User talk:Robert Ullmann/2010–2011

archive pages, page history with archives

• User talk:Robert Ullmann/2006
• User talk:Robert Ullmann/2007
• User talk:Robert Ullmann/2007b
• User talk:Robert Ullmann/2008
• User talk:Robert Ullmann/2008b
• User talk:Robert Ullmann/2009

m:Wiktionary/logo/refresh/voting

I do not want to come across as contumelious but please consider casting your vote for the tile logo as—besides using English—the book logo has a clear directionality of horizontal left-to-right, starkly contrasting with Arabic and Chinese, two of the six official UN languages. As such, the tile logo is the only translingual choice left and it was also elected in m:Wiktionary/logo/archive-vote-4. Warmest Regards, :)--_thecurran Speak your mind ^{my past} 03:15, 2 January 2010 (UTC)

{{t-simple}} and water

water is now using {{t-simple}} for around a thousand of its translations (written in Latn, so no need for other parameters) the existing templates were overstretching the server leading to internal errors on save (presumably out of memory/time) and horrificly slow loading (blamed on {{Xyzy}} by Prince Kassad). Using a template was preferred to using raw links for consistency within the page. At the moment we have no other pages that will benefit from this template, and I can't imagine any arising in the near future. Conrad.Irwin 16:26, 3 January 2010 (UTC)

Cite references prefix

I was wondering if we could remove the "* Notes:" from MediaWiki:Cite references prefix. Sometimes we do use it to show footnotes in a random section (eg 和尚打伞，无法无天) and in those cases the text is okay. But often times we use it show references where that snippet of text looks awkward right after the ===References=== or we show footnotes in a ===Usage notes=== section making the text look even more awkward. I know you talked about this with Ruakh before, but I don't understand the reason. Wikipedia doesn't have an intro text (see w:MediaWiki:Cite references prefix) which makes me think any technical problem must be small. Obviously if we change this we'll have to do some cleanup, but this is probably better than awkward wording on the majority of the pages. --Bequw → ¢ • τ 23:54, 25 January 2010 (UTC)

Any more thoughts? --Bequw → τ 17:20, 18 February 2010 (UTC)

Old Tbot categories

Hi, I notice that at least half of the Tbot entry categories per language are empty. I would have deleted them, but if Tbot plans to restart adding words from these languages, it would make no sense to! Mglovesfun (talk) 11:42, 5 February 2010 (UTC)

Tbot will add more entries for languages as people add more translations and entries in the FL wikts. So the cats should stay. (Harmless anyway ;-) The only cats to be deleted are the monthly ones once emptied, as they won't recur. (sorry for delay, i've been sick and some time in hospital) Robert Ullmann 08:55, 6 February 2010 (UTC)

User:Rising Sun/French verbs needing conjugation

Hello Robert, any chance of updating User:Rising Sun/French verbs needing conjugation? p.s. I hope you're feeling a lot better now --Rising Sun talk? 15:52, 6 February 2010 (UTC)

okay, that's just a matter of re-running the program. Done. Robert Ullmann 17:27, 6 February 2010 (UTC)

Your template

Hi Robert could you please remove Template:xhan from the Category:Chinese_words_needing_attention? Thank you. Tooironic 10:38, 22 February 2010 (UTC)

Hmm, it's unused and unusable in its current form, as you can't substitute it. Mglovesfun (talk) 10:49, 22 February 2010 (UTC)

Hi, my (repeated) apologies for illness with some long time to get back to effectiveness ...

It was used by me to subst to create/fix various entries. Was always supposed to be temporary. Fixed. Robert Ullmann 22:30, 22 February 2010 (UTC)

Portuguese verbs needing conjugation

Hi Robert again. Can I request a page User:Rising Sun/Portuguese verbs needing conjugation, in the same vein as User:Rising Sun/French verbs needing conjugation. My bot's working on Portuguese now, you see--Rising Sun talk? contributions 18:28, 27 March 2010 (UTC)

Yes, should not be difficult. About the same rules? I am still catching up on a number of things left undone from before I was in the hospital. (BD2412 has probably given up on me ;-) I did note AF sorting out some pt entries in common with Spanish. Robert Ullmann 17:03, 28 March 2010 (UTC)

