Open main menu

Wiktionary:Grease pit

Wiktionary > Discussion rooms > Grease pit

Welcome to the Grease pit!

This is an area to complement the Beer parlour and Tea room. Its purpose is specifically for discussing the future development of the English Wiktionary, both as a dictionary and as a website.

The Grease pit is a place to discuss technical issues such as templates, Lua modules, CSS, JavaScript, the MediaWiki software, extensions to it, the toolserver, etc. It is also a place to think in non-technical ways about how to make the best free and open online dictionary of "all words in all languages".

Others have understood this page to explain the "how" of things, while the Beer parlour addresses the "why".

Permanent notice

  • Tips and tricks about customization or personalization of CSS and JS files are listed at WT:CUSTOM.
  • Other tips and tricks are at WT:TAT.
  • Find information and helpful links about modules, Lua in general, and the Scribunto extension at WT:LUA.
  • Everyone is encouraged to expand both pages, or to come up with more such stuff. Other known pages with "tips-n-tricks" are to be listed here as well.

Grease pit archives edit
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018


Contents

October 2018

Changing the order of the characters for category sortingEdit

Is creating a sortkey module the only way to do this, or is there a better option? (Apart from or perhaps even a hacky solution like using a specific sort_key replacement).

The order change in question is for Finnish, where "Å" should be before "Ä" and not after it. SURJECTION ·talk·contr·log· 14:50, 3 October 2018 (UTC)

You can have the sort_key table replace Ä and Å with arbitrary characters that have the right Unicode order, but that means that those characters will be shown in the headers on category pages. (Though if you used the Cyrillic Ӓ in place of the Latin Ä, that would at least not be visually confusing.) This seems like something too simple for a dedicated sortkey module. But you can also have Module:columns use a custom sortkey function, and the sortkey wouldn't be visible as it would be in categories. Then templates such as {{der3}} will have the correct sorting. Egyptian is the only language that makes use of this; see Module:egy-utilities. — Eru·tuon 19:58, 3 October 2018 (UTC)
Terms using å are very uncommon in {{der3}} or other table templates anyway, so categories are really the only place where this would need to be fixed. The sort_keys replacing the headers makes it seem unattractive as a solution to me, even if we use homoglyphs, which means that creating a sortkey module might be the only real way to fix it properly right now. SURJECTION ·talk·contr·log· 20:40, 3 October 2018 (UTC)
@Surjection: But what could a sortkey module do that the sort_key table can't? — Eru·tuon 20:47, 3 October 2018 (UTC)
Does using a sortkey module also affect what shows up in the header? If so, this will be more tricky than I thought. SURJECTION ·talk·contr·log· 20:49, 3 October 2018 (UTC)
@Surjection: Yeah, a sortkey module just lets you do more complicated things when the language object from Module:languages generates a sortkey. It doesn't change the behavior of the server, which always shows the first letter of a sortkey in category headers, uses sortkeys when paging through a category, etc. There've been requests for more sophisticated category features, but so far nothing has been done. — Eru·tuon 21:02, 3 October 2018 (UTC)
I could try some ideas I have - for that, I'd need some way to test different values for the sort_key variable, but it seems the module preview does not allow me to list the contents of a category. I wonder if I just have to publish the test somehow... SURJECTION ·talk·contr·log· 21:11, 3 October 2018 (UTC)
If it's just sort order that you're trying to test, you could adapt the code from Module:egy-utilities that sorts a list of words. — Eru·tuon 21:20, 3 October 2018 (UTC)
The issue is that I'd need to test whether some tricks would work for sorting in categories specifically - the insane idea I had was to use leading spaces and seeing if the MediaWiki code trims them away when sorting the categories. SURJECTION ·talk·contr·log· 21:21, 3 October 2018 (UTC)
You can create your own test categories, just delete them when you're done. DTLHS (talk) 21:28, 3 October 2018 (UTC)
One leading space is retained at least. For instance, in the category All topics, the subcategories List of topics, Body, and so on have a space as sortkey. — Eru·tuon 21:29, 3 October 2018 (UTC)

Category:Categories with invalid labelEdit

There are over 400 entries in this category. Would anyone like to fix them? SemperBlotto (talk) 09:37, 5 October 2018 (UTC)

Anything with 0 members can just be deleted. DTLHS (talk) 16:52, 5 October 2018 (UTC)

West Frisian (fy) in MediaWiki:Gadget-TranslationAdder-Data.jsEdit

Can somebody add the following line on line 294: g: ["c", "n", "p"],

This will add checkboxes for West Frisian in the gadget TranslationAdder. Robin van der Vliet (talk) (contribs) 09:50, 5 October 2018 (UTC)

Added. DTLHS (talk) 16:42, 5 October 2018 (UTC)
Thank you. Robin van der Vliet (talk) (contribs) 13:28, 7 October 2018 (UTC)

Number of equal signs for sections such as 'related terms', 'derived terms', and 'synonyms'Edit

I'm confused as to what the standard is (if there is one) for formatting the section heads of sections such as 'related terms', 'derived terms', and 'synonyms' for entries. I have seen them as regular subheadings with three = on either side, but also on other pages I will see them as sub-subheadings with four = on either side. I'd like to know if there's a standard for the whether these should be subheadings or sub-subheadings for when I add them. If there isn't a standard, maybe one should be established so that different pages' formattings match up. If anyone knows please let me know and ping me. I hope the Grease Pit is the right place for this question, if not, also please let me know and I will move it to the correct discussion section. Thanks! 2WR1 (talk) 21:41, 9 October 2018 (UTC)

If they go under a particular part of speech they should use 4. Sometimes there is no particular part of speech they can be assigned to and then they can use 3 (this is rare). Many editors either do not understand or do not notice header levels so there are a lot of mistakes. DTLHS (talk) 21:42, 9 October 2018 (UTC)
As a general rule they are one level "lower" or more indented than the PoS to which they apply. The exceptions occur if no one has troubled to split them by PoS, which occurs only(?) for related and derived terms (etymological relations), not for the semantic relations (synonym, antonym, hypernym, hyponym, etc.). DCDuring (talk) 13:01, 10 October 2018 (UTC)
I’ve sometimes wondered whether it is appropriate in some cases to place derived and related terms at the end of an entry under a level-3 heading if it’s difficult to tell which part of speech the terms relate to. — SGconlaw (talk) 14:12, 10 October 2018 (UTC)
It's certainly better than omitting them. It seems to me that the uncertainty is mostly between adjective and noun PoSes, though I could see uncertainty occurring between prepositions and adverbs, etc. But, to take an example I am a little familiar with, I don't think it would be hard to split the level-three derived terms at Chinese into level-four noun and adjective derived terms, at least not the blue-linked ones. DCDuring (talk) 14:57, 10 October 2018 (UTC)
At least in the case of Indo-European languages, adverbs are always primary over prepositions, so the derived terms should probably go there. Synonyms are specific to a sense, not even a POS, so they should use {{syn}}. —Rua (mew) 12:46, 17 October 2018 (UTC)

