Wiktionary:Grease pit/2012/February

Definition from Wiktionary, the free dictionary
Jump to: navigation, search
This is an archive page that has been kept for historical purposes. The conversations on this page are no longer live.
Grease pit archives +/-

February 2012


Is there a way to put a six above a four as shown on the ⁶₄ page but on only one line and smaller instead of using a table with the images of both numbers? Celloplayer115 01:15, 1 February 2012 (UTC)

I used head=64. It's not perfect and you can do what I did on  ̥ if you have an image of the thing. —Internoob 02:49, 1 February 2012 (UTC)
How about now? —RuakhTALK 17:54, 2 February 2012 (UTC)
FWIW combining digits have been proposed for inclusion in Unicode, so in the future this ugly hack shouldn't be needed anymore. -- Liliana 17:59, 2 February 2012 (UTC)
<math>6\atop4</math> displays correctly (6\atop4) but is an image. (FYI, the spacing before and after \atop doesn't matter.)​—msh210 (talk) 21:23, 2 February 2012 (UTC)

Combining diacritics

In the entry " ̥", the title is placed in an odd position, since there isn't anything for the character to combine with. It'd be nice to have a placeholder there to combine with:

  1. In MediaWiki:Common.css, add the following rule:
    .combining:before { content: "o"; speak: none; color: #aaa; margin-right: -1ch; }
  2. Create a template {{Combining Diacritical Marks title}} containing with the following wikitext:
    {{DISPLAYTITLE:<span class="combining">{{FULLPAGENAME}}</span>}}
  3. Include the template in entries named after a combining diacritical mark. Unfortunately we can't just stick it in {{Combining Diacritical Marks character info}}, because that infobox is used in some non-combining entries.

In browsers that support CSS generated content, the title would look like:


Browsers without generated content support would be unaffected. An alternative to the letter "o" would be the dotted circle (◌).

 – Minh Nguyễn (talk, contribs) 10:19, 1 February 2012 (UTC)

No objections from me though the dotted circle would really be better than the letter o. -- Liliana 12:53, 1 February 2012 (UTC)
It would, but many fonts render the circle way too small. We could specify font families on the placeholder, but its metrics might not match the diacritic, then. – Minh Nguyễn (talk, contribs) 16:44, 1 February 2012 (UTC)
Could it be used with a no-breaking space? —CodeCat 16:48, 1 February 2012 (UTC)
A non-breaking space likely wouldn't be wide enough (depending on the user's font). Though that is an idea: without modifying any stylesheets, we could just apply padding-left: 1ch; to the title using a template. I still like the idea of indicating somehow that the character is combining, though. – Minh Nguyễn (talk, contribs) 17:25, 2 February 2012 (UTC)
The problem with using the letter o as a placeholder is that it may cause confusion with the actual combination of o + diacritic, especially if that entry exists as well. -- Liliana 16:48, 1 February 2012 (UTC)

That's why I made the "o" a light gray. But it turns out that, in Firefox on the Mac, the diacritic also turns gray if both characters come from the same font. (For me, Firefox was doing font replacement on the diacritic, so it showed up black.) Typically you'd use a dotted-line "o", which is what the Unicode standard uses in its code charts. [1] But ironically, the closest thing in Unicode itself is "◌", which doesn't render well in fonts other than Arial Unicode MS. So how about using a dotted circle and specifying that font for the entire title:

.combining { font-family: 'Arial Unicode MS'; font-size: 110%; line-height: 1em; }
.combining:before { content: "\25CC"; speak: none; margin-right: -1ch; }

Note that this is just a simulation; if these styles are implemented, the placeholder character won't be selectable, just the diacritic.

 – Minh Nguyễn (talk, contribs) 17:25, 2 February 2012 (UTC)

Wouldn't it be better to use the existing unicode class, instead of defining just a single font? -- Liliana 17:31, 2 February 2012 (UTC)
And what about those who don't have that font? —CodeCat 21:27, 2 February 2012 (UTC)
True, for now I've created the template to simply shift the title over by one character. I added it to all the entries I could find. It still would be nice to display some kind of placeholder, to indicate a combining character. – Minh Nguyễn (talk, contribs) 07:11, 5 February 2012 (UTC)

Using language templates in entries directly

As far as I know, our practice has been to subst: these templates whenever they are used directly in entries. But I'm not sure why; it seems to me that a lot of the benefits of using templates are lost this way. Can someone explain this to me please? —CodeCat 12:46, 1 February 2012 (UTC)

There are two places where these templates could conceivably be used in entries: headings, where they break section linking, and translation sections, where they make sorting unnecessarily difficult. -- Liliana 12:52, 1 February 2012 (UTC)
I'm also thinking of language names in etymology sections and in descendants. —CodeCat 12:57, 1 February 2012 (UTC)
In etymology sections you should really use {{etyl|nl|-}} or similar (which displays "Dutch"), but it isn't that widespread yet. As for descendants, same argument from translation sections applies here. -- Liliana 12:59, 1 February 2012 (UTC)
User:msh210/format.js has a 'lang subst' function, in User:Mglovesfun/vector.js I changed this to change {{nl}} to {{etyl|nl|-}}. The script looks for two or three Latin letters wrapped up in {{}}, so codes like {{roa-jer}} get ignored. Mglovesfun (talk) 13:03, 1 February 2012 (UTC)
Is there a reason why that would be preferred in etymology sections? It seems to me that it just complicates things and results in more code and slower pages without any benefit. —CodeCat 13:05, 1 February 2012 (UTC)
It gives a Wikipedia link for (1) languages that aren't linked in translations sections and (2) all languages, if you enable the relevant option at WT:PREFS. —RuakhTALK 13:32, 1 February 2012 (UTC)