The rules for this are: find all Portuguese verbs entries without any pt-verb in them - we use pt-verb for conjugation templates, whereas in French we use fr-conj. And you needn't worry about finding any Tbot entries - they don't concern me. I hope this helps. --Rising Sun talk? contributions 17:16, 28 March 2010 (UTC)

pt has pt-verb (inflection line) and pt-conj (table) just like the French (arranged a bit differently); that isn't a difference. Robert Ullmann 18:44, 28 March 2010 (UTC)

Done. Robert Ullmann 19:07, 28 March 2010 (UTC)

Thanks, I love you Robert --Rising Sun talk? contributions 19:12, 28 March 2010 (UTC)

• Hi again, can I please get a rerun of User:Rising Sun/Portuguese verbs needing conjugation. I've attacked many of these already. --Rising Sun talk? contributions 20:42, 14 April 2010 (UTC)

  Done. Robert Ullmann 22:30, 14 April 2010 (UTC)

• Can I get one done for Dutch verbs too, for CodeCat and James - User:Rising Sun/Dutch verbs needing conjugation. --Rising Sun talk? contributions 09:30, 16 April 2010 (UTC)

  Done. There is nothing to be fixed: the list is all words that are both verb and verb form, and a few verbs not conjugated. Robert Ullmann 11:38, 16 April 2010 (UTC)

• Maybe also User:Rising Sun/German verbs needing conjugation, User:Rising Sun/Spanish verbs needing conjugation, User:Rising Sun/Latin verbs needing conjugation and User:Rising Sun/Italian verbs needing conjugation, for the bot owners to get their teeth into --Rising Sun talk? contributions 12:38, 16 April 2010 (UTC)

  I'd need to generalize the program. I'll think on it. Robert Ullmann 13:05, 16 April 2010 (UTC)

• I must bug you again for this. Thanks --Soleil levant 00:54, 29 April 2010 (UTC)

Pronunciation exceptions and mismatched wikisyntax

At your convenience, please could you regenerate these reports as I might finally be getting some time to do a bit more on them, but given the length of time it's been they're now very out of date.

See also the talk page of the pronunciation exceptions report. Thryduulf 00:39, 14 April 2010 (UTC)

I've run the pron except report, and mismatch running now. This is just the existing code; I spent several hours tracking down a completely gratuitous breaking change in the server code and working around it. I'll think about the talk page suggestion. Cheers, Robert Ullmann 13:53, 14 April 2010 (UTC)

Cheers. Thryduulf 22:20, 14 April 2010 (UTC)

CaBot and Interwikis

Hi Robert,

I'm on the french wiktionary, and I'd like to keep on interwikis too. I suppose your bot is based on english recent changes, isn't it ? So if mine will follow french ones it will work without disturbing yours, isn't it ?

Cordially, -- Quentinv57 ✍ 16:38, 16 April 2010 (UTC)

Interwicket follows RC on all the wikts. It also knows the current rules for link order and policy on redirects on all wikts in namespace zero, which the pywikibot does not have up to date.

If you want to work outside NS:0, cool, but do see if you can get the proper ordering. In NS:0, is handled by Interwicket. If you want to help, is okay, but there are lots of issues. Robert Ullmann 20:26, 16 April 2010 (UTC)

Luckas-bot is working on NS:0 too, isn't it ? -- Quentinv57 ✍ 20:28, 16 April 2010 (UTC)

User talk:AutoFormat

Hi Robert,

When you have time, there's a bit of a backlog of comments and requests at User talk:AutoFormat. You're in high demand. :-)

Thanks in advance!
—Ruakh_TALK 20:35, 16 April 2010 (UTC)

Oh dear. I'll hunt these down. Odd though: when Interwicket talk gets changes I get mail (and that shows on my cell phone even in the middle of the night, at o-dark-whatever. But not AF? maybe has to do with when SUL was started?) Anyway, I'll chase these in the morning! Thank you. If anything needs to be brought to my attention, you or anyone feel free to SMS +254 722 929 463. (and: don't worry about time, it is set silent when I am sleeping ...) Robert Ullmann 21:06, 16 April 2010 (UTC)

There is an "email me if someone writes on my talk page" setting in preferences (not WT:PREF, but the standard Wikimedia one, next to one's watchlist). Perhaps that was set for Interwicket, but not for AF. -Atelaes λάλει ἐμοί 22:25, 16 April 2010 (UTC)

It was set, but the email address was an old one no longer used. Fixed. Robert Ullmann 11:19, 20 April 2010 (UTC)

and it does work now Robert Ullmann 11:59, 21 April 2010 (UTC)

Welcome back?

