settles it then. :-) † Raifʻhār Doremítzwr 19:47, 17 September 2007 (UTC)

mit Sack und Pack

Noun? SemperBlotto 13:45, 17 September 2007 (UTC)

It had been "Noun phrase", which AF put in a cleanup category, but Connel whacked them all to "Noun" without looking at them. Of course "X phrase" headers may become X, or another POS, or Phrase, or Idiom (even Proverb). But now that Connel changed all these to Noun and removed all the rfc-xphrase tags, they are lost unless we notice them at random.

I hadn't looked carefully at the POS ;-) I was taking {t} out of the "Similar" section and renaming it ... Robert Ullmann 13:58, 17 September 2007 (UTC)

Sesotho wiktionary support

Hi,

We talked a while back about the Sesotho wiktionary: http://st.wiktionary.org/wiki/User_talk:Jakoli4

Since then my admin status has expired - I've put in another request what do I do from here? http://meta.wikimedia.org/wiki/Requests_for_permissions#st.wikitionary.org_Sesotho_Wiktionary_Administrator_Request

Thanks! --JAKoli4 06:09, 18 September 2007 (UTC)

Just so you know if you come here; I'm not ignoring this, it is a high priority for me; will do something soon. Robert Ullmann 23:05, 26 September 2007 (UTC)

User talk:Robert Ullmann/Missing

Ooops, I commented on the subpage talk page (it wasn't a redirect to here, but possibly should be.) --Connel MacKenzie 02:50, 19 September 2007 (UTC)

(from subpage, now redirect:)

This is really cool, but can wikified terms be treated separately, like page d'accueil (in page accueil)? It would be really nice if things like that were not broken down into sub-components. --Connel MacKenzie 02:49, 19 September 2007 (UTC)

Yes, I should improve the line parsing, and run it again. Right now it just strips the wikilinking and then parses words. It should be taking out the wikilinks separately first, e.g. "[[parse]]s" should be treated as a link to "parse". not hard to do ... Robert Ullmann 12:51, 19 September 2007 (UTC)

Done. Robert Ullmann 14:37, 20 September 2007 (UTC)

Hmmm. There was something else I wanted to mention there, that I'm sure I forgot to. --Connel MacKenzie 16:13, 20 September 2007 (UTC)

Oh, I remember now: FANTASTIC, DUDE! Very well done! --Connel MacKenzie 16:14, 20 September 2007 (UTC)

Languages in the gray area

Scots is not a gray-area langhuage; it shouyld always be linked. Most people in America don't realize that there is more than one language in Scotland, and many think that Scots and Scottish Gaelic (aka Scots Gaelic) are the same language. --EncycloPetey 14:33, 20 September 2007 (UTC)

Good point. Sometimes I forget I know too much. And we went through this when I ran the 'bot to convert to Scottish Gaelic. Robert Ullmann 14:39, 20 September 2007 (UTC)

Interwiki bots

Where do we say that only one is allowed? If we do say that, I think we should reconsider it. I know I had reservations about AF adding that to its very long list of tasks, but I have not objection to, say, an UllmannInterwikiBot that runs separately. The more I think about it, though, the more I think it should be rewritten to use the Toolserver databases. (I.e. just lookup on toolserver tables, with no httpd hits, to find matches.) --Connel MacKenzie 16:58, 21 September 2007 (UTC)

WT:BOT explicitly resricted interwiki to GeraldM; I changed that to VolkovBot according to the recent vote. It has been explicit policy that only RobotGMwikt was permitted. (So technically we now allow 2, assuming no objections if Gerald's bot re-appeared.) Robert Ullmann 11:11, 22 September 2007 (UTC)

Thank you. I've reworded it slightly (as per the VolkovBot vote) - look good to you, now? What do you think of people's recent requests to have AF do interwikis? I think it is unwise to commingle so many disseparate tasks and would be infinitely more efficient on the toolserver. --Connel MacKenzie 15:27, 22 September 2007 (UTC)

Do you have a toolserver account yet? --Connel MacKenzie 15:33, 22 September 2007 (UTC)

I think they were talking about Tbot, not AF; but either way, it is a separate task. There isn't any reason to glom them together. Should definitely either run on the toolserver, or with a complete set of the article title downloads. (I have one other idea I'm thinking about...). No, I don't have a toolserver account; I'll go look at it. Robert Ullmann 15:44, 22 September 2007 (UTC)

WARNING: Toolserver approval is done either monthly or quarterly (depending on DaB's mood and how many people are nagging him on irc://irc.freenode.net/wikimedia-toolserver )...so get the request filed ASAP. These days we should probably have two dozen Wiktionary accounts there, but instead mine seems to be the only one left. --Connel MacKenzie 15:52, 22 September 2007 (UTC)

I need template help

I started writing a guide to my Ancient Greek IPA token templates, but as I've been expanding it, the last few entries are choking on me. :( It's at Category:Ancient Greek IPA tokens. Can you help out with suggestions, recommendations, etc.? - Gilgamesh 20:42, 21 September 2007 (UTC)

You're hitting the pre-include template size limits. I'll take a look at the templates and see what can be done. Mike Dillon 00:49, 22 September 2007 (UTC)

I didn't even know that was possible. Anyway, I've been thinking about it, and I think I know another way to explain the format of the tokens. I'll describe them in human language to the best of my ability, explaining how to form them and what is supported, rather than just listing every single one's contents. - Gilgamesh 01:14, 22 September 2007 (UTC)

As I told you, the individual templates should not be calling each other! Just put the text in the switch as in the example I wrote on your talk page, and then what you are trying to do should work. As it is, you are making a ridiculous number of template calls.

Do you understand that the way you are doing it, {{grc-ipatok-eeuph}} (for example) expands to hundreds of template calls? (look at this which is only after two rounds of expansion, there is more to go.) WM parses every path. It MUST be just this:

{{#switch:{{{1|cla}}}|cla=ɛːʊp|koi=eːʍpʰ|byz=iff|con=i(f)f}}<noinclude>
[[Category:Ancient Greek IPA tokens|eeuph]]
</noinclude>

The -guiderow thing is just right, it shows all of the calls in the table so it is easy to check. (And I wish you would use "mod" for modern, not "con", "contemporary" doesn't mean now, it means at the same time as. In reference to ancient Greece, it would mean then ;-) Note I put a sort key in the category, else they will all be listed under "T" Robert Ullmann 10:42, 22 September 2007 (UTC)

Daily/weekly changes

Since you are already doing it, I have to wonder if I should too. Can you generate a list (daily or weekly) of all pages that have been modified that week? Using such a list on Special:Export once a day or once a week (for current page versions only) would generate a usable "XML dump delta" file, keeping (in theory) all XML-dump related activities up to date, between XML dumps. Does that sound interesting to you? (I'm not sure if I'm just thinking out loud, or what. We didn't have Special:Export (working) a year ago, so it wasn't feasible then.) --Connel MacKenzie 15:02, 22 September 2007 (UTC)

fountain

You have only partially reverted it! SemperBlotto 16:07, 24 September 2007 (UTC)

Yes, chasing the other user. fixed now. Robert Ullmann 16:14, 24 September 2007 (UTC)

Trans language

The analysis of languages used in translations sections (as of 14 September 2007) is really useful. Is it possible to use the Pythoncode on xml-dumps on other wiktionary's too? I would like to have the same oversight from time to time for the dutch wiktionary. raymondm / 84.87.46.225 23:36, 24 September 2007 (UTC)

Yes, but it may take a bit of modification for the specific format; post a message with your nl.wikt id or whatever, i'll be happy to mail the python code I use Robert Ullmann 23:41, 24 September 2007 (UTC)

That's great. My nl.wikt ID is raymondm. I'll ask the sysop of the nl.wikt if she can help out.

I got the XML dump now, but somehow I get an errormessage when I try to email you for the pythoncode and use the link E-mail this user. Maybe it works when you email me? Raymondm 10:08, 25 September 2007 (UTC)

Hm, there is a checkbox for "enable email from other users". Well duh, why else would I be putting it in? and it seems to default to "off", of all things! Should work now. Robert Ullmann 11:16, 25 September 2007 (UTC)

Ongoing language template saga....