Wiktionary:Per-browser preferences and audioEdit

phabricator:T206942

The audio recording option for Firefox is amazing in theory but it never seems to actually save the .ogg files at Commons. Can someone fix this? I would love to use it and have tried many times to no avail (e.g. https://en.wiktionary.org/w/index.php?title=mompreneur&diff=prev&oldid=50483825) This would be a hugely valuable tool if it worked. —Justin (koavf)TCM 18:21, 13 October 2018 (UTC)

It looks like the script is User:Yair rand/AddAudio.js. @Yair rand: can you help here? —Justin (koavf)TCM 18:30, 13 October 2018 (UTC)
Cf. Commons thread (permalink). —Justin (koavf)TCM 18:30, 13 October 2018 (UTC)
Also reported by me earlier at "Wiktionary:Grease pit/2018/September#"Add audio pronunciation" still not working". — SGconlaw (talk) 20:45, 13 October 2018 (UTC)
On a separate note: I'm not happy with doing on-the-spot recordings due to inevitable background noise (and I worry about fuzzy recordings other people will make, frankly). I have a good tool where I can edit out the noise, but that will never be part of wiki workflow. It would be good to reduce the "file upload" audio approach as much as possible (APIs? tick a box for licence?) for this reason. Equinox 06:48, 14 October 2018 (UTC)

an exceptional Latin declensionEdit

The Latin noun supellex can't properly be said to be a regular noun of the third declension, because it has an anomalous extra formant -il- in its oblique~plural stem. It's close, though, close enough that it has been bodged into the standard Template:la-decl-3rd. I wanted to edit the text "Third declension." appearing above the inflection table to say something to this effect, but learned that that line is emitted by the Lua module and so this is impossible to do easily. What's the best thing to do here? It doesn't really seem sensible to go and make a new subdeclension template (like Template:la-decl-3rd-I-ignis) given that as far as I know only one word would use it. 4pq1injbok (talk) 09:44, 16 October 2018 (UTC)

The stem is unusual, but the declension is determined by the desinence, which in this situation mark it as unambiguously part of the third declension. The template output is therefore fine as is. —Μετάknowledgediscuss/deeds 22:04, 17 October 2018 (UTC)
I agree, but I am curious where the abnormal inflection comes from. That's something the etymology section should explain. —Rua (mew) 10:30, 18 October 2018 (UTC)
Ehhh, "unambiguously" is a strong word. (If we imagine that say the oblique stem of 2nd decl humus had been supplanted by that of humilis, the result wouldn't belong to any one declension unambiguously. Viewed in that light, the fact that we have a 3rd decl nom sg and a 3rd decl oblique paradigm here looks more accidental.) But that aside, if even a normal and well-represented subdeclension like the i-stems can get "i-stem" appended to the "Third declension" in the line above the declension table (Template:la-decl-3rd-I), I think an oddity like this deserves some text appended in the same place.
I'd also like to know (if there's a good answer to) how this came about, but that's a tea-room question... 4pq1injbok (talk) 10:52, 19 October 2018 (UTC)
Again, this is a matter of the stem and not of the declension. The i-stem paradigm does affect the desinence (e.g., in the gen.pl.). —Μετάknowledgediscuss/deeds 15:28, 19 October 2018 (UTC)
Humus, humilis would be more irregular than supellex, because humus looks like a second- or fifth-declension nominative singular. (If it were neuter, it would be expected to have humor- as its stem, like tempus.) The nominative singular form supellex on the other hand looks like a third declension masculine or feminine nominative singular because it has the ending -s. (It resembles lex.) Third-declension nouns often have a nominative singular with a stem different from that of the other forms, like tempus, temporis or lumen, luminis. Supellex, supellectil- doesn't conform to any of the common patterns of the other third-declension nouns with two stems (such as e in the last syllable of the nominative singular alternating with i in the other forms), but that doesn't mean it's not third-declension; the irregular nominative singular still looks like it belongs to the third declension. Maybe it is suppletive though? — Eru·tuon 19:05, 19 October 2018 (UTC)

Creating a new Spanish Reference: RANMEdit

I have been adding some entries about Medical terms in Spanish which are not present in general dictionaries like DRAE. They exist, though, in the dictionary of the Real Academia Nacional de Medicina (RANM), which is an official Spanish scientific institution. I think it could be useful to add a template such as {{R:DRAE}}, but for RANM. I personally don't know how to do this, so I'm either asking for help or for someone to do it, if anyone feels like it. I'll leave the URL to RANM's Medical Terms Dictionary here: http://dtme.ranm.es/index.aspx Pablussky (talk) 12:33, 17 October 2018 (UTC)

@Pablussky: I created a reference template for it at {{R:DTM}}. I'm not sure how useful the dictionary will be, however, because it seems that you need an account to access it. —Μετάknowledgediscuss/deeds 22:02, 17 October 2018 (UTC)
@Metaknowledge It should be {{R:es:DTM}} – no camping on three-letter acronyms, especially with a reference not likely used. Fay Freak (talk) 22:04, 17 October 2018 (UTC)
Fair point. I've moved it. —Μετάknowledgediscuss/deeds 22:06, 17 October 2018 (UTC)

Template:affix doesn't support term requestsEdit

There are a few Proto-Samic entries in Category:Pages with module errors, because of an error message that appears from {{affix}}. I left the second argument empty, to indicate a request for someone to provide it (like {{m}} does), and only provided the third. But this doesn't seem to work, so can someone please make it work? —Rua (mew) 12:45, 17 October 2018 (UTC)

@Rua: Done, for {{affix}} and for {{compound}} because maybe it could be useful there. I saw the module errors but didn't know what you meant to do there. — Eru·tuon 19:37, 17 October 2018 (UTC)
Thank you! —Rua (mew) 21:54, 17 October 2018 (UTC)

