Wiktionary:Grease pit/2011/August

Template:tø
I came up with the idea that templates that don't have two unnamed parameters should be in Category:Translation templates needing attention. It's turned up quite a few results already - ten so far. How would people feel about t, t+ and t- having the same function. Any downsides? Mglovesfun (talk) 22:35, 1 August 2011 (UTC)
 * No, none? --Mglovesfun (talk) 11:07, 3 August 2011 (UTC)
 * It's a good idea, but I think a missing second argument should probably just use the normal, rather than Category:Translation templates needing attention. Actually, thinking about it — maybe there should be a "default" category that uses when no language code is specified. If that were the case, then I think a missing first argument to  should probably also just use the normal . —Ruakh TALK 11:15, 3 August 2011 (UTC)
 * Partly related, the five remaining entries in Category:Translation templates needing attention use a language code and a transliteration with no original script, because the original script isn't known. These are for ancient languages like Sumerian and Hittite. --Mglovesfun (talk) 11:23, 3 August 2011 (UTC)
 * I quite like the idea of a cleanup category just for translations templates, but I wouldn't object if other users preferred using an established category (that is, one we already have and use). --Mglovesfun (talk) 16:03, 3 August 2011 (UTC)
 * Can I go ahead with this? Even doing it temporarily (for a few days) would allow users to clean up quite a few translation templates. --Mglovesfun (talk) 10:56, 9 August 2011 (UTC)

Definitions
There are several (8K) sorted definitions from Wiktionary here. This file may be interesting for somebody, I hope. -- Andrew Krizhanovsky 08:42, 2 August 2011 (UTC)
 * Y'know you can get all the definitions from http://toolserver.org/~enwikt/definitions/. Conrad.Irwin 08:14, 5 August 2011 (UTC)
 * Thank you. It is interesting. -- Andrew Krizhanovsky 14:54, 15 August 2011 (UTC)

boring job to be done by bot or bored user
Can anyone whip up a script to do these sorts of edits, allowing quick navigation between the pages? --Reetmo 13:17, 6 August 2011 (UTC)

Empty positional parameters and default values -- bug?
I'm working on a template that uses named parameters in combination with positional parameters, with default values for some of them. However, test calls are not producing the expected results.

Sample call:

Value 3 defaults to, and value 4 defaults to. These are defined as:

I'd expect to see:

Instead, I get:

What am I missing? When passing the third parameter as an empty value,  should evaluate to   since   is undefined, and this should then evaluate to   since   is empty. But then where is this single quote ' coming from?

Reading around on Wiktionary, Wikipedia, Wikimedia, and Mediawiki (bewildering that the docs are spread around so much, but that's a separate matter for discussion), it sounds like a design mistake early on has caused confusion in terms of how named versus unnamed parameters are handled, and I'm starting to think my bugaboo here is part of that. I'd deeply appreciate it if someone could point me to an elegant way of clearing this up -- I'd hate to have to add  calls all over the place, but maybe that's what it comes to? -- TIA, Eiríkr Útlendi | Tala við mig 18:45, 8 August 2011 (UTC)


 * If  is defined as the null string then   evaluates to the null string; only if   is undefined (, for example) does   get used instead. I have no idea where the apostrophe is coming from, unless you have quotes around the code (like   and they're coming together (as  , which yields a boldface apostrophe) because now only the null string (i.e., nothing) is between them. If you want to use   when   is the null string, then use   or the like. &#x200b;—msh210℠ (talk) 19:54, 8 August 2011 (UTC)


 * I was afraid that would be the case (needing ).  And you're right about the single quote, I forgot to add in the italics in my examples.  Thanks for chiming in, Msh210!  -- Eiríkr Útlendi | Tala við mig 20:28, 8 August 2011 (UTC)


 * I agree, there's no way to do this without using the #if template. Otherwise it will just accept the null string. BigDom 20:42, 8 August 2011 (UTC)