Just to let you know that I have been categorising loads of the plain language templates not in Category:Language templates, I hope to be finished this soon so your XML dump and bot creation of the languages list should include a few more next time.--Williamsayers79 22:03, 25 September 2007 (UTC)

Good, I note you are probably working from the special uncat templates list? I need to keep deleting the bad redirects, should do some more, we'll see where we are in a couple of days. Robert Ullmann 22:47, 25 September 2007 (UTC)

That's all of them on that list now categorised. BTW, do you know how often special uncat templates is updated?--Williamsayers79 20:34, 26 September 2007 (UTC)

The cached specials lists are about twice a week. We should have new XML in a few days, and I'll run that again. Robert Ullmann 23:04, 26 September 2007 (UTC)

redir

okay, roger that ;) Mallerd 14:28, 29 September 2007 (UTC)

pronouns

WT:AJ#Nouns (名詞): Note that pronouns (代名詞) should be considered a subclass of nouns, and not a separate part of speech.

代名詞 don't function differently from nouns (強いあなた "the strong you" => you who are strong), and for the most part are derived from actual nouns (boku was originally something like "the manservant", watashi "the private individual", etc). Do we want to change the header for them now? Cynewulf 17:10, 4 October 2007 (UTC)

My dictionary (Japan Foundation) calls "boku" a pronoun. And certainly a part of speech. Even if they can be considered a sub-class of nouns, from an English perspective (which is what we are about ;-) we want to identify them as pronouns. Note the POV text immediately above whining that "western" models can't possibly apply to Japanese. (Which I really should have deleted; I spent an enormous amount of time cleaning up that doc when I should have just written it anew.) Robert Ullmann 15:02, 5 October 2007 (UTC)

OK, since I lack your dictionary, what makes a pronoun? Anything deictic? There's lots of them -- all the ko/so/a/do (これ etc), all the personal ones... Is この a pronoun or a prefix (it's a 連体詞)? Then just deictic nouns? Or do we just pick some dictionary and go by what it calls daimeishi? (jdict calls boku a noun, though not [8]) I don't object to identifying them, but how do I identify them?

I don't see how "boku" is a different part of speech from, say, "hito" -- can you show a sentence where one would be grammatical and the other not? Using a header to clarify a meaning is fine, but I'm not aware of a difference in word function. Maybe just "a pronoun is a noun that can translate to an English pronoun"... Cynewulf 17:13, 5 October 2007 (UTC)

Japanese grammar is said not to distinguish between personal pronouns and nouns with human referents. There has been quite alot of discussion about how to deal with languages whose grammars differ significantly from English. One set of editors seems to think that readers would be best served by translating foreign languages onto English, then describing the foreign words in terms of layman English grammar (i.e. something that would be taught in a high school English class). The other editors would prefer a technical linguistics approach that follows the foreign grammar as accurately as possible in English terms used in linguistic studies of the foreign language. One option that hasn't really been discussed is to create a WT:PREFS setting that lets readers toggle between “layman” and “linguistics” labels. Then, we could create tags that show the appropriate view to the reader. In this case, the “layman” view would display “Pronoun” and the “linguistics” view would display “Noun”. Does this seem worthy of a pursuit, beginning with a WT:BP conversation? Rod (A. Smith) 18:01, 5 October 2007 (UTC)

That seems like a lot of trouble -- I'd be fine with just a note in WT:AJA to the effect that "pronouns do not behave differently from other nouns in Japanese grammar" and either to require the use of header Pronoun, or to require header Noun and tag (pronoun); "they are marked for the convenience of our English-speaking readers". Like calling 連体詞 prefixes, we can make our own jargon as long as it's explained properly. (maybe we should move this to BP anyway) Cynewulf 18:19, 5 October 2007 (UTC)

If the defining characteristic of a Japanese pronoun is that you'd translate it to English using an English pronoun, then it doesn't seem like there's any need for a context label: the translation will be clear enough. Using ===Pronoun=== has the advantage of being more precise, and the disadvantage of treating pronouns as their own part of speech, when really they're just a kind of noun. On balance, neither seems to make a big difference; the important thing is that it be documented and consistently followed. (I think the easiest is probably to use ===Pronoun===, and have Category:Japanese pronouns continue to be a subcategory of Category:Japanese nouns). —Ruakh_TALK 21:45, 5 October 2007 (UTC)

This is quite simple really: this is an English dictionary of the Japanese language. They should be shown as Pronouns. A Ja/Ja dict will indeed call then "nouns", but in a Ja/En dictionary they are properly pronouns. (And as I commented on Rod's talk page, creating some magic layman/linguist template is out of the question; it will break hundreds or thousands of software programs.) There is a simple right answer here. Even if you had a layman/linguist view, the layman view would show "Pronoun" because that is the simple grammar POS in English, and the linguist view would show "Pronoun" because that is what 代名詞 means (go look: it says: (grammar) pronoun :-). Note that even the quoted text from WT:AJ starts by saying that they are pronouns ... Robert Ullmann 19:17, 5 October 2007 (UTC)

Your point about breaking programs that reuse our database is well taken. Even if the layman/linguist problem is not critical with respect to 代名詞, there are several labels in contention. Some editors prefers English layman terminology but others prefer linguist terminology. Rod (A. Smith) 19:30, 5 October 2007 (UTC)

Sesotho wiktionary

Robert,

Sorry to trouble again...

I'm trying to get some support to get the Sesotho Wiktionary going - I've added a lot of terms myself already.

See my application for adminship: http://meta.wikimedia.org/wiki/Requests_for_permissions#st.wikitionary.org_Sesotho_Wiktionary_Administrator_Request

We need to vote on this matter on the Sesotho Wiktionary: http://st.wiktionary.org/wiki/Wikitionary:Administrators

Thanks!

Kind regards --JAKoli4 08:19, 8 October 2007 (UTC) http://www.sesotho.web.za/

Bible template

I almost did this myself:) Will changing references to {{Bible}} be automated or is help needed? I have also raised a query on the new template's talk page. —Saltmarsh^Talk 13:47, 9 October 2007 (UTC)

Replacements running now. Robert Ullmann 13:49, 9 October 2007 (UTC)

Context cat.

So, the idea's to replace this:

 … <!--
  if topic category
 -->{{#if:{{{topcat|}}}|{{#if:{{{skey|}}}|
 [[Category:{{#if:{{{lang|}}}|{{#ifeq:{{lc:{{{lang}}}}}|en||{{lc:{{{lang}}}}}:}}}}{{{topcat|}}}|{{{skey}}} {{PAGENAME}}]]|
 [[Category:{{#if:{{{lang|}}}|{{#ifeq:{{lc:{{{lang}}}}}|en||{{lc:{{{lang}}}}}:}}}}{{{topcat|}}}]]}}}}<!--
  if pos or pos-like cat
 -->{{#if:{{{poscat|}}}|{{#if:{{{skey|}}}|
 [[Category:{{{{#if:{{{poscat|}}}|language|context 0}}|{{lc:{{#if:{{{lang|}}}|{{{lang}}}|en}}}}}} {{{poscat|}}}|{{{skey}}} {{PAGENAME}}]]|
 [[Category:{{{{#if:{{{poscat|}}}|language|context 0}}|{{lc:{{#if:{{{lang|}}}|{{{lang}}}|en}}}}}} {{{poscat|}}}]]}}}}<!--
  regional templates
 -->{{#if:{{{regcat|}}}|{{#if:{{{skey|}}}|
 [[Category:{{{regcat}}} {{{{#if:{{{regcat|}}}|language|context 0}}|{{lc:{{#if:{{{lang|}}}|{{{lang}}}|en}}}}}}]]|
 [[Category:{{{regcat}}} {{{{#if:{{{regcat|}}}|language|context 0}}|{{lc:{{#if:{{{lang|}}}|{{{lang}}}|en}}}}}}]]}}}}<!--

with this:

 … <!--
  if in article namespace, add cats
 -->{{#if:{{NAMESPACE}}||<!--
  if topic category
 -->{{#if:{{{topcat|}}}|{{#if:{{{skey|}}}|
 [[Category:{{#if:{{{lang|}}}|{{#ifeq:{{lc:{{{lang}}}}}|en||{{lc:{{{lang}}}}}:}}}}{{{topcat|}}}|{{{skey}}} {{PAGENAME}}]]|
 [[Category:{{#if:{{{lang|}}}|{{#ifeq:{{lc:{{{lang}}}}}|en||{{lc:{{{lang}}}}}:}}}}{{{topcat|}}}]]}}}}<!--
  if pos or pos-like cat
 -->{{#if:{{{poscat|}}}|{{#if:{{{skey|}}}|
 [[Category:{{{{#if:{{{poscat|}}}|language|context 0}}|{{lc:{{#if:{{{lang|}}}|{{{lang}}}|en}}}}}} {{{poscat|}}}|{{{skey}}} {{PAGENAME}}]]|
 [[Category:{{{{#if:{{{poscat|}}}|language|context 0}}|{{lc:{{#if:{{{lang|}}}|{{{lang}}}|en}}}}}} {{{poscat|}}}]]}}}}<!--
  regional templates
 -->{{#if:{{{regcat|}}}|{{#if:{{{skey|}}}|
 [[Category:{{{regcat}}} {{{{#if:{{{regcat|}}}|language|context 0}}|{{lc:{{#if:{{{lang|}}}|{{{lang}}}|en}}}}}}]]|
 [[Category:{{{regcat}}} {{{{#if:{{{regcat|}}}|language|context 0}}|{{lc:{{#if:{{{lang|}}}|{{{lang}}}|en}}}}}}]]}}}}}}<!--

, right? (At Template:context, I mean.)

Thanks for your help,
—Ruakh_TALK 16:05, 10 October 2007 (UTC)

Is not that simple, the template also uses the same code to put the context label templates themselves at the top of the categories. But something like that. I was looking at this about an hour ago, and decided to noodle it a bit longer. (There are two other things to fix, and it is kinder to the servers to do just one edit ;-). Robert Ullmann 16:12, 10 October 2007 (UTC)

To do it in essentially that structure, we would use {{#switch:{{NAMESPACE}}|Template|=, e.g. if NS is template or empty, use the following. But as I said, I am going to think this through carefully. Robert Ullmann 16:30, 10 October 2007 (UTC)

O.K., then I'll just wait for you to take care of it. Thanks again. :-) —Ruakh_TALK 16:31, 10 October 2007 (UTC)

{{t}}

Robert - is it possible (or perhaps should say "easy"!)to enlarge the appearance of the output from {{t}} when the Grek script is being used? as in Πράξεις f pl (Práxeis) —Saltmarsh^Talk 14:58, 12 October 2007 (UTC)

svwikt translation

Svwikt translation template {{ö}} is updated now... Will the bot be able to transform [[<word>]] to {{ö|<langcode>|<word>}} and {{ö-|<langcode>|<word>}} respectively, or do you need me to prepare something on svwikt pages to make the tweak in the code for you possible and make the code compatible with svwikt? ~ Dodde 02:44, 13 October 2007 (UTC)

Excess User templates categories to delete

As per your request, the Category:User templates subcategories are no longer needed (they were admittedly somewhat superfluous, but I was just following the Wikipedia's pattern). Unfortunately, I do not have category deletion privileges, so could be so kind as to take care of that? The categories to delete are:

• Category:User templates zh
• Category:User templates vi
• Category:User templates th
• Category:User templates te
• Category:User templates sv
• Category:User templates sr
• Category:User templates sq
• Category:User templates sh
• Category:User templates sco
• Category:User templates ru
• Category:User templates ro
• Category:User templates pt
• Category:User templates pl
• Category:User templates non
• Category:User templates no
• Category:User templates nn
• Category:User templates nl
• Category:User templates nds
• Category:User templates mo
• Category:User templates lt
• Category:User templates li
• Category:User templates lb
• Category:User templates la
• Category:User templates ku
• Category:User templates ksh
• Category:User templates ke
• Category:User templates ja
• Category:User templates is
• Category:User templates id
• Category:User templates hu
• Category:User templates hr
• Category:User templates hi
• Category:User templates gil
• Category:User templates gd
• Category:User templates ga
• Category:User templates fy
• Category:User templates fr
• Category:User templates fo
• Category:User templates fiu-vro
• Category:User templates eu
• Category:User templates et
• Category:User templates es
• Category:User templates eo
• Category:User templates en
• Category:User templates el
• Category:User templates de
• Category:User templates da
• Category:User templates cs
• Category:User templates ca
• Category:User templates bs
• Category:User templates bn
• Category:User templates bg
• Category:User templates ar
• Category:User templates ang
• Category:User templates an
• Category:User templates am
• Category:User templates als
• Category:User templates af

Thanks. Urhixidur 20:47, 16 October 2007 (UTC)

'missing' word report

I added a link to User:Robert Ullmann/Missing to Wiktionary:Requested articles under see also since it already listed the inactive Special:Wantedpages. I hope you don't mind. I'm finding the report you generated very useful. RJFJR 02:38, 19 October 2007 (UTC)

That's fine, thank you! (and time to run it again) Robert Ullmann 07:14, 19 October 2007 (UTC)

Wash - Thanks

I appreciate the corrections. I find it difficult to keep track of all the formatting standards on WP and, moreso, on Wiktionary. Unfortunately, like most folks, I learn best by really doing, not by playing in a sandbox and not by studying rules. Would more inclusive templates help folks like me? I think so. I wonder how hard it would be to have a template that created all of the headings in a way that they wouldn't appear (except when editing) if there wasn't any content entered beneath the heading. Hmmm. Some kind of conditional HTML? DCDuring 16:43, 20 October 2007 (UTC)

re:tr=

Thanks for the tip! I love that template. It's so great for those languages without their own infl-templates. :-) — [ ric | opiaterein ] — 16:40, 24 October 2007 (UTC)

feature for missing list

On User:Robert Ullmann/Missing it would be nice, for entries that are misspellings, to be able to run the wiki search on the word to see all the places it is misspelled by clicking a link. Unfortunately, I'm not sure the increased size of the page, which would slow loading, is paid for by the convenience. RJFJR 22:48, 27 October 2007 (UTC)

How does it know which are misspellings? (;-) yes, adding a link for every one would be a lot. It already has trouble with the number of links on each page. Robert Ullmann 13:43, 28 October 2007 (UTC)

I think it would have to be all of them. I'll stick to typing it into the search box. (I may start opening another widnow though). It's a pretty big page already. Thank you. RJFJR 18:47, 28 October 2007 (UTC)

Translation tables

In thinking about a new version of Wiktionary:Votes/2007-10/Lemma entries, I need to be sure to address concerns like the one you expressed in that vote. A big concern of mine is from your comment, “The forms may have different translations depending on tense and/or sense used only by that form”.

Hopefully you understand that if an inflected form of an English word has additional senses absent from what would otherwise be considered the lemma entry, that form is to be given full treatment. I will endeavor to make that more clear in the next version of that vote.