Reason dropdown list when deleting a page is brokenEdit

The Reason dropdown list when deleting a page is broken. (Well, actually it's just a custom GUI component that doesn't support basic vital functionality.) Specifically: F4 doesn't pop it open like a normal combo box, and also you can't type a letter on it to make a selection (e.g. "V" for "vandalism"): it is deeply crippled compared to the proper Windows GUI control, and has no keyboard support. Can this be fixed? Equinox 18:06, 19 October 2018 (UTC)

Typing letters works for me. Is F4 a standard key shortcut that should be expected to work in all browsers? DTLHS (talk) 18:10, 19 October 2018 (UTC)
On Windows yes. That's the native behaviour for dropdown lists. This is latest Chrome on Win64. Might try Firefox later. Anyway it worked until very recently, probably today. Equinox 20:22, 19 October 2018 (UTC)
Update: this started working again recently, it seems. Equinox 21:07, 10 November 2018 (UTC)

Remove template testcases from CAT:EEdit

Could an admin change the line

-->|Module={{#ifeq:{{SUBPAGENAME}}|testcases<!--

in MediaWiki:Scribunto-common-error-category to

-->|Template|Module={{#ifeq:{{SUBPAGENAME}}|testcases<!--

to remove template testcases (like Template:quote-journal/testcases) from CAT:E? As with module testcases (which aren't added to the category when they have module errors), sometimes the desired outcome of a template testcase is a module error. — Eru·tuon 19:50, 19 October 2018 (UTC)

  Done. I had to read through the documentation for the {{#switch}} parser function to understand why I was doing it before I felt comfortable doing it- I don't like making changes on a system-wide level that I don't fully understand. Chuck Entz (talk) 04:30, 20 October 2018 (UTC)

Sinhala FontsEdit

How does one fix the selection of fonts for the Sinhala script? For headwords some style sheet is imposing the ordering:

Sinhala Sangam MN,KaputaUnicode,KandyUnicode,Dinamina,DinaminaUniWeb,Potha,Madhura,sans-serif;

I think the offender is MediaWiki:Common.css.

The problem is that I don't have Sinhala Sangam MN, KaputaUnicode has very poor support for 'touching letters', and I would like to see by reading whether a link to a Pali entry is to the spelling with or without touching letters. (I'm lucky that hovering over a link displays the URI in a better font.) I would like to have the Windows 7 (and later?) font Iskoola Pota added in first or second place - it has a good coverage of touching letter pairs for Pali. RichardW57 (talk) 13:37, 20 October 2018 (UTC)

"Lua error in Module:category_tree/poscatboiler/data at line 38: attempt to index field '?' (a nil value)"Edit

... is displayed in some categories, e.g. Category:English list templates. --Der Jud (talk) 11:31, 21 October 2018 (UTC)

See Wiktionary:Information desk/2018/October#URGENT: Category tree not working. SURJECTION ·talk·contr·log· 11:56, 21 October 2018 (UTC)

Merging of Template:t-needed into Template:tEdit

A little while ago I proposed merging {{t-needed}} into {{t}}, by making use of the fact that linking templates request a term if you don't provide one. There was no objection to that, so I want to get it done, but I'm not sure how the end result should be. [Term?] and please add this translation if you can obviously don't look the same, so should we just use the standard "term?" text that Module:links produces? Modifying Module:links so that it displays something else, for this specific template only, would involve a lot more work. There is a technical aspect too: the current translation adder gadget may require the format that {{t-needed}} currently displays (in particular the span tags surrounding the text), and could break if the format is changed even if the wikicode in the entry stays the same. @Erutuon maybe you can help with this. —Rua (mew) 11:52, 23 October 2018 (UTC)

please add this translation if you can appears more mannerly to me – [Term?] lets it appear as though there were a lacuna. You could stick with {{t-needed}} though, there aren’t catchy reasons to do anything. You can spare workforce too and save the modules from complexity (it becomes increasingly hard to read the module code, no?) by keeping the templates split. Fay Freak (talk) 17:25, 25 October 2018 (UTC)
Perhaps, but people do already use {{t}} without a term, simply because other linking templates already work that way. Special:WhatLinksHere/Template:tracking/translations/no term. Making such usage generate an error seems counterintuitive to me. —Rua (mew) 17:29, 25 October 2018 (UTC)
My first instinct is not to modify Module:links at all and to have Module:translations output the text of {{t-needed}} when a condition is met. What should the condition be: when only a language code has been supplied to {{t}}? — Eru·tuon 19:18, 25 October 2018 (UTC)
I think this change might result in less Lua memory usage, which is a good thing. At the moment {{t-needed}} invokes two separate module functions that use language data (via {{categorize}} and {{#invoke:languages/templates|getByCode}}); this would be reduced to one if it's merged into {{t}}. As each invoked module function adds a little overhead, fewer of them might mean less memory usage, even when they are doing roughly the same work. — Eru·tuon 19:46, 25 October 2018 (UTC)
Now that’s what I call a reason to merge. Fay Freak (talk) 20:26, 25 October 2018 (UTC)
I realize I misunderstood you, Rua, when I wrote my first post (as I guess I'm wont to do), so now Module:translations/sandbox outputs the text of {{t-needed}} under certain conditions, which isn't what you were proposing. Oh well. I'll have to look into your actual proposal. — Eru·tuon 00:17, 26 October 2018 (UTC)
I looked at the translation adder gadget, and it does mention the class name that is added by {{t-needed}}. I don't really understand the gadget though (yet?) so I can't say whether it needs the class or whether it could be changed to look for the term request instead. — Eru·tuon 23:52, 26 October 2018 (UTC)

teEdit

This page te has "not enough memory" Lua errors. I have already added it to the exclusion list in the source code of {{redlink category}} and it helped a little. Are there any other tricks to make these errors go away? —Internoob 04:57, 25 October 2018 (UTC)

Probably replace the uses of {{der3}} and {{rel3}} with manually balanced and sorted lists. DTLHS (talk) 05:01, 25 October 2018 (UTC)
Removing the linking templates from inside {{der3}} reduced memory enough to remove the module errors for now. {{der3}} creates the same links as {{l}} but with less overhead. In general it's a good idea to reduce the number of linking templates if you can, because each one adds a bit of overhead. Another way to reduce memory is to replace {{der3}} with {{der3-u}}, which doesn't alphabetize the words. — Eru·tuon 05:53, 25 October 2018 (UTC)