 * Ta, thanks BigDom -- Eiríkr Útlendi | Tala við mig 22:50, 8 August 2011 (UTC)

isValidPageName
Curiously returns (that is, a blank string) although - is valid, as you can see by the blue link. Why is this? --Mglovesfun (talk) 10:57, 12 August 2011 (UTC)


 * If by "why is this?" you mean, "what causes this?", the answer is that that behavior is hardcoded into the template: if its argument is -, it returns a blank string, without doing its normal check. If by "why is this?" you mean "why was this done?", I have no idea. I can only imagine that it was intended that to be callable as  , but I don't know why. Something like    would work just as well. —Ruakh TALK 12:31, 12 August 2011 (UTC)
 * I never realized it was intentionally coded like this, I just assumed it checked for validity! As for why, no I don't know either. --Mglovesfun (talk) 12:59, 12 August 2011 (UTC)

request for transliteration
What template should I use to mark that a Persian word in the "Translations" section is lacking transliteration? I use but then entry goes to the Category:Entries which need Arabic script, not to the Category:Persian terms lacking transliteration. Is there a template for this? Maro 21:48, 13 August 2011 (UTC)
 * Ok, there was no such template. I created . Maro 19:34, 28 August 2011 (UTC)

Duplicate (but different) moves
At 17:17 / 17:18 this after noon two people moved the same page to different targets. Both moves worked, no redirect was left. Is this a bug? SemperBlotto 17:20, 14 August 2011 (UTC)
 * Grr. The first move left a redirect. The second just moved the redirect! SemperBlotto 17:28, 14 August 2011 (UTC)

Issue with not using the script specified in the sc= parameter
Take a look at he and shi. Even with the sc= parameter specified, the template is still picking up the script associated with the lang= parameter, making the end result a tad difficult to read for some (and god awful to look at). seems to exhibit the same issue. I had a look at both and, but couldn't find anything wrong. I thought the sc= param was supposed to take precedence over the the script specified by the lang= tag? Is this not the case anymore? Jamesjiao → T ◊ C 00:08, 15 August 2011 (UTC)
 * The template wasn't accepting sc. Now fixed. --Yair rand 00:37, 15 August 2011 (UTC)
 * Thanks Yair rand! Jamesjiao → T ◊ C 02:34, 15 August 2011 (UTC)

Issues with search AND secure server
Bug description: What do we do about this? The set of steps seems to work on e.g. en.wikipedia and sv.wiktionary, so I conclude the issue is local for English wiktionary. \Mike 04:58, 15 August 2011 (UTC)
 * 1) Log in to the secure server.
 * 2) On any page, do a search for a non-existent page title.
 * 3) On the search page, use the special search box unique to this page, and search for anything. Hit the search button on the page (not in the margin).
 * 4) Find yourself taken to a "Wiki does not exist" page.
 * Did you file a bug report at ? --Yair rand 05:02, 15 August 2011 (UTC)
 * Actually, this seems to be an issue with MediaWiki:SpecialSearch.js. --Yair rand 05:06, 15 August 2011 (UTC)
 * Fixed. --Yair rand 05:15, 15 August 2011 (UTC)
 * Thank you. \Mike 15:57, 15 August 2011 (UTC)

Deletion perms within one's own User area?
I've been working on some template ideas, and have rethought how to organize things a couple times. Moving works well enough, but then I'm left with all these stub pages with unneeded redirects. Is there any way of granting folks deletion permissions just within their own User areas? -- Curious, Eiríkr Útlendi | Tala við mig 09:13, 17 August 2011 (UTC)


 * No, but you can ask an admin to delete such pages for you. —Ruakh <i >TALK</i > 12:30, 17 August 2011 (UTC)
 * ...by adding to each page, unless there are too many to bother doing that to, in which case identify them (e.g. here). &#x200b;—msh210℠ (talk) 14:53, 22 August 2011 (UTC)