That would be good, as the present vote language strictly prohibits that! Now, what happens when someone expands an entry that doesn't "require" that full treatment to make it more useful? (i.e. adding the useful simple translations, not dumping the FL conjugation table in ;-) Is ELE going to make it very, very clear that the "extra" content is not to be removed? Robert Ullmann 23:13, 2 November 2007 (UTC)

The proposal says that pluralia tantum are to be treated as lemma entries. I will make it more clear in the next version. If a contributor expands a non-lemma English entry (e.g. (deprecated template usage) talked) by adding a translation table, that translation should be moved into the lemma entry (e.g. (deprecated template usage) talk). Otherwise, we will end up with other editors wasting their time trying and failing to translate the nuances of foreign language person, number, tense, aspect, mood, and formality within a translation table. It should be very clear that translation tables are not the proper place for such details. Such "extra content" is not only unnecessary, but it damages this project and should be moved to the proper entry. Rod (A. Smith) 23:24, 2 November 2007 (UTC)

I'm not talking about dumping the FL conjugation table into the translation table. I'm talking about things that are useful, but that this vote language prohibits. see below Robert Ullmann 00:23, 3 November 2007 (UTC)

For translations, are you really under the impression that English entry translation tables are a suitable place to show multiple persons, numbers, tenses, aspects, moods, and formality levels of each foreign lexeme or do I misunderstand your comment? Rod (A. Smith) 17:31, 2 November 2007 (UTC)

Why would they have all that? And I said no such thing! As you note I said "The forms may have different translations depending on tense and/or sense used only by that form." which is not at all what you are expanding it to above. MAY, MAY, MAY!!! (;-) It is (very probably) not useful to try to translate (say) an English past form to all of the past forms in a heavily inflected language; it is useful to translate in some cases where the form depends on the sense, or a simple (particularly when obscure) tense in the FL, or the tense is a different word in the FL, e.g. what I said. Robert Ullmann 23:13, 2 November 2007 (UTC)

I don't understand. Please give an example of a translation we should keep from a non-lemma English entry (i.e. an entry without unique senses beyond the lemma form). Rod (A. Smith) 23:25, 2 November 2007 (UTC)

I don't know of any FL language with an interesting irregularity for "talked", just the simple translation to past:

• Swahili: -linena, -lisema, -liongea

"nena" means speak or articulate, "sema" is talk, say, "ongea" is chat, gossip. "-li-" is simple past tense. In this case these might very well be linked to the lemma forms, as they aren't complete words. "tuliongea" is we chatted. So now the user can see which tense marker they want. They aren't going talked->talk->ongea->(what tense were we at?). Do remember that most people (in English) have no clue what tenses are beyond "present, future, past", if that.

In the FL wikt, with the same principle applied; which may be a clearer example, because it is almost always going to a simpler tense system: (e.g. in sw.wikt at nilisema):

• Kiingereza: I talked

Robert Ullmann 00:23, 3 November 2007 (UTC)

So, you do advocate a translation table on (deprecated template usage) talked? And you would be happy to see this in it?

• Swahili: -linena, -lisema, -liongea

That's terrible. I have no idea from that when to use which form. Worse, some editors will see that translation table and waste time translating "talked" into various languages. By not discouraging such waste, we would be encouraging it. Still worse, a reader who wants to translate "walked" into Spanish will have no clear place to turn. If the reader has seen the entry for (deprecated template usage) talked, she might expect to see the translation at (deprecated template usage) walked, but that entry will not clearly tell her to look at (deprecated template usage) walk for the translation. Translations from English belong in the primary English entry. The various forms of the translated target word can be effectively listed nowhere outside of the English Wiktionary's FL entry. Rod (A. Smith) 01:03, 3 November 2007 (UTC)

Re "trying to get people to "focus on" what you want is totally non-wiki", do you oppose WT:CFI? Rod (A. Smith) 23:32, 2 November 2007 (UTC)

Why do you leap to a conclusion only about a thousand miles beyond what I said? I don't "oppose CFI", don't be ridiculous. I do think it ought to be expanded, but it isn't there to "focus" attention; it is there to define the outer boundary of what the en.wikt is. Totally different concept from saying what useful content that is clearly within the project users are allowed to work on.

Come on Rod, just flip the mindset from creating restrictions to allowing people to create expanded entries (the status quo pro ante). We don't put conjugation tables in translations sections on lemmas either. The non-lemma entries can be just the same, the most useful FL word and tense. ELE is just fine, none of this is needed!

The perceived problem arises because the 100,000+ "form of" non-lemma entries do not follow ELE: they lack definitions/translations!

All we would need to do to resolve that is to note that some non-lemma entries may be in the simplified stub form if they have not been expanded. That's it. Nothing else! Robert Ullmann 00:30, 3 November 2007 (UTC)

Or, what I would prefer: treat those stubs as inadequate, that must be improved to put "xx form of (lemma)" on the inflection line, and the English translation on the definition line so that they then meet the present ELE. Robert Ullmann 00:38, 3 November 2007 (UTC)

I agree that "xx form of yy" is a poor definition. The proposal is not about forcing such definitions. The new proposal will improve on that detail by encouraging a brief gloss and a clear link to the primary entry. The proposal is about consolidating certain details (form-invariant nuance definitions, translations, synonyms, etc.) into primary entries. We do not and should never put foreign conjugations in English translation tables because it's only effective to consolidate them into the en.wikt foreign term lemma entry. CFI and ELE tell us to move content from a page like eat dinner to more effective locations (the entries (deprecated template usage) eat and (deprecated template usage) dinner), just as they should tell us to move translations, synonyms, and nuance definitions from (deprecated template usage) eats to (deprecated template usage) eat. Rod (A. Smith) 01:03, 3 November 2007 (UTC)

Re "move[ing] content: it should be noted here that neither CFI nor ELE say anything of the sort. Robert Ullmann 12:02, 7 November 2007 (UTC)

CFI clearly is our instrument for determining which entries stay and which entries do not. If a contributor adds an entry eat dinner with definitions that consist of synonyms of eat and dinner, other contributors would move those synonyms into (deprecated template usage) eat and (deprecated template usage) dinner because CFI does not cover combinations like eat dinner and because ELE tells us where in an entry synonyms belong. Did you really need me to explain that, or are you just being difficult? Rod (A. Smith) 16:14, 7 November 2007 (UTC)

Yes, it says which stay, and which do not. If someone creates an entry for "eat dinner" that consists of synonyms for "eat " and "dinner", it just goes away. CFI does NOT SAY IT IS "MOVED". You are pretending that they say that so that you can justify your desire to dumb-down non-lemma entries by moving content. Neither ELE or CFI make any reference whatsoever to MOVING content to different entries or forms. You apparently think they should, but they DO NOT NOW. In plain language. The reason they don't is that we do not, and never did want, anything less than full entries whenever someone wants to take the time. Robert Ullmann 18:18, 7 November 2007 (UTC)

No shit. I want full entries, too. That does not include misleading, unusable translation tables. Rod (A. Smith) 18:24, 7 November 2007 (UTC)

Thanks for good edit summary help

On Wiktionary I learn much more by doing than from the documentation. I learn most from constructive advice in edit summaries and on talk page. DCDuring 11:54, 4 November 2007 (UTC)

Reasonable page size

I've just finished preparing a concordance of words to check coverage of the wiktionary. It came out bigger than I anticipated. If I split it into pages what do you think is a reasonable size to make each page composing it? RJFJR 13:01, 4 November 2007 (UTC)

It depends on the density of links. More that (very roughly) 5000 seems to cause serious trouble. The overall page size 100 to 200KB? Smaller pieces are always easier to deal with until there are simply too many of them. I split "missing" up into 7, it could've been split more (but I hope it will just get smaller ;-). If you have something like 1 template call/line, it will have to be smaller. (In case you are doing that for some reason.) "missing" has 4-6000 links per page, pages 150-200KB. Robert Ullmann 14:21, 4 November 2007 (UTC)

Boxes

Hmmm - well I thought that the box in question was OK as well, but felt that things were moving against them. I have suggested to User talk:DAVilla that consideration should be given to moving towards his vote Wiktionary:Votes/2007-07/Sister project links. Even if it is at the risk of being divisive, it is surely better to decide these things earlier rather than later? —Saltmarsh^Talk 06:58, 7 November 2007 (UTC)

WT:A

What do you reckon for me? There were a whole bunch of promotions today, so I must be in with a shout. I was once blocked here for altering policy document without approval, but that was a while ago and I was young and naive then. As for admins having no fun tasks, I'm sure that's not true. I'll sign up for it tomorrow (or when I'm next online) if you reckon I'll pass. It would be a waste of time to have the vote and fail it. --Keene 02:16, 9 November 2007 (UTC)

Someone will nominate you presently I'm sure. Connel has said he'll nominate anyone who does serious work on the Transwiki pile. (Self-nominations are allowed, but generally seem less successful, the initial endorsement by the nominator is apparently significant.) We are always looking for candidates. Robert Ullmann 03:06, 9 November 2007 (UTC)

Tbot gone astray?