Sorting out the multitude of linking templates

Right now we have a wide variety of templates used to create links. Some are more commonly used than others, some are used only to support other templates. But in many cases there are several templates that fulfill a similar function. The overlap is quite huge, as people who don't know of the other templates decide instead to create their own. It's hard to see the template forest with the trees in the way, and I wouldn't mind pruning some of them! I'd like to make an inventory of linking templates and their different functions. I think it would be nice if we had a set of small templates that perform one function and nothing else. For example, one template that creates just a link, nothing else without formatting or scripting. (I propose that we use {{l}} for that purpose, by offloading the additional things like scripts, transliterations, glosses and genders to another template (which in turn transcludes {{l}}).)

Are there any others? Please add them to the above list. —CodeCat 16:40, 1 February 2012 (UTC)

Category:Internal link templates, though some of them aren't quite 'linking templates' in this way. I created recons to do the same as {{proto}} but for any language, not just one's that start 'proto'. So you can link to Appendix:Frankish/foo or even to languages that are attested but have unattested forms, like Appendix:Old Dutch/foo. Mglovesfun (talk) 16:44, 1 February 2012 (UTC)
{{recons}} is a nice template in itself, because it can be widely used. To me it makes {{proto}} a little redundant, especially now that reconstructed languages have their own language codes. Instead of using {{proto|Germanic|dagaz}} why not use {{etyl|gem-pro}} {{recons|dagaz|lang=gem-pro}}? This practice might even make {{termx}} redundant as well, as it does the same thing. —CodeCat 16:55, 1 February 2012 (UTC)
{{termx}} is needed for Qapla' I think -- Liliana 16:57, 1 February 2012 (UTC)
Proto may do the same job as etyl and recons separately, but it's easy to use and saves a little bit of space. I find that a bit like saying {{feminine of}} is redundant to {{form of|feminine}}. Mglovesfun (talk) 10:33, 4 February 2012 (UTC)

Recent changes

What is all this rubbish at the top of "Recent changes" - after "Wanted" and before the "options"? SemperBlotto 19:56, 1 February 2012 (UTC)

Developers must be fiddling with something, it's happening at en.wikiquote as well. -- Cirt (talk) 19:57, 1 February 2012 (UTC)
Gone away now. SemperBlotto 20:01, 1 February 2012 (UTC)
Yeah, must've been a temporary glitch. -- Cirt (talk) 20:03, 1 February 2012 (UTC)

Language templates listed as unused templates

Currently most of the templates in Special:UnusedTemplates are language code templates along with their corresponding script subtemplates. This doesn't really seem desirable to me because they aren't 'unused' in the sense that they may be deleted; they're just waiting to be used. And in the mean time they clog up the list, which is limited to 5000 entries, so there is no way to see the remaining actually unused templates. Is there a way to make it so language templates are not shown in the list? —CodeCat 18:29, 2 February 2012 (UTC)

Create a page that embeds them. It can even be in the User namespace. As long as they're embedded *anywhere*, they won't show up in the list. -- Liliana 18:30, 2 February 2012 (UTC)
Oh I didn't think of that... Wiktionary:Etymology/language templates and Wiktionary:Index to templates/languages seem like good candidates, but the bot that generates those pages would need to be modified. (Why do we have both pages though, aren't they the same?) —CodeCat 18:40, 2 February 2012 (UTC)
I think you just need to stop interpreting 'unused' as 'may be deleted'. That seems like the quickest possible solution. --Mglovesfun (talk) 19:24, 2 February 2012 (UTC)
I think the issue is the interaction with the 5000 entry limit, which problem occurs in some other Special Pages lists as well. DCDuring TALK 19:34, 2 February 2012 (UTC)
Oh I see. I think 5000 is the limit for all special pages. The number of unused categories is over 1500, which is why I'm in favor of regularly deleting unused categories. Mglovesfun (talk) 19:42, 2 February 2012 (UTC)
In Wiktionary:Index to templates/languages, instead of linking the name of the language like [[English]], use [[{{en}}]]. That'll do it. Just need to get the agreement of the bot owner that updates this list. Mglovesfun (talk) 15:06, 12 February 2012 (UTC)

Import bot for translation pairs

What follows is a request for help (or input) with creating a bot for volume import of translation pairs into Wiktionary.

I am in contact with a person who is considering to donate some English-Czech translations to English Wiktionary. He asked me whether the following could be done using a bot:

Let us have a list of tab-separated records of the form <headword>\t<translation-gloss>\t<translation>\t<gender>. For each row of the table, the bot should add the translation to a suitable translation table in the entry <headword>, by finding "{{trans-top|<translation-gloss>". As the tab-separated list has been created using translation glosses extracted from Wiktionary, the bot can assume that <translation-gloss> perfectly matches the string found after "{{trans-top|". To make it generic, one column of the input table could contain language code. For simplicity: if the translation table already contains a Czech translation (or that of another specified target language), the bot should skip the record or row; this can be made more complex by requiring the bot to append the translation to the list of translations of the target language unless the translation is already in the list.