Missed masculine and feminine counterpart parameters in some headword templatesEdit

Pray, can someone add |m= and |f= to {{sh-noun}}/Module:sh-headword (@Vorziblix?)? I would have used it often already, like adding poslanica to poslanik, as is done with {{pl-noun}}, {{uk-noun}}, {{be-noun}}{{ru-noun}}, {{sl-noun}}, {{dsb-noun}} and others, to name only the Slavic, I have checked all: {{bg-noun}}, {{mk-noun}} (Martin123xyz would surely have used them heavily already), {{cs-noun}}, {{cu-noun}}, {{sk-noun}}, {{rue-noun}} lack them too, {{hsb-noun}} does not even exist, ignoring some less well known languages … I wish the head templates would be unified a bit more, one would learn and use them faster. At least this random defect can be remedied. Fay Freak (talk) 18:04, 25 October 2018 (UTC)

@Fay Freak Unfortunately I still haven’t gotten around to learning Lua properly, so it might take me a good while to figure out what’s going on in the code of Module:sh-headword and add the missing parameters. You might have more success asking Erutuon or Rua, since they’re the ones who worked on the module, although perhaps it’s best not to pester Rua given that she says she ‘will not solve technical problems’ on her user page. — Vorziblix (talk · contribs) 08:47, 26 October 2018 (UTC)
@Vorziblix: Done, I think. — Eru·tuon 20:05, 26 October 2018 (UTC)
Thanks! — Vorziblix (talk · contribs) 20:20, 26 October 2018 (UTC)
Thanks! — Fay Freak (talk) 20:31, 26 October 2018 (UTC)

Can not interwiki link categoriesEdit

When I want to interwiki link a category such as sv:Kategori:Kroatiska/Djurungar (Croatian words for baby animals; in Swedish Wiktionary) and its counterpart in Russian Wiktionary, I click "add links" in the left margin and the normal pop-up appears, but when I try to enter the language code (sv or ru), this is just dead text. No text completion happens and the next target field is greyed-out. This has been so for a full day now. What's wrong? --LA2 (talk) 19:11, 26 October 2018 (UTC)

[1] DTLHS (talk) 22:27, 26 October 2018 (UTC)

-- 周恩恩來; -- 陶行行知; -- 賈平平凹; 路過-- 終有一日,村子裡來來了一個天文學家。他在我家門前路過,突然發現發現了這塊石頭,眼光立即就拉拉直了。Edit

There is a problem with zh-x that is causing a duplication of 恩, 行, and 平. It happens when I add the dash between 恩 and 來, 行 and 知, and 平 and 凹. I don't think I'm making any mistakes. Conversation moved here from Beer parlor. --Geographyinitiative (talk) 02:06, 26 October 2018 (UTC) (modified)

You can see a similar, probably related problem in the second quotation for , sense #4. Maybe more an issue for the Grease pit.  --Lambiam 09:43, 26 October 2018 (UTC)
I very distinctly remember being able to use this...
  • CHAR-CHAR
    (Mandarin) Produce pinyin of CHARCHAR but link to CHAR and CHAR.
...dash between characters without causing this appalling duplication of characters! I don't think I'm making any mistakes, but look at the new example I added at 路過- truly 'sad'! --Geographyinitiative (talk) 03:35, 27 October 2018 (UTC)
了不起-- 他跑完完了馬拉松 Tā pǎo wánle mǎlāsōng --Geographyinitiative (talk) 03:52, 27 October 2018 (UTC)
-- 佘賽賽花 Shé Sàihuā --Geographyinitiative (talk) 07:04, 27 October 2018 (UTC)
@Geographyinitiative: Okay, I think I found the reason for the problem. Let me know if it's fixed. — Eru·tuon 07:37, 27 October 2018 (UTC)
Whoops, other freaky stuff is happening. I'll work on it. — Eru·tuon 07:42, 27 October 2018 (UTC)
As far as I can tell, you have solved the problem. Thanks! --Geographyinitiative (talk) 08:16, 27 October 2018 (UTC)
The problem at remains. {{zh-x|人手.一冊|It's a bestseller}} produces “手一冊 / 手一册rénshǒu yīcè ― It's a bestseller”. It is not obvious, but there are four red links there.  --Lambiam 11:09, 27 October 2018 (UTC)
@Lambiam: I added that to the informal list of testcases on Module:zh-usex/documentation. If you note any other buggy examples, please add them there. — Eru·tuon 20:03, 27 October 2018 (UTC)
Fixed. It was because of my changes. — Eru·tuon 20:19, 27 October 2018 (UTC)
@Erutuon Seems to be still broken on 武林 .. thanks! Wyang (talk) 05:28, 3 November 2018 (UTC)
@Wyang: The hyphen isn't being removed by the module because it's not between two Han characters after the module has added bolding around the term in the pagename (internally, <b>武林</b>-高手. Do you know if all hyphens (-) in {{zh-x}} mean "link separately but don't add a space to the pinyin"? That would make this easy to fix: all hyphens can be removed. — Eru·tuon 06:07, 3 November 2018 (UTC)
Oh, actually this version from this summer assumed that all hyphens can be removed. So I can proceed. — Eru·tuon 06:10, 3 November 2018 (UTC)
Thanks! Wyang (talk) 06:43, 3 November 2018 (UTC)
@Erutuon: I don't think we can assume that all hyphens can be removed. I was trying to look for an example, but I couldn't until @Mar vin kaiser edited 阿啄仔 recently. The word chit4-tho5 needs its hyphen retained in the example there. — justin(r)leung (t...) | c=› } 17:37, 9 November 2018 (UTC)
@Justinrleung: I checked and the previous version of the module linked above did manage to keep those hyphens. It only removed all hyphens from a word if that word contained a pair of bold tags, or if it contained a hyphen between two Han characters. I'm not satisfied with that solution because it doesn't seem to get at why the hyphens should be kept. Is it that hyphens should be kept if they're a part of romanization? I'm not totally sure how to determine that reliably, but maybe by checking if the hyphens touch a Latin letter or a digit, or are separated from one by bolding. — Eru·tuon 20:04, 9 November 2018 (UTC)
@Erutuon: I think hyphens should be retained only if both characters surrounding the hyphen (not including bolding) are non-Chinese (i.e. Latin letter or digit). — justin(r)leung (t...) | c=› } 20:30, 9 November 2018 (UTC)
@Justinrleung: Okay, done, I think. — Eru·tuon 00:26, 10 November 2018 (UTC)
@Erutuon: Looking good! Thanks! — justin(r)leung (t...) | c=› } 03:57, 10 November 2018 (UTC)