Is Tbot supposed to be creating entries? Is that an impostor account or something? --Connel MacKenzie 06:37, 9 November 2007 (UTC)

Experimenting with something. Noted on the BP. Need to write some serious doc later today. (You think I was doing all that work over the past months with iwikis and {t} and tbot just to fix some links? ;-) See Category:Tbot entries Robert Ullmann 06:40, 9 November 2007 (UTC)

Tbot and interwiki

I think that when t+ template is used in translation, your bot could add one interwiki in applicable language to the newly created entry. --Derbeth ^talk 10:11, 10 November 2007 (UTC)

In general, the template is only t+ if the entry already existed and had the interwiki link ... In most cases these either didn't have the template at all, or it was just {t}. The interwikis are added by Iwicketbot and others that are checking for the existence of the FL.wikt page. Robert Ullmann 11:40, 10 November 2007 (UTC)

Do note that it is only creating entries when the template is {t}, which means the local entry doesn't exist (if it did, the template would have been changed to t+ or t- in the previous step) So there presently isn't any case where the template was t+ and the bot is creating an entry. Robert Ullmann 12:08, 10 November 2007 (UTC)

Fiji Hindi

(moved from my use page)

This is Girmitya. I am using anonymous login because my editing has been disabled by Connel MacKenzie. I have found this to be really surprising as discussions on this topic had not been finalised. I am also disappointed with, and confused about the role of Administrators within Wictionary. As far as I am concerned, all that I wanted to do was document words from a language spoked by 310,000 people in Fiji and another 140,000 people who have left Fiji for (Australia, New Zealand, United States, Canada and the United Kingdom). This language is the one one thing that unites this diaspora. At this stage, I do not care whether the Administrators prefer to cal it "Fijian Hindustani", "Fijian Hindi" or "Fiji Hindi". All that I want is a chance to contribute!!! My personal preference, and that of speakers of this language, is "Fiji Hindi", but please restore my editing rights. — This unsigned comment was added by 58.165.19.37 (talk).

I do not know what he is on about. More on your talk page. Robert Ullmann 11:47, 10 November 2007 (UTC)

DEW

Hello Robert Ullman -- Right you are, about changing DEW to an acronym. But you observed that it is always pronounced "du", not D-E-W ;-). Actually, most Canadians—at least those who can remember the DEW line—probably pronounce it "dyoo" ;-). I really do appreciate your eye for detail, though. -- WikiPedant 16:38, 10 November 2007 (UTC)

IPA templates

Someone told me to ask you about this. I'll repeat what I said at Template talk:IPA: Can't we just remove forced fonts and let users customize their User:YourUserNameHere/monobook.css? I set a style for my .IPA class, but it's being overridden by this template, making its .IPA class reference meaningless. And for {{IPAchar}} too. I use FireFox 2.0.0.9. - Gilgamesh 20:09, 14 November 2007 (UTC)

You can use !important in your Monobook to override the default font set. —Ruakh_TALK 20:29, 14 November 2007 (UTC)

Regarding User:DragonBot

Hi Robert, thanks for the note on my talk page. I do wish to run the interwiki bot and has been thinking about requesting permission. The reason why I am looking for another interwiki bot is because I need one that starts on Malayalam and other South Asian Language wikis and updates when sufficient new pages are entered. At present,

1. the update frequency is not sufficient (as I observe from mlwikt)
2. many mlwikt links are hidden since users don't add ml links to other wikis where the current two interwiki bots operate. So the bot has to start from mlwikt.

Further, DragonBot in wikipedia has over 100-thousand edits across over 40 wikis; so it is proven reliable.. Thanks very much ! --DragonBot 10:38, 15 November 2007 (UTC)

Do you understand the difference in running it on wiktionaries? (The track record on the wikis isn't that relevant ;-)

We run a bot called User:Interwicket that reads the union index of all 170 wikts (172 -tokipona -klingon ;-) and adds all of the iwikis. So it finds all the cases where a word in ml.wikt is in the en.wikt; the other bots always include the en.wikt, so they then get propagated.

That said, running it should be okay, but we have an approval process. Robert Ullmann 10:54, 15 November 2007 (UTC)

database dump

Odd, there doesn't seem to have been a datbase dump since October 16th, over a month. Have you heard anything? RJFJR 18:15, 19 November 2007 (UTC)

They tried to run an "all revisions" dump on en.wiki; it ran for a month or so and had an expected completion time in December (originally in January). That has been stopped now (they need a better procedure). It is happening right now. The usual dump we use (pages with current rev) is done. ) 2:51 UTC today. Robert Ullmann 10:03, 20 November 2007 (UTC)

Valid/Invalid L2 headers

Lately you've been updating the list about once a month (which is fine). Could you step that up for the next month-and-a-half (through early January) to update about every week or two? I've got more free time now and coming up, and it would help me to have the processed items cleared out more frequently during that time. After the start of the new year, I'll be busy again so you could drop back to monthly updates again. Thanks. --EncycloPetey 22:41, 20 November 2007 (UTC)

I can do it with each XML dump, and it is the first thing I run. The dumps have been uneven; "they" (not quite sure who that means) tried to run an all-rev-history dump on en.wiki which delayed things this time. If the dumps are available sooner, sure; in any case within 12-24 hours of each ;-) Robert Ullmann 23:08, 20 November 2007 (UTC)

Romanian gender

I was wondering if it would be possible to have a bot find every instance of m & f and f & m in Romanian entries and translation tables to switch that stuff to n. (Neuter nouns use masculine words in singular and feminine in the plural, so although it looks like they're both or something, it's a seperate gender) That would be really really awesomely awesome. — [ ric | opiaterein ] — 19:23, 21 November 2007 (UTC)

Can you give me an example? What sort of syntax? (m & f exactly?) I'll try to identify them first. Robert Ullmann 07:04, 22 November 2007 (UTC)

Because neuter nouns don't have their own set of words (generally adjectives and stuff like Latin and Slavic languages) they use masculine words in singular (un vis frumos) and feminine ones in plural (doua vise frumoase). I guess someone saw that Italian nouns can be m & f, and followed that suite or something. That better at all? :-s lol — [ ric | opiaterein ] — 14:55, 22 November 2007 (UTC)

On a similar note, I'm not sure how often it occurs, but if you could have a bot switch ã for ă in every Romanian translation (I don't think this mistake is made in any article) that'd also be really kickass. — [ ric | opiaterein ] — 19:25, 21 November 2007 (UTC)

There don't seem to be a lot. Afghan, Santa Claus, São Tomé and Príncipe (is that wrong? given the source name?), ball, burrow, capital, happy (twice), peafowl, phantom. That's all of them; easier to just fix them. Robert Ullmann 07:04, 22 November 2007 (UTC)

I didn't think there would be a lot, but I wouldn't have known how to find them. Thanks. :) — [ ric | opiaterein ] — 14:55, 22 November 2007 (UTC)

Habari Nd. Robert

Salam nyingi Nd. Robert. Kuna swali kidoogo: unajua nd. yangu mimi sijaelewa bado lengo la kuwepo kwa wikitionary, kwasababu niliangalia baadhi ya majadiliano katika sw.wkt, nikakuta mengi yanaongelewa Kiingereza, je hapa ni sehemu ya kamusi au nini? Usishangae mimi ni mmoja wa wachangiaji wa Wikipedia kwa Kiswahili, hivyo ninavyoona mambo mengine katika mradi huu lazima niangalie vipi. Tafadhali naomba nifahamishe. --Muddyb Blast Producer 13:10, 23 November 2007 (UTC)

Correct format for translations in languages with multiple dialects

Hi Robert, Eveningmist had brought to my attention to adhere to WT:ELE guidelines for the correct format for the different Kurdish dialects, but after I started using the suggested format I noticed that AutoFormat changed the format in the Translations section for the Sorani Kurdish dialect in lazy. It changed it from:

 *Kurdish:
 :Sorani:  to 
*Kurdish:
 *:Sorani: 