Does anyone have a bot whose parts could be used to create such an importing bot? Does writing the bot seem doable with a fairly low amount of effort?

Here is a test set for import:

pumpkin fruit of this plant     dýně    f
decade  a period of ten years   desetiletí      n
Armenian        person  Arménec m

The target language is Czech, thus "cs".

Thank you for any help. --Dan Polansky 18:22, 3 February 2012 (UTC)

  • coding modified so it won't interfere with the rest of the page Chuck Entz (talk) 14:20, 27 June 2012 (UTC)


Remember this page, which caused us so much trouble due to starting in an invisible character so it can be allowed as a page title?

Well as I just found out, Unicode has a solution for this: . This isn't the IPA colon, but a new character, which is intended for languages which use the colon orthographically. Even though it might not be supported by older systems, it would be a perfect fit in this case, in my opinion. -- Liliana 02:39, 4 February 2012 (UTC)

No idea. Use your judgment, it's been serving us well so far. Mglovesfun (talk) 14:39, 12 February 2012 (UTC)
Brilliant!​—msh210 (talk) 07:19, 13 February 2012 (UTC)
Moved to na꞉k. -- Liliana 17:26, 13 February 2012 (UTC)

Right-to-left Babel boxes

So I know WT does a lot of things different from WP (and in this case from Commons and WB as well). But is there a particular reason that the language codes for languages written right-to-left, like Arabic and Hindi are on the left rather than the right? --Quintucket 21:10, 4 February 2012 (UTC)

That's what Wikipedia does as well. Or did I miss something? -- Liliana 21:16, 4 February 2012 (UTC)
Odd. I just realized that I'm completely wrong. Wikipedia, puts the language code for rtl languages on the left for native speakers, and right for level 1 speakers, but all other wikis put both on the left. I wonder why it is that I didn't notice it when I put my Babel boxes on Commons and Wikibooks. --Quintucket 21:25, 4 February 2012 (UTC)

Georgian blends

I've just added the Georgian language tag "ka" to არაფერი to move it out of the "English blends" category and into a new category "Georgian blends". First of all, that is the right language tag, isn't it? Secondly, the new category isn't showing up here. Is there something I need to do to make it appear? — Paul G 15:37, 5 February 2012 (UTC)

Yes, that is the right language tag (/ language code). The category wasn't showing up in Blends by language because it hadn't been created yet. :) --Yair rand 16:23, 5 February 2012 (UTC)
Ah, OK :) Thanks for doing that. — Paul G 18:02, 5 February 2012 (UTC)

Bot not working

Although I can edit normally (but somewhat slow?) my bot is not working - I get "HTTPError: 504 Gateway Time-out". As far as I can tell, no bot has run since 18:11 yesterday. (They are running OK on -pedia) SemperBlotto 08:44, 12 February 2012 (UTC)

  • Now it is almost working. It adds one or two words then goes into a loop of dozens of "Pausing 300 seconds due to database server lag" messages. It will never catch up at this rate. SemperBlotto 14:34, 12 February 2012 (UTC)
  • Server lag now down to single figures. Light at the end of the tunnel. SemperBlotto 15:00, 12 February 2012 (UTC)
    • I had the same problem using AWB this morning; a couple of hours later, it was okay. --Dan Polansky 15:47, 12 February 2012 (UTC)

Template not discriminating language when categorizing

This discussion (Wiktionary:Requests for cleanup#Category:en:Chemical elements) points out that the template in question is adding the word to an en: category, even though the template is in the section for a specific non-English language and not in the English section. How hard is it to make the template aware of what language section it's in?—This unsigned comment was added by Chuck Entz (talkcontribs).

Very. But a bot can add a lang parameter based on section. (Liliana, you reading this?)​—msh210 (talk) 07:13, 13 February 2012 (UTC)
To repeat myself, the problem with this template is quite small; with context labels, I wouldn't be surprised if there were between 10 000 and 100 000 pages using context labels with the wrong language. Mglovesfun (talk) 12:04, 13 February 2012 (UTC)

Subpage problem

Talk:error/mistake distinction ought not to be a subpage of Talk:error; its content relates to that of Citations:error/mistake distinction. How do I prevent the software from treating it as a subpage of Talk:error? — Raifʻhār Doremítzwr ~ (U · T · C) ~ 12:01, 13 February 2012 (UTC)