Sorry about your health issues. I hope they have been resolved acceptably.

I hope you will have enough time to participate again. If you have been working on tech matters and have been on IRC, I wouldn't see that. Anyhoo, welcome back. If you aren't staying, come back soon. DCDuring TALK 11:35, 28 April 2010 (UTC)

I remember learning while studying public health how long it takes people to recover from being in hospital, but hadn't experienced it. Two days in ICU, plus another 5 in a private room, and one leaves feeling like everything is pretty much fine. Then I discovered that getting anywhere near fully functional takes months. Wow. (No wonder modern medicine tries to do almost everything they can on out-patient or short-stay (24-36 hours), recovery time is much better.) Will be able to do more now. Robert Ullmann 12:07, 28 April 2010 (UTC)

Don't get me started. That is yet another reason I try to avoid or postpone encounters with doctors and hospitals and resent being compelled to pay the expenses of those who overuse medical services. In my case, total knee replacement is likely to become necessary, but as much delay as one can stand seems to be the best course, both for the expense (if one pays one's own way) and the recovery. Overuse is sometimes even against the patient's wishes (excessive end-of-life treatment), to the patient's detriment (medications without demonstrated benefit for average recipient), or to the detriment of public health (overuse of antibiotics).

But when medical care is needed, it is really needed.

I hope wiktionary remains high on your list of priorities. DCDuring TALK 14:24, 28 April 2010 (UTC)

Google mining

Robert, I'm so sorry to hear that you're recovering from a hospital stay. Hope things pick up for you. I have a request, maybe for too much, but we'll see. Would you be able to generate a list of all words (including hyphenated forms and contractions) between, say, five and fifteen letters, which return fifteen or more Google Books hits? I've done a little scratch testing on this, and I think that those parameters will yield a usable list primarily composed of actual words-in-use that meet our CFI. That would, in turn, ease the process of verifying words, and would also provide an enormous bank of what is missing from our dictionary. Be well, and cheers! bd2412 T 14:41, 28 April 2010 (UTC)

Are you serious? If you were to generate such a list for every language and every word between 5-15 letters in every language that get >15 google hits, you would be looking at lists of millions, if not tens of millions of words...Razorflame 16:24, 28 April 2010 (UTC)

That is the point. All words in all languages. bd2412 T 16:53, 28 April 2010 (UTC)

Just to ease up on the request a little bit, we could limit it to words written in the Latin alphabet (with Asian languages, it's hard to tell where words begin and end anyway). Let's suppose we generate a list of 2.5 million words that way - that would be 100 pages with 25,000 words each. In terms of storage space it would be fragmentary, and we can very quickly weed out the ones we have, and easy-to-make missing derivations. We could also start with a test run, for example, only five-letter words. bd2412 T 19:31, 28 April 2010 (UTC)

Wow. (Yes, I've used that word twice in 24 hours. Sorry.) This is a really good idea. It will require formal cooperation with Google (one can't just run automation on their search, let alone "Books") but that may be an advantage: they could easily gen the all-words list, and I could screen it. And it is the sort of thing they'd like. (We can add lots of Google citations too.) Must think on this. (is 2 AM here, watched the UEFA match) I did do some things with hunting words in current newspapers, but never fully integrated it. hmmm .... :-) Robert Ullmann 22:56, 28 April 2010 (UTC)

I asked the Google Books help desk if they could provide such a list, and their reply was "we're sorry we can't help you". Perhaps they were leery of running the search themselves, and would welcome someone else designing the search and just slipping it into their database. Do we really need their cooperation anyway? Given their volume, how would they even know they're being auto-searched (and why would they care)? bd2412 T 02:27, 29 April 2010 (UTC)

And how would we feel if someone else took that attitude? Eh? (;-) They (meaning the s/w) notice very quickly. On several occasions I have (manually) done a bunch of searches fairly rapidly, and their servers started giving me CAPTCHAs. Any kind of automation beyond long random intervals for parts of a day will probably result in my IP address getting blocked. And: we need a few million results, right? Using individual queries would be very inefficient.