Lua error: bad argument #1 to 'match' (string expected, got boolean)Edit

--Geographyinitiative (talk) 07:20, 27 October 2018 (UTC)

Reduplication template? (continued)Edit

Continued from Wiktionary:Grease_pit/2015/May#Reduplication_template?.
Is there any template for reduplication for Wiktionary:Grease_pit/2015/May#Reduplication_template? I will be grateful if someone can make a template {{reduplication}} to be used in etymology sections (not a definition-line form-of template, though, which is why it probably shouldn't be called {{reduplication of}}) that will categorize terms into the language's reduplication category. --Xbypass (talk) 05:31, 28 October 2018 (UTC)

Done. It does not take the terms however as templates like {{affix}} does, so you will use {{m}}, but one likely writes some additional words between the terms and the word “reduplication” anyway or maybe there isn’t even a known term where the reduplication is from (often it certainly was a reduplication from the beginning). In other words I don’t know what they tossed around in 2015, I have created the template in ten minutes. @Xbypass Fay Freak (talk) 20:33, 28 October 2018 (UTC)
Thanks a lot. Is it possible to make the {{reduplication}} like {{calque}} which has source word, translation and transliteration? Regardless of that, this is pretty helpful than before. --Xbypass (talk) 16:03, 29 October 2018 (UTC)

Egyptian hieroglyphs in etymologyEdit

Is there a way of making these smaller, and so a large gap is not created beneath them? For instance, see all the entries for basalt. DonnanZ (talk) 12:09, 28 October 2018 (UTC)

As far as I know, not really, unless we use images in place of WikiHiero. —Μετάknowledgediscuss/deeds 19:28, 28 October 2018 (UTC)
There is Unicode, which looks better on my machine (and for search engine optimizers …). I cannot say how much is covered by Unicode however. Fay Freak (talk) 20:35, 28 October 2018 (UTC)
There is no large gap with the English example, which is odd, it may be something to do with the added reference. DonnanZ (talk) 21:02, 28 October 2018 (UTC)
I had not counted Unicode, given that the encoding does not allow for arranging the characters properly, but I just remembered that Unicode is expanding the Egyptian blocks, so that may be (or will soon be) an option as well. —Μετάknowledgediscuss/deeds 21:24, 28 October 2018 (UTC)
Unicode will be able to properly stack hieroglyphs starting in March 2019, if I’m not mistaken. (Its repertoire of glyphs will still be small, but then, so is WikiHiero’s.) Until then we’ve got to make do with WikiHiero. I tried an extremely ad hoc solution using images over at basalt, which seems to work at yielding smaller glyphs, but I don’t think it’s sensible to do that and make a mess of code every time Egyptian hieroglyphs appear. In any case, the ‘large gap’ under the glyphs was not much larger than the gap before any other L3 header (28 pixels for glyphs versus 25 for Latin-alphabet text in my browser). — Vorziblix (talk · contribs) 22:01, 28 October 2018 (UTC)
Oh, never mind, I misinterpreted what you were saying; I see the problem (you were talking about the non-English entries primarily, and I was just looking at the English one). The full stops after the glyphs are the issue. Because glyphs aren’t rendered inline unless the block of text containing them is enclosed in a <div> element, the periods are getting placed on their own separate lines. I’ll go fix that. — Vorziblix (talk · contribs) 22:04, 28 October 2018 (UTC)
The unsightly huge gap should be gone now. (I’ve switched back to using WikiHiero glyphs, though, as I don’t think the ad hoc image solution is the right way to go.) — Vorziblix (talk · contribs) 22:11, 28 October 2018 (UTC)
I'll be damned, simple full stops were the spanners in the works. That's something to remember, I'm sure there's more like that. Anyway, thanks a lot, that leaves the question of hieroglyph size... DonnanZ (talk) 22:30, 28 October 2018 (UTC)

WikiHiero just brokeEdit

The inline display of WikiHiero hieroglyphs inside divs and tables has been repeatedly breaking and unbreaking over the past few hours. It’s currently broken. (See, for example, the ‘alternative hieroglyphic writings’ templates and quotes at jz.) Does anyone know what’s going on? — Vorziblix (talk · contribs) 22:24, 31 October 2018 (UTC)

Ah, I figured it out; the MediaWiki devs have inexplicably removed .mw-hiero-outer { display: inline-block; } from the WikiHiero css. I’ve raised a concern over at Phabricator, but if they don’t get around to fixing it in a few days over there, I’ll re-add that to the common css over here instead. — Vorziblix (talk · contribs) 00:28, 1 November 2018 (UTC)
@Vorziblix I just noticed that the Hieroglyph script broke. Thank you Vorziblix, I really appreciate your help. Aearthrise (talk) 14:45, 1 November 2018 (UTC)
@Aearthrise: No problem! Until it gets properly fixed for everyone, you can fix it for yourself alone by adding .mw-hiero-outer { display: inline-block; } to User:Aearthrise/common.css. — Vorziblix (talk · contribs) 15:53, 1 November 2018 (UTC)
Fixed now. — Vorziblix (talk · contribs) 16:12, 2 November 2018 (UTC)

NEC not workingEdit

Not the only thing that's broke (referring to the above), New Entry Creator isn't working when I click on "Create entry". It's not the first time this has happened. DonnanZ (talk) 23:04, 31 October 2018 (UTC)