I'm not sure that you can. I tried adding the entry (with a definition based on guesswork) - but that didn't help. SemperBlotto 12:11, 13 February 2012 (UTC)
It's strange; neither the mainspace page nor the Citations: page is affected. — Raifʻhār Doremítzwr ~ (U · T · C) ~ 12:15, 13 February 2012 (UTC)
That's because we don't have the 'subpage' feature turned on in those namespaces. (See mw:Manual:$wgNamespacesWithSubpages.) —RuakhTALK 14:49, 13 February 2012 (UTC)
Do we want subpages to be enabled for mainspace talk pages? — Raifʻhār Doremítzwr ~ (U · T · C) ~ 17:35, 13 February 2012 (UTC)
Good question. We don't use talkpages anywhere nearly as much as WP (which needs subpages for archiving), but we do use them. Conceivably, we'll need to archive one at some point. Can you think of a way to do so without the use of subpages? (The subpage feature is necessary if we're using subpages for the "Entry" link atop the archival page to link to the correct entry.)​—msh210 (talk) 18:15, 13 February 2012 (UTC)
I wouldn't know; I'm no good at this sort of thing. Do we have any mainspace talk pages with archival subpages? — Raifʻhār Doremítzwr ~ (U · T · C) ~ 04:26, 14 February 2012 (UTC)
Not that I can find. I think we have to assume that we will, though. I suppose we can put them in project namespace (Wiktionary:Talk archive/entrytitle). (Archived talk is as valid there as in the talk namespace, since it's Wiktionary-related and is no longer actual talk.)​—msh210 (talk) 16:32, 14 February 2012 (UTC)
Is there a way to use JavaScript to test for the existence of a page? If so, we can have JS test for wgNamespaceNumber=1 && wgTitle.indexof('/')>0 && exists(wgTitle) and set the CSS of body.ns-1 div#contentSub span.subpages to display:none. I think that would work (but only, of course, for users with JS and CSS, which is far from ideal), though it may need some tweaking.​—msh210 (talk) 16:56, 13 February 2012 (UTC)
But Talk:error does exist... —CodeCat 17:05, 13 February 2012 (UTC)
Right, but that JS would test for the existence of [[error/mistake distinction]]. (Come to think of it, that might not be ideal, as we sometimes have talkpages for nonexistent pages, especially where the latter have been deleted. I suppose if the JS could check for the existence of a page or a log entry therefor....) The existence or non- of [[talk:error]] has nothing to do with it.​—msh210 (talk) 18:10, 13 February 2012 (UTC)
That's certainly possible, though I'm not sure it that would be a good idea... --Yair rand 16:43, 14 February 2012 (UTC)
Technically speaking, to test for the existence of the mainspace page, we can just check if the "Entry" tab is a redlink, with something like document.getElementById('ca-nstab-main').className.search(/(^| )new( |$)/). That said — this approach seems error-prone. Note that the talk-page that sparked this discussion, Talk:error/mistake distinction, was orphaned (or rather, not-yet-parented) when the discussion began. If we want to suppress the breadcrumb for non-archive pages, I think it's simpler to check if the pagename looks like .../Archive.... (But to be honest, I'm not sure how much harm the breadcrumb really does. And that's the only effect of being a subpage, so far as I can tell, aside from parser variables like {{SUBPAGENAME}} that we're not using, and that couldn't readily be fixed by JavaScript even if we were.) —RuakhTALK 17:53, 14 February 2012 (UTC)
I suppose there's no way to escape characters in strings used as headers, but is there any look-alike Unicode character that could be substituted? It might be good to look for such look-alikes for all the syntactically-significant characters, which would be used like a non-breaking space substitutes for a space. Chuck Entz 18:45, 20 February 2012 (UTC)
There are some, yes — see mw:Help:Subpages — but that cure would be infinitely worse than the disease. —RuakhTALK 18:56, 20 February 2012 (UTC)
The bug is still present at enwiki more than 10 years after it was brought up there, so that pages like w:Talk:OS/2 still have a "backlink" to w:Talk:OS. Soap (talk) 02:53, 26 February 2012 (UTC)
That's because it's not a bug. —RuakhTALK 19:48, 28 February 2012 (UTC)


An editor has asked whether Wikipedia has a concordance of its own most-used words. This seems to me like a useful thing to create here, if we can do so. Obviously, since Wikipedia is in a constant state of flux, it would need to be date-specific. Cheers! bd2412 T 16:46, 20 February 2012 (UTC)

Someone was reaping words common on WP that we don't have entries for. (That's not the same thing, of course, but I thought I'd mention.) I forget who, or where it is.​—msh210 (talk) 19:10, 20 February 2012 (UTC)
Haven't found it. bd2412 T 23:23, 29 February 2012 (UTC)

creole languages linking to the creole cat

Languages like Category:Greenlandic Eskimo Pidgin language are placed in the blue Category:Pidgins and creole languages, but link (in their descriptive text) to a nonexistant Category:Creole and pidgin languages. This is possibly Template:etyl:crp's fault, but I can't work out how. Can this be fixed, so that the descriptive-text link is also a blue link to Category:Pidgins and creole languages?

Yes check.svg Done -- Liliana 22:19, 20 February 2012 (UTC)
Aha, thank you! And thanks for expanding the number of countries langcatboiler can handle a language being spoken in... I've started to fill up Category:Languages of the United States now, and have just about reached the old limit. :P - -sche (discuss) 22:29, 20 February 2012 (UTC)

Wiktionary trigraphs

Which language does ang stand for? Is there anywhere I can get a complete list of Wiktionary trigraphs (and digraphs) —This unsigned comment was added by Rdurkan (talkcontribs) at 21:59, 22 February 2012.