Like any other organization with a "Help Desk", that is the last place you should go for help. We need to talk to someone fairly senior, or some interested engineer in Google Labs or similar. Maybe the "Google Summer of Code" people might help? thinking on it Robert Ullmann 14:01, 29 April 2010 (UTC)

I see your point. Let me know if there's anything I can do to help! bd2412 T 18:40, 29 April 2010 (UTC)

Get well soon, RU. [1] help any (with the Google issue, not with getting well)?​—msh210℠ 18:59, 29 April 2010 (UTC)

{{language}}

Given that there are now no pages at Special:PrefixIndex/Template:lang: it seems that much of this template's complexity is now un-needed. I thought I would double check with you before changing it to {{{{{1}}}|l=}} (alternatively or additionally, we could begin to deprecate its use completely as it doesn't represent a saving in effort). Conrad.Irwin 22:29, 30 April 2010 (UTC)

Yes, we can take out the "lang:" code. I want to keep the template because part of its purpose is to hide the method used. If we ever get a extension or function like #language that would return the names (but in English ;-) we can change the mechanism. If the "l=" hack is all over everywhere this will be a lot harder! (Just identifying the code isn't so easy, is it a language reference or some other template using "l" as a parameter?)

Do note that now calling {language} with blank returns blank, that is the function of {lang:}. If you over-simplify, this breaks. We want something like: {{#if:{{{1|}}}|{{{{{1}}}|l=}}}}

The overhead of having the template is low, it is often used dozens of times in an entry, but only loaded once (one db SELECT) and is straight line code, the only branch being nil. Robert Ullmann 14:54, 1 May 2010 (UTC)

I've kept it (as now) with an absent first parameter causing an obvious error, but a blank first parameter returning blank. Conrad.Irwin 15:26, 1 May 2010 (UTC)

chizi

Umm... I think this tbot Swahili entry speaks for itself. Please help. --EncycloPetey 23:39, 30 April 2010 (UTC)

Wow, that's freakin' hilarious. I wouldn't want to run into that chizi in a dark alley. -Atelaes λάλει ἐμοί 23:41, 30 April 2010 (UTC)

you clearly have not met me in person. CHEESE!!

a very good entry

I am perfectly serious. Really!

(and yes, the pix is for the other definition of the word, which is why review is needed; but I really am mad about CHEESE which Kenyan dairies are just learning to make; I am about to raid my refrigerator right now ...) Robert Ullmann 00:03, 1 May 2010 (UTC)

Thanks, Wallace. --EncycloPetey 02:47, 1 May 2010 (UTC)

Tbot

Hi, Rob. Do you have your Tbot code available somewhere? I'm interested in arranging something similar for pt.wikt and your code would be a good place to start. Thanks, Malafaya 11:30, 3 May 2010 (UTC)

User:Tbot/code not completely up to date, but good enough for stealing. (And it isn't really good code, mostly kludged together.) Robert Ullmann 11:35, 3 May 2010 (UTC)

Great! Thanks a bunch, Malafaya 11:38, 3 May 2010 (UTC)

---

Hello, Rob. I'm afraid I don't understand what you meant by "We don't use a -5 level. Either -4 or native ;-) " on my talk page. Anthonybaldwin 12:53, 6 May 2010 (UTC)

Automated templating in Rhymes namespace

Would it be possible to go through the Rhymes namespace and automatically template any of the following patterns:

-foo /-foo/ /-foo/ → {{enPR|-foo}}, {{IPA|/-foo/}}, {{SAMPA|/-foo/}}
-foo -foo -foo → {{enPR|-foo}}, {{IPA|/-foo/}}, {{SAMPA|/-foo/}}
-foo /-foo/ <tt>/-foo/</tt> → {{enPR|-foo}}, {{IPA|/-foo/}}, {{SAMPA|/-foo/}}

/-foo/ /-foo/ → {{IPA|/-foo/}}, {{SAMPA|/-foo/}}
-foo /-foo/ → {{enPR|foo}}, {{IPA|/-foo/}}