I think the New Entry Creator code is located here. When I click the "create it using the New Entry Creator" link on the search page, I'm seeing a message in the browser console Error: "Unknown dependency: mediawiki.toolbar". mediawiki.toolbar is mentioned in MediaWiki:Gadgets-definition as a dependency of MediaWiki:Gadget-legacy.js, and MediaWiki:Gadget-legacy.js loads the new entry wizard, so I guess the inability of the JavaScript apparatus to load mediawiki.toolbar is preventing the new entry wizard from being loaded. Why mediawiki.toolbar can't be loaded, I don't know. — Eru·tuon 23:35, 31 October 2018 (UTC)
  • Other things may have been fixed, this may have been forgotten about. It's still not working. DonnanZ (talk) 13:30, 2 November 2018 (UTC)
    @DonnanZ: Could an interface admin remove mediawiki.toolbar from the dependencies for "LegacyScripts" at the top of MediaWiki:Gadgets-definition? I can't figure out how that could be causing the problem with the New Entry Creator, but it's the only place where I can find mediawiki.toolbar in the MediaWiki namespace. — Eru·tuon 15:48, 2 November 2018 (UTC)
@Erutuon: I’ve removed it. — Vorziblix (talk · contribs) 16:15, 2 November 2018 (UTC)
@Vorziblix: Thanks! That fixed the NEC. — Eru·tuon 16:38, 2 November 2018 (UTC)
@Erutuon, Vorziblix: The ping didn't work, wrong address but not to worry. Yep, up and running again, it makes the task of writing ==Norwegian Bokmål== easier for a start. Thanks chaps. DonnanZ (talk) 18:09, 2 November 2018 (UTC)

Edittools is brokenEdit

Unable to insert characters ... today. (Chrome) Wyang (talk) 23:42, 31 October 2018 (UTC)

@Wyang: MediaWiki:Edit.js handles Edittools. It uses mw.toolbar, which has been removed, so maybe that is why it's not working, or maybe the JavaScript error mentioned above is causing a crash even before Edit.js is loaded. — Eru·tuon 00:23, 1 November 2018 (UTC)
Same problem on ca.wikt I solved it just removing mediawiki.toolbar from MediaWiki:Gadgets-definition and a hard purge of my brower. Not sure of collateral effects. --Vriullop (talk) 12:16, 1 November 2018 (UTC)
  • Fixed it. The devs have removed classic toolbar and the javascript libraries it used. Dixtosa (talk) 21:04, 1 November 2018 (UTC)
    Works now. Thanks! Wyang (talk) 23:05, 1 November 2018 (UTC)

General, serious sluggishness in basic functionalityEdit

Using Chrome, sometimes for hours at a time:

  • When I click on a Citations tab, I am unable to access it, or to stop downloading using the browser. My only recourse seems to be to close the browser tab. The option of then reopening the tab is not available.
  • The edit window is sluggish and sometimes requires that I click to make visible the characters that I have entered.
  • Some pages take very long to load.

Have others had similar problems? DCDuring (talk) 12:17, 1 November 2018 (UTC)

Problem has gone away without any action on my part. DCDuring (talk) 17:12, 1 November 2018 (UTC)

Wiktionary is brokenEdit

Everything on the site broke. I can't even edit the Grease Pit. --XY3999 (talk) 18:11, 2 November 2018 (UTC)

It is the website finally pushing back against your relentless campaign of new/blocked usernames. - TheDaveRoss 18:17, 2 November 2018 (UTC)

November 2018

Character table below the editing box no longer worksEdit

All of a sudden, the character table that appears below the editing box, which one can use to insert characters by clicking on them, no longer works for me. — SGconlaw (talk) 03:13, 1 November 2018 (UTC)

mw.toolbar has been removed, someone needs to update the script with whatever the replacement is. DTLHS (talk) 03:24, 1 November 2018 (UTC)
Gaah. — SGconlaw (talk) 03:25, 1 November 2018 (UTC)
I'm not sure what characters you want. The editing pop-down "Special characters" still works, if it's language characters you want you can install keyboards for different languages if you use Windows. I have installed six different keyboards, the problem is finding out which key for which character. DonnanZ (talk) 12:40, 1 November 2018 (UTC)
Well, for starters, I can't find a long s in any of the menus. If I've missed it, please point it out to me. I might find other missing symbols in the next couple of days, but apart from that I guess I'm just used to the character table. — SGconlaw (talk) 16:39, 1 November 2018 (UTC)
The problem appears to have been fixed. Thanks to whoever did it! — SGconlaw (talk) 03:04, 2 November 2018 (UTC)

Reference placementEdit

I can't work out why a reference in the etymology of English halogen is now appearing below other languages, both Norwegian (today) and Czech have been added in the last fortnight. It must be something to do with the ref itself, no problem before the latest additions. DonnanZ (talk) 16:58, 1 November 2018 (UTC)

You need to put
===References===
<references />
inside the English section. Right now there’s only one at the end of the page. — SGconlaw (talk) 17:05, 1 November 2018 (UTC)
That fixed it, thanks. DonnanZ (talk) 17:31, 1 November 2018 (UTC)

Collapsed articlesEdit

It has happened before, but somehow got cured. Now it seems permanent: all the articles appear in collapsed view (only the title is shown and the "Unchanged" button that doesn't work). How am I supposed to fix it and return to "unchanged" view? Al Silonov (talk) 08:51, 2 November 2018 (UTC)

@Al Silonov: There have been a number of changes behind the scenes in MediaWiki which have been breaking things, can you try disabling all of your preferences and gadgets and then re-enabling them one-by-one to determine which of them (if any) are breaking the functionality? Also, what browser and OS are you using? - TheDaveRoss 12:53, 2 November 2018 (UTC)
Thanks, I'll try... Al Silonov (talk) 13:22, 2 November 2018 (UTC)

Can the language name be suppressed when using {{inh}}, {{bor}}, {{der}} etc.?Edit

The new etymology templates are very nice and useful, but it seems to me that they have one disadvantage compared to the old {{etyl}} {{m}} sequence, namely that you can no longer omit the language name. Now, I realize one would not normally want to do this, but in some cases it seems to me it would be useful. For instance, if we have a term deriving from a proto-language, but we have two different reconstructions of what the ancestral term was in the proto-language, we'd want the text to say e.g. "Inherited from Proto-Language *sample or *sahmple …", but it seems to me that this isn't possible using the {{inh}} template or the other new etymology templates. I've not been able to find a way to stop it from writing out the full name of the language for each invocation. Is there an undocumented parameter that I'm missing? --Pinnerup (talk) 00:17, 3 November 2018 (UTC)

It is usually just {{inh|foo|ine-pro|*foo}} or {{m|ine-pro|*fooh}}.Suzukaze-c 00:20, 3 November 2018 (UTC)
We already have to clean up Category:etyl cleanup no target/language, a lot of these can be substituted with {{cog}}. DonnanZ (talk) 00:42, 3 November 2018 (UTC)
{{m}}, though for machine-readability it would be better to use instead {{der}} etc. with suppressed language name. Or you put both reconstructions into the same template call and put square brackets around them. Trying to do this annoys you with bidirectionality problems however if the script linked is RTL. Fay Freak (talk) 00:52, 3 November 2018 (UTC)
You shouldn't do this. If things are separate terms, they should be in separate template calls. If multiple terms are wrapped in one template, then that is to be understood as a single combined term, which is probably not what you want.
As for machine readability, the first term (the one wrapped in {{inh}}) should always be the lemma and any following terms should be alternative forms of that lemma. A bot should only follow this first link, and will find alternative forms on the entry there. The following links are only for humans. —Rua (mew) 19:24, 9 November 2018 (UTC)

Latin declension tablesEdit

It looks like somebody fucked them up.