{{ang}} -> Old English. You can find a complete list here: Category:Language templates. Maro 21:22, 22 February 2012 (UTC)

unpatrolled edits in own userspace

This edit showed up in Recentchanges as unpatrolled. I marked it as patrolled, but I thought all edits by users to their own userspace were supposed to be auto-marked as patrolled. - -sche (discuss) 01:28, 23 February 2012 (UTC)

No, currently only edits to [[Special:MyPage]] and [[Special:MyPage/Sandbox]], not edits to other subpages; see Wiktionary:Whitelist?oldid=5774202. But that can be changed quite easily, if so desired, by modifying [[MediaWiki:Gadget-PatrollingEnhancements.js]]. —RuakhTALK 02:01, 23 February 2012 (UTC)
Ah. That's a good point about promotional material. OTOH, there's no reason to think promotional material couldn't just be added to a /Sandbox page, so... I would favour extending protection to all userspace pages (or removing it entirely). - -sche (discuss) 03:16, 23 February 2012 (UTC)

Edits reverted, but still marked as unpatrolled

New question: this edit showed up as unpatrolled, although by the time I looked at it, it had (long) been reverted. Is it possible/practical to mark any reverted edit as patrolled? - -sche (discuss) 07:27, 23 February 2012 (UTC)

I think that feature would be very useful too. It would also be nice if an edit were marked as patrolled if a patroller makes any further edits to the page, or perhaps just the sections that were edited before, with the assumption that if they made further changes, they also saw the edits and therefore agree with them. —CodeCat 13:36, 23 February 2012 (UTC)
de.Wikt's "Stabilversionen" (apparently an official Mediawiki feature) function that way: patrollers making edits to a page must specify if they don't want to mark it patrolled despite making a change. Patrollers can also simply approve the current version, which approves all revisions up to that point. However, that system also has the downside of not displaying the actual/current revision of a page until it's approved, which can be a long time. - -sche (discuss) 14:04, 23 February 2012 (UTC)

{{only in}} and WT:LOP

What would it take to enable {{only in}} to direct users to the subpages now at WT:LOP? Doing so would enable us to eat our cake and still have it, by letting us:

  1. keep new terms out of principal namespace until they met our standards,
  2. appear to be reasonably open to new terms, and
  3. discourage creation of full entries which then attract RfVs and RfDs.

We could even go so far as to set up a tickler system, based on a parameter in {{only in}} to remind us of when we might have hope of meeting the "spanning one year" requirement. DCDuring TALK 03:34, 25 February 2012 (UTC)

Separate issue, as long as we're modifying {{only in}}: per Template talk:only in, let's simplify it to simply take the target (e.g. "w:Word" or "WT:LOP/Word") as a parameter, rather than the convolution it presently requires. - -sche (discuss) 03:39, 25 February 2012 (UTC)
Could Conrad.Irwin have been trying to make the display consistent? How could we otherwise consistency of appearance? DCDuring TALK 04:11, 25 February 2012 (UTC)
Set the text of {{only in}} to something like Wiktionary does not have a dictionary entry for this term. See [[{{{1}}}|{{PAGENAME}}]]. Help us collect evidence of this word being used at Citations:X.? Or we could make it obvious that the link wasn't to a main-namespace place, ...See [[{{{1}}}]]. Help... That would work for any link, whether to [[Appendix:whatever]], [[WT:LOP/whatever]], [[w:Whatever]]. - -sche (discuss) 06:31, 25 February 2012 (UTC)
Would the shortcuts appear as shortcuts or expanded? DCDuring TALK 06:39, 25 February 2012 (UTC)
What do you mean by "shortcuts"? Viewers would see "w:Whatever" if they looked at the second code I suggested; they would only see "Whatever" (actually, PAGENAME) if they looked at the first code. - -sche (discuss) 06:43, 25 February 2012 (UTC)
I was focused on the second approach because I think we want to have a match between what the link says and the landing page after clicking. I guess someone would have to type in Wiktionary:List of protologisms/A-P#L to get within a couple of pgdns of the entry for "Linsanity". That looks ugly. And it doesn't really get the user very close, so that a new user might miss what we would have. I suppose diligent page maintenance (adding two- and three-letter section headings when required) would address the second concern. DCDuring TALK 07:06, 25 February 2012 (UTC)
Well, we could make the person adding the template specify a long, 'piped' string like Wiktionary:List of protologisms/A-P#L|Linsanity (and w:Whatever|whatever, and Appendix:Things#W|whatever). The would produce something that didn't look ugly. It's a long string to put in, but I wouldn't mind it: it's still easier that the current mechanism, which requires users to remember and call different templates for different places. In fact, the more I think about it, the better I like it, letting the person adding the template specify exactly where the template should point and what the displayed link should say, all with just one template and one parameter. I would change the template right now, but I'm not sure how to not break existing uses of the template. - -sche (discuss) 09:05, 25 February 2012 (UTC)
Would it work to have named parameters for the new options that bypassed the error-handling for the old? What about another template like {{only-in}} or {{only in2}}? DCDuring TALK 15:17, 25 February 2012 (UTC)
A new template would be the easiest thing (although my ideal solution would be to switch all the existing uses with the aid of a bot). I think it'll be easy enough that even I can design it, and make it accept links to anywhere: WP, Appendices, LOP... now the question is, does anyone have any objection to linking to LOP? - -sche (discuss) 19:48, 25 February 2012 (UTC)
I had been thinking that I should have gone to WT:BP first, but I was wanted to make sure that there wasn't some fatal flaw or that it would require a technically complicated solution. There might be some better idea, but at least it seems that there is a simple solution and no obvious problem. DCDuring TALK 19:54, 25 February 2012 (UTC)
I've created {{only-in}}, which can handle links to anywhere. (Even off-wiki webpages, I think, although I presume that's disallowed.) - -sche (discuss) 01:16, 28 February 2012 (UTC)
Thanks. We might need it for unattestable terms in WT:Glossary. It might be desirable for terms that appear in Citations namespace, but not principal namespace. Gee, we could have many terms that are "only in" both LoP and Citations namespace. At the very least, the LoP entry and the Citations namespace entry for a word should link to each other. DCDuring TALK 01:40, 28 February 2012 (UTC)