/-foo/ → {{IPA|/-foo/}}
-foo → {{IPA|/-foo/}}

Adding the lang= parameter based on the page title (e.g. Rhymes:French:foo should get lang=fr)

If you could then produce a report showing the pronunciation lines of all the rhymes namespace pages with no pronunciation template, or a pronunciation line with templatesd and untamplated text, in a similar fashion to the Pronuncation Exceptions report so they can be fixed manually or other automatable patterns discovered. Thryduulf 22:03, 7 May 2010 (UTC)

On second thoughts maybe you could do the second part by simply extending the existing pronunciations sections report to work on Rhymes namespaces as well. Thryduulf 22:04, 7 May 2010 (UTC)

Wiktionary:About Serbo-Croatian

Jesus, could you try to be a little more inflammatory? I agree that a nod to the vote is worthwhile, but let's work out some different wording, shall we? Something a little more along the lines of "the community has not reached a consensus on this topic."? -Atelaes λάλει ἐμοί 01:05, 13 May 2010 (UTC)

Ok, I've taken a stab at it. Let me know what you think. -Atelaes λάλει ἐμοί 01:11, 13 May 2010 (UTC)

Fair enough on removing the official looking banner, but you can't make a definitive statement that editors must or must not conform to the proposal. We have no official policy stating that Bosnian, etc. must be kept either. Since there is a lack of policy either way, interested editors are free to do as they see fit. I'm sorry that nearly all of the interested editors disagree with you, but such is the reality. -Atelaes λάλει ἐμοί 01:18, 13 May 2010 (UTC)

It does need to say that the other sections shouldn't be deleted; people are doing a lot of damage that was in no way agreed on. No, a few trolls disagree with me (and Stambuk is not one of them; he has added other sections and stated repeatedly that others may!). (The proposal is nonsense, which is why every single authority has rejected "SC". Eh?) Robert Ullmann 01:23, 13 May 2010 (UTC)

I've told you this so many times, that I really tire of doing it again. I (and many others) don't see the sections as being deleted, but as being merged. If someone created an entry with "Classical Greek", "Koine Greek", and "Byzantine Greek" sections, I'd merge them all into "Ancient Greek" and you'd pat me on the back for doing it. Ivan has cited many authorities who embrace SC. -Atelaes λάλει ἐμοί 01:27, 13 May 2010 (UTC)

Are you aware of how many "authorities" he has cited have recanted, now that they aren't under threat of being imprisoned or shot for not supporting SC? And your comparison is completely invalid, as Croatian and Serbian etc are different languages, with large active populations that do not consider them the same language, and find proposals to merge them in large part extremely offensive. If I forcibly merged Bokmal and Nynorsk would you "pat me on the back?" Eh? (;-) best, Robert Ullmann 01:35, 13 May 2010 (UTC)

No, I guess I am not. It's funny that you mention the whole Norwegian thing. I find it a bit humorous that you spent that entire debate with such obvious disdain for Nynorsk, trying to relegate it to some second class language, and then make accusations of genocide against a Serb and a Croat who are trying to treat their "languages" as a unified whole. I've seen a number of linguists say that all the Scandinavian languages could easily be considered a single language, owing to their extreme similarity. So, if you spent the time to create a robust proposal for merging the two Norwegians, and got feedback from people, then yes, I might just pat you on the back. Concerning the whole Chinese thing, you've stated many times that SIL is more than happy to bow to political pressure in their coding, and that we shouldn't follow the whims of some political entity with a dogma to promote. But, the specific SC debate aside, on Wiktionary, when there is no policy, editors are free to do as they wish. I wish we could've resolved the whole Norwegian thing back when it happened, but we didn't. So we had to leave the Norwegian editors alone to let a policy evolve through edits. Same applies here, though it would happen a bit more peacefully if we didn't have all these fanatics showing up with a bone to pick. -Atelaes λάλει ἐμοί 01:47, 13 May 2010 (UTC)

Herr Ullmann, trolling as usual. If I see you blocking "trolls" such as Kilibarda, you are so dead. --Ivan Štambuk 07:02, 13 May 2010 (UTC)