Is this the correct format to use? Gbeebani 01:21, 25 November 2007 (UTC)

ELE should never have specified just : in the middle of the * list; that breaks the list. For example, one doesn't want to write

# one
# two
: something
# three

1. one
2. two

something

1. three

Because it breaks the list; is less obvious with the bullets, but does mess up the HMTL list structure. So that was wrong (and Dijan has fixed it). But also, that is for multiple lines within a language (the Roman/Cyrillic in the example); for languages, it probably should be **:

• Kurdish:
  • Kurmancî: ...
  • Soranî: ...

but the real resolution is to figure out which language names and codes we want to use. "ku" seems to be used in WMF for Kurmancî w:ku:Destpêk, and maybe we should use code ckb = "Kurdish Soranî" as another language name? (I don't know if this precedence is reasonable in the case; in Norwegian, where there is Bokmål and Nynorsk, we use no = "Norwegian" and nb = "Nynorsk"). Whichever way it is worked out, we then simply use two language names without the nesting; OTOH, it is looking more likely that we need the nesting in the general case. Robert Ullmann 08:25, 25 November 2007 (UTC)

Galicia

I can fix that. Kevlar67 23:12, 26 November 2007 (UTC)

Etymology and Pronunciation sections order / DerbethBot

I remember someone asked me the same question in Beer Parlour. I just follow Wiktionary:Entry layout explained: it says that Pronunciation section should always go after Etymology - so my bot just adds pronunciation below first Etymology section. --Derbeth ^talk 10:53, 27 November 2007 (UTC)

When there is more than one Ety section, and the pronunciation is the same, WT:ELE doesn't cover the case; we put the pronunciations first. Putting it after/inside the first section is wrong. And will require cleanup. Has it already done this? Do you know? Robert Ullmann 11:13, 27 November 2007 (UTC)

I can't imagine making Wiktionary entries more complicated than you have. This is completely insane, I have to put 10 times more work than in German or Polish Wiktionary to do the same thing. --Derbeth ^talk 19:27, 27 November 2007 (UTC)

The bot has so far added audios only for Czech, Danish and Swedish; I don't expect these entries having multiple etymologies or something like this, but I can give you a plain text file with a list of modified entries if you want to run AutoFormat on them or something. I have changed the bot code and hope it will work fine now. --Derbeth ^talk 21:31, 27 November 2007 (UTC)

"hope"? Have you tested it once? While what you are doing is very good, this is not good at all. Robert Ullmann 22:23, 27 November 2007 (UTC)

I have always tested my bot after every change in the code, but no testing can prove that the bot will always work fine in all cases, as you can't predict any abnormal entry form. When I have written "I hope" I have ment "I have tested my bot and therefore I can hope". Well, but know better than me what I do.

What I find "insane" is that noone on English Wiktionary has been able to tell me how exactly my bot should work, although it was about a month from my first post on Beer parlour to the moment when my bot was given a bot flag. I have clearly written how my bot works and noone has objected; noone has ever mentioned sections like "Etymology 1". And in the middle of the work of my bot you have written on its discussion page, that I should completely change its behaviour, because it was completely wrong. When I said that I had written the bot according to Wiktionary official rules, you replied that these rules are wrong. How can I write a bot when noone on the English Wiktionary actually knows, what is the structure of entries?

Even the algorithm you have provided me is not right. The edit on theater you have found wrong is actually the only possibility: pronunciation section is in wrong place, because the rest of the entry is being transcluded as a template, so my bot cannot reach it. Even if I dealt with the case of (multiple?) page transclusion (which would require major code rewrite; I don't have enough time to do such complicated job), I think that soon an another bizzare case similar to "Etymology 1" or "page transclusion" would appear. I think that (lack of) organisation of English Wiktionary makes any automatic processing impossible as there are more exceptions than rules.

Currently I don't have time for anything; I am not able to implement complicated-as-hell handling of every possible exception on English Wiktionary. I see two solutions: either I will forget about audio files on English Wiktionary or I will modify the bot code, so that it would add pronunciation in a simple but safe way (in order not to remove anything) and would insert a cleanup template, so that people can fix them later. I think that for most languages except English such job would be just removing the template. There are more than 10 000 audio files to add, the choice is up to you. --Derbeth ^talk 19:14, 9 December 2007 (UTC)

taciturn

Hi Robert Ullmann, I wanted to both thank you and apologize for my unintentional lack of manners for your help in WT:ID#taciturn. Your reply was very helpful and interesting. I was very busy with my exams and school clubs that I may have appeared unappreciative in my short "Thanks!" reply. The reason for my short answer was that I wanted to thank everyone in time before I forget and get busy again with school. I hope I did not upset you or anyone else. Gbeebani 05:04, 28 November 2007 (UTC)

No, no, not at all. (And given the context, appropriate. ;-) Robert Ullmann 10:30, 28 November 2007 (UTC)

Code for programs

I've been running some simple perl code to find lacking Spanish entries. I was wondering if you (could) publish your code you use (especially for list of "red links" for Spanish you made). Anything would be great, I know a bit of python too. --Bequw 22:32, 28 November 2007 (UTC)

Re: Deletion of Transwiki talk:Steups

Hi Robert,

Apologies if I mucked anything up; this was my first transwiki, and I thought I followed the directions fairly well. Since I'm not a sysop on the English Wikipedia or here on Wiktionary (and thus can only use Export and not Import), I was wondering if you could tell me how best to go about copying the page history in the future.

Thanks! --Jonny-mt 09:08, 29 November 2007 (UTC)

see User talk:Jonny-mt Robert Ullmann 15:21, 29 November 2007 (UTC)

Thanks for the response--I copied and pasted per item one in the help article I linked, which suggests that course if you don't have import/export access. Based on what you're saying, though, it seems that those help guidelines are dead wrong, as they suggest both copying-and-pasting the content of the page and copying-and-pasting the talk page (and offering this page as an example of how to do this). If you could confirm that these instructions are indeed 100% wrong, I'll start a discussion on the meta side to see if we can't get them cleaned up to prevent similar problems in the future.

Thanks again, and sorry for the trouble! --Jonny-mt 04:04, 30 November 2007 (UTC)

The instructions are not wrong. They say explicitly that if/when an automated procedure is available, it should not be done manually. And explicitly states that en.wp to en.wikt must not be done manually. They should probably be edited to stress this. Robert Ullmann 11:25, 30 November 2007 (UTC)

Bug?

増大する is showing a literal {{{g}}} apparently as part of {{infl}} -- I haven't noticed this before, is this related to your changes to infl (semi-)recently? Cynewulf 19:20, 29 November 2007 (UTC)

Er, en-noun just went wacky too, maybe it's not just infl. I'll raise this in GP instead. Cynewulf 19:29, 29 November 2007 (UTC)

Never mind, somebody forgot to mount a scratch monkey. Back to monkey business as usual. Cynewulf 19:46, 29 November 2007 (UTC)

I haven't heard that term in a long time. Maybe, ah, well, since 1600 bpi x 2400' was state-of-the-art ;-) Robert Ullmann 22:17, 29 November 2007 (UTC)

Main Page vote

Re: Hi, yes I was aware that there had not yet been universal participation - is there a way to encourage more people to have a look (the vote has been in a proposed status for half a month, and it has been mentioned on the beer parlour at least twice in that time). I was assuming that there was just a general lack of interest in any more changes, but it could be that people just can't pinpoint what needs making better. Your input would be great (even though I shouldn't think of it as "mine" - I seem to - though I am happy for other people to mutilate it beyond recognition ;) Conrad.Irwin 11:52, 30 November 2007 (UTC)

People have varying amounts of time and interest, and usually serious contributors have interests that already exceed time available. One of the reasons to have a vote is to prod people into looking at it. That's why voting on using the page Dec 10-31, and then looking at where we go from there, will serve to get more response. Robert Ullmann 12:09, 30 November 2007 (UTC)

flat store

You said here that you are familiar with this term. I’ve tried, unsuccessfully, to verify it — do you want to try to do so? (As well as defining the three other partly attested senses?) † ﴾^(u):Raifʻhār ^(t):Doremítzwr﴿ 14:46, 30 November 2007 (UTC)

Conversion script redirects.