Implementing Wiktionary:Votes/pl-2009-03/Context labels in ELE v2

We voted to not link the display of context information and topical categorization. We have not implemented the change. To see the problem consider Category:en:Arithmetic. It contains such terms as odd, even, half, fraction, sum and more than a dozen others which clearly have senses that are not limited to an "arithmetic" usage context. Other terms are clearly so restricted, eg, rabdology, augend. Others may be more uncertain, eg, integer, divisor.

Though the fundamental problem is that the usage context labels should never have been used for the purpose of populating the topical categories, a simple solution is to edit each and every context label inserting a bit of code that allows a switch that suppresses the display by default. If a registered user wants to display topical contexts, they can do so by some appropriate preference selection. What say all? DCDuring TALK 17:14, 25 February 2012 (UTC)

Be careful, the vote does only apply to context labels, not to categories. So manually adding [[Category:en:Arithmetic]] is ok. Mglovesfun (talk) 18:38, 25 February 2012 (UTC)
Of course. Though I think the implementation of topical categories is an arbitrary and closed process, it is not inherently wrong. DCDuring TALK 19:08, 25 February 2012 (UTC)
It's subjective, but isn't all human activity subjective? Mglovesfun (talk) 19:27, 25 February 2012 (UTC)
I'm just an engineer (at heart) not a metaphysician. DCDuring TALK 19:55, 25 February 2012 (UTC)

http links in Mediawiki interface pages

Because Wiktionary can be accessed directly using https for some time now (without secure.wikimedia.org), a problem with interface (Mediawiki:) pages still including http:// links or which load pictures or scripts using http only has shown up (those should be replaced by links like //en.wiktionary.org/wiki/User:Hoo_man which are protocol relative). If Wiktionary gets accessed using https:// now, this can cause a security risk and annoying warnings might appear in some browsers. The problem has already been discussed on meta and I fixed all the smaller Wikimedia wikis. But we've decided to let the big wikis decide over that at their own. That can be resolved either by me, because of my global rights (if you reach consensus for that just contact me) or by a local administrator using pywikipedia with the following fixes.py extension (I've used it to fix all the smaller wikis, you can start the script then using python replace.py -fix:https -namespace:8 -start:MediaWiki:! -family:wiktionary -lang:en). All edits need to be reviewed per hand (sometimes manual fixes are required and sometimes fixes would break something). Because of that the admin who does that should know what he/ she does (Javascript and CSS knowledge is a must-have) - Hoo man (talk) 01:52, 26 February 2012 (UTC)

Redirect classes

Lately I notice <span dir="auto">....</span> appears in redirects from capitalized titles. Is this a bug, or is it something to do with my browser? I know it's not a big deal, most of those pages don't have any useful content, I just think if it's possible to change it should be changed. Soap (talk) 02:50, 26 February 2012 (UTC)

Fixed, thanks. Previously the page-title appeared directly inside <h1> tags, but now it appears inside <span dir="auto"> tags within said. —RuakhTALK 20:16, 27 February 2012 (UTC)

Using Template:l on rhymes instead of bare links

Could the template {{l}} be used to link to rhymes on rhyme pages, instead of bare links? This will need some bot work to update and the script that adds the rhymes would need to be changed as well. —CodeCat 19:04, 26 February 2012 (UTC)

Sounds good to me, but it would need to be used consistently. If it was only used on some rhymes pages it would break the alphabetizing system in the rhymes adding script. --Yair rand (talk) 22:17, 26 February 2012 (UTC)
Yes that's why I'm discussing it here. If it's changed, it would have to be changed all in one go, or things will start breaking... —CodeCat 22:49, 26 February 2012 (UTC)
It's a good idea... don't ask me how to actually do it. Mglovesfun (talk) 23:05, 26 February 2012 (UTC)

Google books and quotes

The search google books:"lap cheong bao" seems to treat the quotes poorly; this used to work correctly. Any ideas how to fix it? --Dan Polansky (talk) 18:43, 27 February 2012 (UTC)

It's a bug in Google Books. I'm guessing they don't plan to fix it, because they haven't kept the books.google.com search interface up-to-date with the new interface that's more closely integrated with web search. So, I've changed {{google}} (and hence {{b.g.c.}}) to link to that new interface instead. —RuakhTALK 20:25, 27 February 2012 (UTC)

Mass entry of words

I wonder if someone could guide me a little with this: I am trying to mass-enter a number of nouns with identical declinations into the Norwegian Wiktionary (my native W). I already have a handy template (called side) that creates all the necessary information when you want to create one word. A subst works fine. So opening a new entry I do like this: {{subst:side|no|sub}} and that word is created.