  1. Compare capio (noun) or captus (participle, noun) with capio (verb form) or Brücke: The former isn't collapsable while the latter is (if you have javascript).
  2. Compare Pontus with capio or captus: The former lacks proper table design with borders and colours.

... --80.133.101.81 17:43, 3 November 2018 (UTC)

The second was easy to fix- a stray ")", apparently added by accident. The first, however, doesn't seem to be anything new- I don't see any evidence that declension tables were ever collapsible, and I don't remember them being that way, either. Chuck Entz (talk) 19:29, 3 November 2018 (UTC)

Template:setnEdit

I created this because I couldn't find an existing template to perform the simple and basic function of changing the line numbers in an ordered list created by # in the wikitext. w:Help:Lists had a trick using an html tag which seems to work fine, so I incorporated it into a template. Still, any time something this basic and obvious is missing, it makes me nervous.

My questions:

  1. Do we already have something that I missed?
  2. Is there a technical reason I'm unaware of not to do this?
  3. Is this going to encourage bad practices by editors that might make it a bad idea?

I'm rather incompletely self-taught when it comes to HTML and Mediawiki, so I don't want to unintentionally introduce something based on an undocumented and ephemeral quirk of current implementations, nor do I want to create invalid HTML, even if it works on some browsers. I also certainly don't want to make entries harder to work with.

Any thoughts from those who know more than I do? Chuck Entz (talk) 23:37, 3 November 2018 (UTC)

You haven't provided any actual example of where you'd want to use this. DTLHS (talk) 00:32, 4 November 2018 (UTC)
Not into any entries, but I've been working on splitting some frequency-list appendices that have too many expensive parser function calls. The results aren't quite where I'd like them to be, but at least you can see the template in action: Special:WhatLinksHere/Template:setn, with the original pages linked to from Appendix:Mandarin Frequency lists. The main potential bad usage I can see is that they allow the numbers on definition lines to start up where they left off when there's something in between to interrupt the sequence. It might encourage people to put things where they don't belong, since it provides a way to fix the disruption to the numbering- but by hard-wiring in a number that won't reflect changes to the lines above it. Chuck Entz (talk) 02:00, 4 November 2018 (UTC)

Problem with Template:eo-headEdit

Template:eo-head seems to be miscategorizing nouns, adjectives, and adverbs that end in ato, ata, ate, ito, etc. It is evidently assuming based on this ending that they're participles, when in fact there are plenty of Esperanto nouns, adjectives, and adverbs that just happen to end in these letters. This problem can be seen, for instance, at ŭato and mito. —Granger (talk · contribs) 15:20, 6 November 2018 (UTC)

@Mx. Granger: I've changed the participle-detection pattern so that the verb stem has to be at least two letters long. (I looked through part of Cat:Esperanto verbs for one-letter verb stems and didn't find any, though that's not quite systematic.) It fixes the two examples that you mention, at least. If there are more non-participles being categorized as participles, you can probably override the part of speech with the |pos= parameter in {{eo-head}}, rather than using {{head}}. — Eru·tuon 03:35, 7 November 2018 (UTC)
On azoto I used "pos=noun" but the accusative became the same, what is the reason for this? J3133 (talk) 07:17, 7 November 2018 (UTC)
Thanks for the response. I'm afraid requiring two-letter verb stems doesn't solve the problem at all—the two examples I gave happened to be four letters long, but there are many other words with this problem, such as akurata and monato. Using {{eo-head}} with the override parameter solved the problem in akurata, but it's inconvenient to have to do it in such a complicated way instead of just using {{eo-adj}}.
I have no idea what the problem is with azoto, but it needs to be fixed. —Granger (talk · contribs) 08:55, 7 November 2018 (UTC)

Fatal exception of type "UnexpectedValueException"Edit

I keep getting errors like [W@OLHwpAAEEAAEYSby8AAAAE] 2018-11-08 01:02:23: Fatal exception of type "UnexpectedValueException" like when trying to preview or save changes. Started happening just a few minutes ago. Happens even when I do a random test, like start editing a new page, e.g. yaddasnorkelweaselblah, put some test content into it, and click "Show preview". It's not happening with preview on this discussion page. — SMcCandlish [talk] [cont] ‹(-¿-)› 01:08, 8 November 2018 (UTC)

[2] DTLHS (talk) 01:49, 8 November 2018 (UTC)
Ah, thanks for the pointer. — SMcCandlish [talk] [cont] ‹(-¿-)› 16:54, 8 November 2018 (UTC)

Editor preview now defaulting to some broken beta feature?Edit

I've changed no settings, and suddenly, the Show preview button no longer previews the content of the edit textbox, but instead generates some broken edit conflict diff resolution mess. I was editing a section over in the Tea Room a moment ago and the preview suggested I was trying to replace the whole page with my tiny little edited bit. I hit Back and then Publish changes and, lo, just the section I was trying to edit was actually changed. I see just now that Show preview is causing the same behavior here too.