Thank you both. I'll add the once I get enough deletable pages to where it makes sense to get rid of them all in one batch. -- Eiríkr Útlendi | <small style="position: relative; top: -3px;">Tala við mig 16:43, 22 August 2011 (UTC)


 * Oh, don't worry about that. If anything, it might be better to add to just an entry or two at a time, rather than flooding the category all at once. —Ruakh <i >TALK</i > 23:30, 22 August 2011 (UTC)


 * Sounds good, I may have time for that tomorrow. (There's not much that needs tagging, it's just that I saw this right as I'm about to get back to work. :) -- Eiríkr Útlendi | <small style="position: relative; top: -3px;">Tala við mig 04:43, 23 August 2011 (UTC)

Removing prefixes from language templates
There was an earlier discussion at Grease_pit_archive/2010/December on removing the prefixes "proto:" and "conl:" from language template titles. The only objection given was that moving the templates would complicate AutoFormat's updating of User:AutoFormat/Languages. Since AutoFormat is no longer running, is there any reason not to remove the prefixes? --Yair rand 01:13, 23 August 2011 (UTC)


 * The current system is a disgusting mess of ad-hoc code. The real problem is not the prefixes themselves — <tt>proto:</tt> and <tt>conl:</tt> are a good way to mark these non-mainspace language codes — but the fact that a large number of existing templates have been retrofitted to support both prefixed and unprefixed language codes by using . But since we're never going to fix that problem (one editor in particular is strongly dedicated to the proposition that the difference between mainspace languages and non-mainspace languages should be completely editor-transparent), removing the prefixes is probably the next-best option. —Ruakh <i >TALK</i > 12:43, 23 August 2011 (UTC)


 * Transparent can mean either that the thing is easy to see (figurative sense) or that it's impossible to see (more literal sense). Often the intended meaning is clear; it's less so in the above. &#x200b;—msh210℠ (talk) 18:11, 25 August 2011 (UTC)


 * I don't mean either of those. I was pretty sure that my sense was common in programmer circles, but maybe not? Anyway, I mean that the editor in question feels that something like should just work; to the extent that we treat Klingon differently from other languages, the editor feels that templates should handle that internally, rather than editors having to be aware of the distinction. (That is, the distinction would be "transparent" to editors, in that they could see right through it: both kinds of languages would be handled the same, from their perspective.) Anyway, thanks for making be clearer. :-)   —Ruakh <i >TALK</i > 19:12, 25 August 2011 (UTC)
 * Ah, so "impossible to see", then. Not in them, I don't know what's common in programmer circles. &#x200b;—msh210℠ (talk) 19:15, 25 August 2011 (UTC)

Right alignment of images, wikipedia box
What method was used to right-align images and the wikipedia box even in long articles, so they don't interfere with the left side and they don't create blank gaps on the left? It would be great to use the same in the Hungarian Wiktionary. Thanks. --Panda10 21:34, 23 August 2011 (UTC)
 * Without checking CSS that governs, I'll guess its . &#x200b;—msh210℠ (talk) 16:28, 25 August 2011 (UTC)
 * Actually, my question is more about the relationship of the left and right sides. What did we do here that the right side is not interfering with the left side? Before, whenever we added multiple images on the right, the trans-top table remained as wide as the page and there were big white gaps. Now it shrinks to allow the right side co-exist with the left and there are no gaps. --Panda10 17:27, 25 August 2011 (UTC)
 * Big white gaps before something as wide as the page are usually because the lower-appearing thing has  set (or has something appearing right before it (like ) that has either   or  ). &#x200b;—msh210℠ (talk) 18:08, 25 August 2011 (UTC)

Adding translations
I just tried to add the Luxembourgish 'hoppelen' as a translation of the verb 'hop' using the Add translation tool. It usually works fine, but this time I got the following error message: "Could not find translation entry for 'lb:hoppelen'. Please reformat." Obviously I can still add it manually, but does anyone know why this happened? Cheers, BigDom 11:02, 25 August 2011 (UTC)