I have created a long list of words that do not have any definition. And a template that marks an entry as «entry to be checked for quality». These would take forever to enter manually one by one. I have no idea if one can build a template around another template that does this or if there is a specific javascript that does the trick. I am no programming expert but can read and understand templates as long as they aren't overly complicated... Any help would be welcome! --Teodor (talk) 21:03, 27 February 2012 (UTC)

no:template:side already includes {{uttale mangler}} and others; it can likewise include {{whatever the name is of the template that marks the entry as to-be-checked}}, can't it? Or am I misunderstanding your question?​—msh210 (talk) 21:06, 27 February 2012 (UTC)
Yes, I think you are talking about the actual template. This works fine. I have my own template based on this one that does include some more information. What I am talking about is creating a routine that can create, say, 50 words at a time. If all words are of the same type I should get 50 new masculine nouns at the speed of one. I want to do this as a pilot to see if it is feasible to enter literally thousands of words very fast. This, however is something we won't do until we are sure we are all right with so many words lacking a definition. That still has to be entered manually I think. --Teodor (talk) 21:13, 27 February 2012 (UTC)
Sorry, I think I have been unclear about an important detail. The list I was referring to is from a public domain list of Norwegian words. These are currently not in the Norwegian Wiktionary, or washing the two databases is fairly simple. Copying these into Wiktionary and making entries for them is what I want to achieve :) --Teodor (talk) 21:18, 27 February 2012 (UTC)
Assuming you get permission from the no.wikt community, I think you should create a bot account and run this as a bot. If you have a Perl interpreter, I can help you with this. —RuakhTALK 21:28, 27 February 2012 (UTC)
Thanks for your offer to help. I assume I would get that kind of permission. I am probably the most active contributor there at the moment. But I am not familiar with bots and Perl. Does it take a lot of effort to get the knack of it? --Teodor (talk) 21:34, 27 February 2012 (UTC)
If you have programming experience, then it's pretty simple. If not, then . . . I don't know. I guess you can try it and see? —RuakhTALK 21:53, 27 February 2012 (UTC)
You can also use Python instead if you prefer, using the pywikipediabot framework. My bot User:MewBot uses it too. —CodeCat 22:01, 27 February 2012 (UTC)
I have installed both Perl and Python and tried to suss out how to use these. Since I have very little programming experience I realise this would be a very tall order for me. So I won't try to learn them for the moment. But could someone tell me if it's possible to write a short programme which could batch submit some new articles based on a common template, a subst and a list of words to be created? I would appreciate if someone could give me a code I could try out with a few words. I could tehn post-correct these manually if anything went wrong. Best regards, --Teodor (talk) 23:44, 27 February 2012 (UTC)
If you e-mail me detailed information about the format of this list, and how to translate from that list-format to Wiktionary-entry-format, I can send you a Perl program to run. —RuakhTALK 02:47, 28 February 2012 (UTC)
That is very kind of you. The list is at :no:Bruker:Teodor605/Ordbank bmma and the template I intend to use is {{subst:sideteodorm1|no|sub}}, located at the Norwegian Wiktionary. Not sure how to link that template directly for you... That template marks an article as one that needs to be checked, as well as a number of other things, so it is okay to create a small number of articles for test purposes. However we need to discuss how to proceed before we start using it. If you write such a Perl programme I would appreciate it very much. If you are able to write comments to the places where I need to change templates etc that would also be excellent. I obviously need to use different templates for different word classes and declension styles.--Teodor (talk) 11:09, 28 February 2012 (UTC)
Ah, that's simpler than I expected. In that case, a JavaScript solution might be more convenient; you can follow these steps:
  1. Create a preload template, say, no:Mal:preload-subst-sideteodorm1-no-sub, with the contents {{<includeonly>subst:</includeonly>sideteodorm1|no|sub}}.
  2. Modify no:Bruker:Teodor605/common.js to do two things:
    • When you visit no:Bruker:Teodor605/Ordbank bmma, it can go through all the redlinks and change their URLs from http:///no.wiktionary.org/w/index.php?title=...&action=edit&redlink=1 to http:///no.wiktionary.org/w/index.php?title=...&action=edit&redlink=1&preload=Mal:preload-subst-sideteodorm1-no-sub&save=instant
    • When you visit a URL of the form http:///no.wiktionary.org/w/index.php?title=...&action=edit&redlink=1&preload=Mal:preload-subst-sideteodorm1-no-sub&save=instant, it can automatically click the "Lagre siden" button.
  3. Slowly go through the page, opening each appropriate redlink in a new browser tab.
Does that sound good to you? If so, I'm pretty sure I can put together the JavaScript you need. (I've done the same sort of thing on en.wikt before.)
RuakhTALK 12:40, 28 February 2012 (UTC)