Can we get a list of all conversion script redirects with nothing linking thereto? bd2412 T 06:57, 2 December 2007 (UTC)

Well, what I'm doing now is checking Special:Whatlinkshere for each as I (code ;-) go. A lot of the links are from user, transwiki, etc namespace that are not in the XML I download. (They are in a bigger one I don't.) So doing some sort of XML analysis won't give the same result. (also there were a lot of links from old cleanup lists now blanked and/or deleted).

What are you interested in? (it may be a function of purpose) Robert Ullmann 13:52, 2 December 2007 (UTC)

If I had a list of redirects that were ready to be deleted, I figure I could stick them all in a push-the-button-to-delete template, and just roll down the page, a few hundred at a clip. bd2412 T 16:03, 2 December 2007 (UTC)

It's already automated, I'm not doing those blocks by hand ... (!). I can just leave it running and it will munch on them. The interesting ones are those that have links. Robert Ullmann 16:06, 2 December 2007 (UTC)

Heh. I've been doing them by hand. bd2412 T 16:32, 2 December 2007 (UTC)

Blocked

Hello, and thanks for the edit of μύθος. However, I have been blocked by Connel MacKenzie for my edit of Μέγα Ετυμολογικόν in the wrong place. He is right about that, but I don't think he is right about some other comments he has made for false etymologies etc. I have tried to communicate with him but it is impossible. What do you think? - Athang1504 213.16.149.78 23:17, 2 December 2007 (UTC)

Unblocked. Would be a good idea to create your user page, and explain who you are? Robert Ullmann 17:04, 3 December 2007 (UTC)

haor

Thanks. I hope nothing is harmed. Take a look at the entry now. Aditya Kabir 16:45, 3 December 2007 (UTC)

Thanks for the help. I just found out that haor also means "light" in hebrew. The word is used in the Torah, and is parts of a couple of place names in Israel. How do we integrate that information? Please, reply to my talk page. Aditya Kabir 16:42, 4 December 2007 (UTC)

christmas light

Just thinking christmas should be capitalized, at Christmas. sewnmouthsecret 18:55, 3 December 2007 (UTC)

I jumped the gun on this. Thought I had seem suggestion at info desk. Moved to Christmas lights as plurale tantum. DCDuring 19:46, 3 December 2007 (UTC)

So...

I guess that means my entry will be gone within next 5 hours then? 82.148.70.2 10:22, 4 December 2007 (UTC)

noms

How the hell do I nominate someone as an admin these days? there seem to be a whole load of templates involved... Widsith 18:12, 4 December 2007 (UTC)

I worked it out. Ignore me.. Widsith 18:25, 4 December 2007 (UTC)

rabbitlike

The adverb definition of rabbitlike is fine. Since "way" is a noun, the rabbitlike in "rabbitlike way" is an adjective, which is defined in the adjective section. But you're right that it's exactly the kind of thing Encyclopetey wants people to elaborate further on. Language Lover 21:11, 5 December 2007 (UTC)

łoś

The word łoś evaded your variations search - was it the ł? bd2412 T 03:10, 7 December 2007 (UTC)

The entry didn't exist as of the last XML dump. (20 November, you created it on the 25th) Robert Ullmann 09:43, 7 December 2007 (UTC)

Ok, I see. Thanks again! bd2412 T 14:41, 7 December 2007 (UTC)

Hey, I'm hitting a lot of redirects - can you sort those out? Cheers! bd2412 T 16:08, 8 December 2007 (UTC)

Redirects that should be entries, redirects that are bad, or (not sure if this can occur) valid redirects? Want me to separate them in each line? Should I do anything about the ae and oe ligatures, or is it good enough that you know where they end up? Robert Ullmann 04:40, 10 December 2007 (UTC)

I think most of them are "bad" in the sense that they do not constitute a valid separate sense (like caps from the conversion script, or slant quote forms redirected to straight quote forms). However, we may be missing some legit entries (especially for capitalized forms) because of the false positive of a blue link, so separating them out would be best. The ligatures are not a problem - it's easy enough for me to plug an æ or œ into a word with an a, e, or o to see if anything comes up (which they rarely do). Cheers! bd2412 T 06:07, 11 December 2007 (UTC)

Thanks, this creates a much clearer picture of how many variations there are. I'll get back to tackling the list after my move. Cheers! bd2412 T 21:50, 11 December 2007 (UTC)

main page for wiktionary day

Keep in mind that 12 December starts in the Pacific at about 1100 UTC on 11 December ;-) Robert Ullmann 05:11, 10 December 2007 (UTC)

As far as the word of the day is concerned, it starts at 0:00 UTC. Ironically, I will be somewhere over the Pacific myself at exactly that time. If you want to do the switch at a more convenient time, just update those links (the one in Mediawiki as well as any redirects) and change Wiktionary:Word of the day/December 11 to use {{WOTD}} instead of {{wotd}}. DAVilla 06:33, 10 December 2007 (UTC)

You weren't planning on moving the new design over the existing page name? Ouch! Not good. Should be a move op, with the Mediawiki link and the redirects left alone. Do you have any idea how many links there are from outside the en.wikt?

Glad I brought this up ... Robert Ullmann 06:40, 10 December 2007 (UTC)

For example, http://wiktionary.org links to http://en.wiktionary.org/wiki/Wiktionary:Main_Page changing the local link(s) is out of the question. (we won't even discuss the iwiki links on 170+ other wikts) Robert Ullmann 06:46, 10 December 2007 (UTC)

A redirect from Wiktionary:Main Page couldn't handle all of those? Anyways, I won't be doing the switch, so however you think it best. DAVilla 06:48, 10 December 2007 (UTC)

And you plan on eventually removing the redirect how? Surely you weren't thinking of leaving it that way?

For a temporary move? Of course not! I'd undo it by reverting the last change. :-P DAVilla 14:59, 10 December 2007 (UTC)

It should just be move op, with a bit of care to keep the page history attached to the relevant version. Robert Ullmann 06:52, 10 December 2007 (UTC)

keep one's eye on the ball

My experience is that is is used mostly in situations (such as cricket) in which the ball moves (or by extension, the situation changes) rather than in which the ball is stationary. SemperBlotto 09:15, 10 December 2007 (UTC)

In those cases one fairly automatically watches the ball. The idiom comes essentially from golf, where the inclination is to look where you want to send the ball, but one must—somewhat unnaturally—keep looking down at the ball, not the target. This gives the idiom its precise connotation: keep in mind where you are going, but look at what you are doing now. Robert Ullmann 09:20, 10 December 2007 (UTC)

old conversion scripts

Hi Robert, if you don't mind sharing your secrets, where are you finding all the redirection you're deleting in the impressive deletion you are doing from "old conversion scripts"? The list of redirect only caches 5000 words for me and stops in the middle of A.... Goldenrowley 04:08, 11 December 2007 (UTC)

I'm reading the XML dump of the entire en.wikt. I made lists of all of the Conversion script redirects, see Special:Prefixindex/User:Robert Ullmann/CS/ and note the trick I used to keep the lists from adding links ... also see the list of exceptions, which have links and have to be resolved. Robert Ullmann 05:34, 11 December 2007 (UTC)

Conversion script redirect links.