 * I had these issues as well. There seems to be a conflict with (translation request) for other languages and  (translation to checked) even for the language you're actually checking. --Anatoli 02:49, 26 August 2011 (UTC)

WT:RFV slow, reluctant to load
Has anyone else had trouble making changes to WT:RFV — encountering Wikimedia Errors ("Error: ERR_READ_TIMEOUT, errno [No Error]") when trying to add or archive sections? It's almost as if archiving those 300 000 bytes made the page slower... I do not encounter the error when adding to other pages, even pages of comparable size, like this one or WT:RFC. - -sche (discuss) 02:36, 26 August 2011 (UTC)


 * I ran into that a lot earlier today with WT:RFD and RFC; I figured maybe there was a lot of traffic on the server, or a lot of editors at work at once, or maybe even a DDOS or something. -- Eiríkr Útlendi | <small style="position: relative; top: -3px;">Tala við mig 06:13, 26 August 2011 (UTC)


 * Yes. When I update it I get a Wimimedia Error message or similar and I try again - and again... All the edits were successful - so we end up with multiple additions that have to be undone. It's getting to be a pain. SemperBlotto 21:28, 26 August 2011 (UTC)
 * The trick, I find, is to look at the recent changes, or Special:MyContributions in a different window. If the message has been posted, simply close the first window. Mglovesfun (talk) 10:26, 27 August 2011 (UTC)


 * FWIW, I tried making a couple edits to de:WT:WL (the biggest page I know of on de.Wikt) and had no trouble, though it's possible that page wasn't big enough (only 120 000 bytes). Is this a local en.Wikt issue, or does it appear to be a Wikimedia-wide server issue? - -sche (discuss) 11:14, 29 August 2011 (UTC)


 * Feedback is giving the same problem now (but not all the time). SemperBlotto 14:17, 29 August 2011 (UTC)

Ranunculus acraeus
This is coming up in Special:UncategorizedPages. For some reason, the server (not because of anything the creator has done) is interpreting all the templates with an uppercase first letter. The entry is correct, as it uses, the server however is interpreting this as , which doesn't exist, hence uncategorized. Mglovesfun (talk) 10:04, 27 August 2011 (UTC)
 * And now fixed; and I have no idea why. Mglovesfun (talk) 21:22, 27 August 2011 (UTC)

Correctly sorting categories
I'm trying to get Category:Old Irish noun forms to sort properly by treating "á é í ó ú" as equivalent to "a e i o u". Can anyone tell me why the <tt>|sort=</tt> parameter of works at [[láime]] but not at [[féile]], [[díu]], or [[óil]]? —Angr 17:28, 27 August 2011 (UTC)
 * Apparently, that's because and  try to categorize and sort the entries again. --Daniel 17:43, 27 August 2011 (UTC)
 * But [[láime]] is using and it's sorting correctly. —Angr 17:55, 27 August 2011 (UTC)
 * I'm with Angr on this; the syntax currently used in óil really ought to work. Perhaps it's the server then, when it refreshes it will be ok. Mglovesfun (talk) 18:05, 27 August 2011 (UTC)
 * No, not actually. [[láime]] does not have a <tt>|lang=sga</tt>, which makes a difference, for some reason. --Daniel 18:07, 27 August 2011 (UTC)
 * Indeed it does. I took the <tt>|lang=sga</tt> out of the other three, and now they sort correctly too. —Angr 18:17, 27 August 2011 (UTC)
 * It's because categories in

, and it's trumping the sort key in. Mglovesfun (talk) 21:22, 27 August 2011 (UTC)
 * Daniel C said this above, my bad. Mglovesfun (talk) 21:23, 27 August 2011 (UTC)

AWB being an ass
I'm trying to get all the transclusion of, including the templates that use it as a base (see Category:Swedish declension-table templates. I'm using AWB's 'skip' tab, selecting Contains: (that is, skip if the page contains) which is what I used to find out which pages in Category:French verbs weren't using . I then ticked regex and put in . All I want it to do is skip pages containing that, thinning the list. Then I put in after that   and that should skip anything.