This sounds good to me. I am just not sure if you wanted med to modify my common.js file before you start testing out a javaScript. I haven't modified an javascripts before; only copied other user javascripts to my own page (not really understanding the code).
Seems to me what you are suggesting is the same thing as what I have been doing previosuly, only without a script. I visit no:Bruker:Teodor605/Ordbank bmma and then click on a redlink, paste {{subst:sideteodorm1|no|sub}} in the edit window and save. So 50 new entries takes me about 50 minutes. And that is before the real work of writing a definition, checking the etymology and writing any translation has even begun.
Feel free to do the necessary changes to my common.js page if you think that is wise. I can try out the javascript.
Before I start any mass production of entries I am going to discuss my experience with this apporach with the other contributors at the Norwegian Wiktionary. The no:Bruker:Teodor605/Ordbank bmma list is very big, so I am dead scared to let any javascript run wild, thus creating thousands of new words lacking definitions. I would like to be able to excercise some control, like being able to decide if I want to do 1, 10, 50 or any specific number of new entries. Once again, thanks a million for you willingness to help! --Teodor (talk) 21:25, 29 February 2012 (UTC)
BTW the no:Bruker:Teodor605/Ordbank bmma list is only one of several. It contains regular declension masculine nouns starting with ‹A›. There is one for all the others letters of too, and new lists for verbs, adjectives etc. Only these have not been uploaded yet. I need to be able to substitute no:Bruker:Teodor605/Ordbank bmma with any other of these lists. Hope that isn't complicating things for you. --Teodor (talk) 21:36, 29 February 2012 (UTC)
If you copy the contents of no:Bruker:Ruakh/common.js and no:Bruker:Ruakh/common.css into no:Bruker:Teodor605/common.js and no:Bruker:Teodor605/common.css, you'll be set for no:Bruker:Teodor605/Ordbank bmma. You should find it easy to create entries as quickly or as slowly as you want. (Or, nearly so.) After you've used that approach for a bit, you can decide for yourself whether you're happy with it, or whether you'd prefer to try the bot approach instead. —RuakhTALK 22:29, 29 February 2012 (UTC)
Yes, that looks beautiful! One entry takes just one click. I was thinking more like batch submissions, but this is blisteningly fast anyway. Thanks a lot!!! Now if I want to use this approach on no:Bruker:Teodor605/Ordbank bmmb I understand I substitute that in the common.css. But if I want to use it simultaneously for many list, do I then just write several more if-statements, like
if(mediaWiki.config.get('wgPageName') != 'Bruker:Teodor605/Ordbank_bmma')
else if(mediaWiki.config.get('wgPageName') != 'Bruker:Teodor605/Ordbank_bmmb')
Correct? Hope you can guide me through that last question too! --Teodor (talk) 23:12, 29 February 2012 (UTC)
Re: "One entry takes just one click. I was thinking more like batch submissions, but this is blisteningly fast anyway": I use Firefox, which lets me middle-click on a link to load it in a new tab in the background; so when I use this sort of interface, I can quickly load a dozen links at a time, then quickly go through and close those tabs. Does your browser not support that?
Re: using it simultaneously for many lists: I've modified no:Bruker:Ruakh/common.js to be more extensible in that respect.
RuakhTALK 00:39, 1 March 2012 (UTC)
Thanks, I see how you modified the common.js file. Think I know how to modify it further now. I use Firefox too, so I'll have to look into that. Didn't know one could... I usually right-click and choose "open in new tab", but your suggestion should be even faster. --Teodor (talk) 00:56, 1 March 2012 (UTC)
Are you using a Firefox plugin to be able to middle-click to open a link in a new tab? If so, which one? I tried to to this yesterday but I couldn't find any solution. --Teodor (talk) 09:00, 1 March 2012 (UTC)
No, no plugin; it's the default behavior. To verify it, open a new tab, navigate to about:config, and type the filter browser.tabs. The browser.tabs.opentabfor.middleclcik option and the browser.tabs.loadInBackground options should both be true. —RuakhTALK 13:25, 1 March 2012 (UTC)


There seems to be some vandalism somewhere in regard to Navajo entries, but I can’t locate it. For example, the Navajo translation in goat says "{{hot spisy deder dum duderdum ow '|lang=nv|tłʼízí}}". At tłʼízí, under Noun it says "{{ hot spisy deder dum duderdum ow ' |tłʼízí| face=head | lang=nv }}". —Stephen (Talk) 15:38, 28 February 2012 (UTC)

I found it. The vandalism was in Template:nv/script. Fixed. —Stephen (Talk) 15:51, 28 February 2012 (UTC)
I have applied cascading protection to Wiktionary:Index to templates/languages/protection (warning: massive page!) and the pages of families and scripts it links to. At the moment, the protection is so high it prevents any non-admin from editing the pages. If a helpful IP wanted to add family info, that would be a problem, so other admins might consider reducing the protection to "block new and unreg users only". For now, I explained in the protection that users who need to edit the templates should comment at WT:ID. - -sche (discuss) 19:25, 28 February 2012 (UTC)
It's okay. If someone wants to change these, they can come here or to the BP. -- Liliana 22:38, 28 February 2012 (UTC)