A handful originate solely from Appendix:Roget's thesaurus classification, so I just went ahead and made all the linked words on that page lowercase. That should take care of those. Cheers! bd2412 T 00:27, 12 December 2007 (UTC)

More than a handful, it would eventually have hit every word on the page. (it is going longest titles first, then alpha for each length, an excellent suggestion by DAVilla ;-) Thanks! Robert Ullmann 10:05, 13 December 2007 (UTC)

Actually, it would be something of a boon to my appendix project if you could eliminate the shortest titles first - there should only be a handful of two and three letter redirects anyway. Cheers! bd2412 T 02:30, 20 December 2007 (UTC)

67.53.130.69

I noticed you've had issues with User:67.53.130.69. You may wish to look at w:User:EverybodyHatesChris (block log) and w:User:67.53.130.69 (block log)on Wikipedia. The IP is currently blocked on enWiki as the source IP for banned abusive user and sockpuppeteer EverybodyHatesChris. This IP, since being blocked, has been trolling other Wiki sites leaving abusive attack messages directed at myself and other Wikipedia editors who reported him for evading his block. If you require further information, please leave a message here. Thanks. --Ckatz 08:54, 13 December 2007 (UTC)

67.53.130.69 aka Webster Boy

Hi there, I do feel that this user is being treated very unjustly at the moment, (both here and on 'pedia). The only issue I can find about him is that he created multiple sock puppets on 'pedia, which is against their policy. (Our policy is against "disruptive" sockpuppets, which this user certainly is not). I disagree with your reversion of all their edits, some of them were improvements, some of them were steps sideways, very few of them were steps backwards, the same is the case with the few items of history I have looked at on 'pedia. I have asked Ckatz for more information, and am happy to be proved wrong, but there is still no reason to undo non-detrimental edits. If you know more, feel free to point me in a better direction. Conrad.Irwin 09:37, 13 December 2007 (UTC)

I did not see any good edits, all bad; in every case the preceding version was better. (If you want to suggest a counterexample, go ahead ;-). I blocked for continuing to use given names in example sentences after being specifically told not to on his user talk page, without replying on the talk page. Further: this user's activity on WP is entirely sufficient cause for a block in itself. The one day block on the login should be infinite, and the IP blocked long term.

This user has already been reverted by 3 different sysops, and blocked twice.

We do not allow disruptive sockpuppets; we have tolerated users (one in particular) using different logins serially, but not at the same time, and not to evade blocks etc. The fact that this has already involved 4+ sysops on two different projects is sufficient to meet the "disruptive" ("bad") criterion. Robert Ullmann 10:02, 13 December 2007 (UTC)

Conrad, think about this: suppose a contributor came here and made a few edits in good faith, and CM (or someone) reverted them. The good faith user would post a note on his talk page asking politely why there was a problem, or what he might have done wrong.

This user, OTOH, immediately attacked CM, saying he was being harassed, and that he would "report it". (Talk:wise guy)

We do assume good faith; but this user has exhibited bad faith from the start. Is a troll. Robert Ullmann 10:25, 13 December 2007 (UTC)

The situation of your instant, unwarranted block and baseless reversions of all my edits, with no reasoning provided in your edit summaries, like this [9], this [10], and this [11] for example, was enough to make Google want to have nothing to do with anything Wikipedia-related. It certainly seems like that's the case. After I e-mailed Google this activity of yours from last week and added that it was the activity of a moderator at that, they issued a statement a couple days later about how they planned to completely replace Wikipedia with something more efficient. I plan to phone up The Wikimedia Foundation tomorrow and report this too them verbally and let them know that this is how disturbed I am with your wrongdoings as an admin of their site. That entire situation with you last week is a prime example of how administrators can get away with all kinds of garbage around here. I believe it was that very incident, in which I provided examples of exactly what you were doing, that made google become disgusted with Wikipedia. I couldn't care less if you try to block me or ban as a result of this letter because I refuse to edit wiktionary anymore with the uncontrolled behavior that's going on around here (perfectly proven by your actions last week).

You were a third party in a situation that had NOTHING to do with you where Connel was reverting all of my edits with no basis provided in the edit summaries. I see that you have had all these friendly chats with him in the past, evidenced by your discussion pages. This business on these sites with the taking of sides and third parties butting in where they have no business (exactly like you did) is something that I have brought to the attention of Google, to which they issued a statement of how they plan to replace wikipedia within months. I am calling up the Wikimedia Foundation tomorrow because it went too far last week and I am sick and tired of it Webster Boy 01:05, 19 December 2007 (UTC)

Oh P.S., here's the edit summary you provide before instantly blocking me afterwards [12] . Well take a look at this definition [13] . Next time give people a chance to argue their case before immediately going ahead and blocking them. I will discuss this baseless block and let them know how uncomfortable I am editing here knowing administrators can get away with such behavior Webster Boy 01:17, 19 December 2007 (UTC)

You think Google announced knol, a business-model experiment they have been working on for many months, because of your pathetic complaints? Boy, you are sure full of yourself.

I blocked you because your "contributions" were unmitigated crap.

Go ahead, they've dealt with trolls before, and will again, long after you are gone. Robert Ullmann 05:18, 19 December 2007 (UTC)

User:Athang1504 and etymologies

After talking with Versageek, I have left him a warning about the continued use of his outdated etymology source: User_talk:Athang1504#Content_and_format_of_Ancient_and_Modern_Greek_entries. If you think this is not appropriate, please let me know. I figured I should give you a heads up since you were involved in this at some point. Thanks, ArielGlenn 21:52, 16 December 2007 (UTC)

User page edits

Judging from the history, you may want to semi-protect your user page to avoid trouble since your conversion script deletions appear to attract a lot of randomness. The latest edit looks like it was trying to fly below the radar by being semi-useful. Globish 20:23, 18 December 2007 (UTC)

Would you kindly enable your email? Robert Ullmann 00:05, 19 December 2007 (UTC)

Recent comments

Thank you for retracting your comments against EncycloPetey on the vote page. Seeing someone else lose their cool reminds me just bizarre my comments (born from similar frustration) must seem. EncycloPetey has done an enormous amount of cleanup work over the years and has greatly (very greatly) helped parsing, cleanup and automation efforts.

With any new (beyond cutting edge, beyond bleeding edge) technology, everyone has questions. Everything from methodology to algorithm review that isn't explicitly described (and, lo, even sometimes then) puts questions in people's minds. I described my sentiment as "hebbie-jeebies" as I'm not sure how to accurately describe the unease about the unknown. EncycloPetey, I presume, was just looking for corner cases. The fact that you do handle it correctly is praise in and of itself, for Tbot. But disparaging one of our very best contributors (reactively) can instead make you (not Tbot) look bad.

So, I wish to request that you back off. You have encroached on my turf (that is, saying ridiculous things about good contributors.)  :-)   As I am apt to forget, the basic power of the wiki-concept, is that we are able to build off an enormous variety of talents. This very minor faux pas is a good reminder (to me, at least) how damaging an errant comment here or there, can be.

Best regards,

--Connel MacKenzie 20:29, 20 December 2007 (UTC)

fr-noun edit request

Right now, there's two things the template need todeal with most possibilities:

1. Eitehr A way to replace the entire content of the parenthesis, or a long additional series of replacements (namely un-wikied replacements for all three options and an "invariable" option)
2. An extra option for gender covering nouns wher eusage is uncertain, like après-midi.

Circeus 03:33, 21 December 2007 (UTC)

Isn't it already handling invariant (2nd parameter '-')? Or is it replacing that that is the issue? Can you give me an example? Adding something to gender is fairly easy; just need to know what it might be? Robert Ullmann 11:32, 21 December 2007 (UTC)

Re:Numbers

Hi, I would love to convert Sanskrit numbers written in Roman script into the proper Devanagari. The only problem is that I currently don't have a Sanskit dictionary or book with number vocab with me. From my knowledge of Sanskrit (having studied it for two years), I am 99% sure that if I convert these number into Devanagari, the translation would be correct. There is a very small chance that it would not however. Thanks for asking and I will see what to do! Have a nice day! DaGizza 09:27, 23 December 2007 (UTC)

userpage =

Why you have protected your usrpage?--Acdeoliopyut 18:44, 24 December 2007 (UTC)

because it has been vandalized, and there is no legitimate reason for anyone to edit it. Robert Ullmann 23:25, 24 December 2007 (UTC)

DerbethBot

I won't have time to run my bot for next few months. If you want, I can give you plain text files with a list of pronunciation files existing on Commons matched to existing English Wiktionary entries. If someone makes a bot for inserting audios, they can use these data, or they can just be simply used for manual insertion (for example, there are about 1100 entries in English on the list so it might be as well performed by hand). The data is grouped by language. --Derbeth ^talk 10:43, 20 January 2008 (UTC)