But it's not working, I'm not sure if it's the \n syntax that it won't accept or what. PS it might be possible to combine all of these into one single, but bigger expression: how about ? Mglovesfun (talk) 21:29, 27 August 2011 (UTC)
 * If I understand you correctly, building a list this way in AWB might be hard because you will have to log by hand all the pages it didn't automatically skip. Using the database dump tool is probably easier. If you do that, you might have to check for \r's with your \n's. A simplified regexp could be  --Bequw → τ 03:53, 29 August 2011 (UTC)
 * You're correct, every time it doesn't skip an entry, I move on to the next entry and press 'Start'. It's not really 'hard' though, I just keep checking until the list is finished. Time consuming in some cases yes, hard, no. Mglovesfun (talk) 09:47, 29 August 2011 (UTC)

URL simplifier
Does anyone know of resources that can simplify long URLs, stripping out unnecessary field-value pairs from long query strings? For instance, Google Books URLs are often really long, but we can shorten it down to include a more manageable link the wikitext, while still having it link to the correct book and page number. These aren't URL shorteners but the intent is similar. --Bequw → τ 16:41, 28 August 2011 (UTC)

Javascript minification
Has anyone debated minifying our Javascript code? I'm not sure if post-compression it would change the file sizes much, but I thought I'd ask. --Bequw → τ 19:03, 29 August 2011 (UTC)
 * Don't the Wikimedia servers serve all pages gzipped already? -- Liliana • 19:23, 29 August 2011 (UTC)
 * I believe so, but minified & gzipped is still smaller than just gzipped. See this related post on wikipedia where pre-compression minification reduces final file size by 47%. --Bequw → τ 19:43, 29 August 2011 (UTC)

Importing a batch of entries
I have about 2,500 pre-formatted English noun entries to import (specific minerals; and a small handful exist already in Wiktionary and should be skipped). What is a good way to do this, or can somebody do it for me? (Although I've used DotNetWikiBot in the past for retrieving pages, it will not let me save a page &mdash; "500 Internal Server Error" &mdash; and the inadequate documentation makes this impossible for me to diagnose.) Thanks. Equinox ◑ 15:13, 31 August 2011 (UTC)
 * Is it merely a matter of rote copying and pasting from a master document? Send me a batch and I'll help (my email is my username at hotmail). Cheers! <i style="background:lightgreen">bd2412</i> T 17:44, 31 August 2011 (UTC)
 * I mailed you a copy. I've actually pasted a bunch of them in manually since then, but it is slow going and not good for the fingers. Equinox ◑ 20:48, 31 August 2011 (UTC)
 * I can do it for you, or if you have Perl installed, I can send you code to do it yourself. —Ruakh <i >TALK</i > 20:44, 31 August 2011 (UTC)
 * Could you mail me the code, please? It's been a while since I've used Perl but I am vaguely familiar with it, enough to make any tweaks I would need. Thanks! Equinox ◑ 20:48, 31 August 2011 (UTC)

Etymologies
I am trying to pull out all Wiktionary entries of Akkadian origin. I have tried using category:en:akkadian derivations but it doesn't work, yet I know Akkadain etymologies do come up in Wiktionary entries e.g. tandoor, souq.

Is there any way I can draw out all such entries?
 * See Category:Terms derived from Akkadian. --Yair rand 21:34, 31 August 2011 (UTC)

I did already find this page but it didn't list tandoor and souq. They must have been added since. Is there any page which lists all Akkadian derivations? Is there a way I could also pull out all Aramaic and more generally, all Semitic derivations?


 * Also see Category:English terms derived from Akkadian. Besides this, you could search for Etymology Akkadian. —Stephen (Talk) 15:11, 1 September 2011 (UTC)