What the heck is this beta feature? Where did it come from? Why is it enabled by default? How do I make it go away? And who do I talk to (sternly) about not forcing such incomplete and broken features on users with no warning? ‑‑ Eiríkr Útlendi │Tala við mig 02:01, 8 November 2018 (UTC)

Wow, spontaneous enablement? You can disable it in the Beta features tab of your Preferences ("Two column edit conflict"). It used to work, aside from some bugs. Looks like you've already found the discussion at Help talk:Two Column Edit Conflict View on MediaWiki. — Eru·tuon 02:39, 8 November 2018 (UTC)
Thanks for the pointer! I've now disabled this. I just tried Show preview here and things are back to the expected behavior. Cheers! :) ‑‑ Eiríkr Útlendi │Tala við mig 02:52, 8 November 2018 (UTC)
I'd had miscellaneous, mostly minor problems with Beta features. I unchecked the "Automatically enable all new beta features" box and review the beta features tab every month or two. Those with more patience than I have are performing a community service by being guinea pigs. It helps if you can efficiently articulate any complaints and suggestions to the right parties. Thanks. DCDuring (talk) 17:18, 8 November 2018 (UTC)

Problem in Limburgish conjugation templateEdit

I'm having in issue at li.wikt concerning a conjugation template. Template:V-w-del seems to work just fine on beuke and blebbe, but not on aaddinke. The problem arises in the IPA of the form "gaaddink" (deilwaord - vergangen tied). For some reason, an extra break is inserted before the IPA input, while that doesn't occur in the form "gaaddinke" (participium), which works with the same parameters (fon-pref and fon-pref2). Does anyone know what is the problem, and how it can be fixed? --Ooswesthoesbes (talk) 10:20, 8 November 2018 (UTC)

Qualifiers in Module:homophonesEdit

Can someone please add qualifiers in Module:homophones, like in Module:nyms? It’s needed so {{homophones|lang=en|cot}} {{qualifier|accents with cot-caught merger}}, {{l|en|court}} {{qualifier|non-rhotic accents}} can become {{homophones|lang=en|cot|q1=accents with cot-caught merger|court|q2=non-rhotic accents}}.Jonteemil (talk) 15:47, 8 November 2018 (UTC)

All of these really would be awesome of they were implimented:

ParametersEdit

|1=
The language code (see Wiktionary:Languages) of the language whose sense this appears under.
|2=, |3=, |4=, ...
One or more synonyms to be listed.
|alt1=, |alt2=, |alt3=, ...
Link text for each of the synonyms, if different from the entry name.
|tr1=, |tr2=, |tr3=, ...
If necessary transliteration for each of the synonyms, some languages are done automatically.
|q1=, |q2=, |q3=, ...
If necessary, qualifiers for each of the synonyms.

Jonteemil (talk) 17:53, 8 November 2018 (UTC)

I've implemented trN and qN. There isn't a good way to do first parameter as language code, since it wasn't required (kept it like that so as not to cause module errors and added tracking). If we see usage without language code isn't wide spread, someone with more knowledge than me should be able to do this relatively easily. —Enosh (talk) 20:37, 10 November 2018 (UTC)
@Enoshd However, your addition of Module:parameters has resulted in over 150 module errors (see CAT:E). Chuck Entz (talk) 06:44, 11 November 2018 (UTC)
@Chuck Entz: Fixed. (I oughta make a function for this.) — Eru·tuon 08:16, 11 November 2018 (UTC)
@Erutuon Not that I'm a big fan of Module:parameters (the image of swatting a fly on someone's head with a sledgehammer comes to mind), but if you're going to be changing it to log-only in some cases, perhaps we need to have a hidden category so people will know what needs to be cleaned up- perhaps something like "Templates with unrecognized parameters". Chuck Entz (talk) 01:33, 12 November 2018 (UTC)
@Chuck Entz: You're right, it needs to have a category or tracking template. A category is easier to notice on the page, but a tracking template is easier to add. Since logging unrecognized parameters is temporary, I've just added a tracking template.
I'm puzzled because several pages had the |sort= parameter but I can't find any cases in a search (hastemplate:homophones insource:/\{\{homophones[^}]+?\|sort=/). Maybe someone removed them. If any remain, they'll surface. Japanese probably does need a |sort= parameter. — Eru·tuon 02:19, 12 November 2018 (UTC)
Try searching for {{hmp}} instead of {{homophones}} Chuck Entz (talk) 02:25, 12 November 2018 (UTC)
Perfect, thank you!Jonteemil (talk) 12:43, 12 November 2018 (UTC)

U+2019 in notWordPuncEdit

Is there any opposition to adding U+2019 (RIGHT SINGLE QUOTATION MARK) into notWordPunc in Module:headword? This change would eliminate it from the punctuation marks, automatically preventing stuff like this (just look at the headwords; they shouldn't be linked separately because they aren't separate words). SURJECTION ·talk·contr·log· 14:03, 9 November 2018 (UTC)

Why do you use this character then and not U+02BC MODIFIER LETTER APOSTROPHE which should be used in its place?
And Ukrainian lemmas, the same: обо́в'язок (obóvʺjazok), з'їзд (zʺjizd) should use U+02BC MODIFIER LETTER APOSTROPHE, this stroke is just an equivalent of ъ. The fact that after I doubleclick onto the word and my browser marks only half of it though otherwise all of the word is another example why it should be U+02BC MODIFIER LETTER APOSTROPHE. Fay Freak (talk) 18:15, 9 November 2018 (UTC)
@Fay Freak: Per Talk:п'яны, there seems to be some opposition to using the curly apostrophe. See also User talk:Mahagaja/Archive 18 § Apostrophes in French and Belarusian. Per utramque cavernam 19:11, 9 November 2018 (UTC)
Same with Macedonian entries beginning with an apostrophe. Per utramque cavernam 19:11, 9 November 2018 (UTC)
Different functions of the apostrophe require different apostrophes. Thus unlike Mahāgaja says there should be different standard apostrophes for different languages, even multiple apostrophes in the same language if applicable. You have correctly used U+2019 in French entries. Ukrainian and Belarusian have to use U+02BC because of the behaviour of the characters. It’s only marginally about curliness. If Surjection wants the apostrophe to be treated like a letter it is likely that U+02BC is the intended character. See also Ukrainian Stackexchange. Some people say English has to use U+02BC too, but I have no opinion on this and it does not matter on Wiktionary any more. Fay Freak (talk) 19:30, 9 November 2018 (UTC)
We use U+2019 because that is the standard for Finnish - for instance, it's used on this website of the official body governing Finnish. It's not used as a letter per se but as a typographical symbol; regardless the words should not link separately since they have no meaning. Besides, I'll start looking for all the entry titles using U+2019 soon and if all of them use it as an apostrophe, I'll likely just add it to notWordPunc. SURJECTION ·talk·contr·log· 15:36, 14 November 2018 (UTC)
This claim “they have no meaning” is of course telltale. Apparently they are used to convey something, innit? I don’t think either essentialist considerations if something “is a letter” have a place here. The apostrophe is “not a letter” in Ukrainian and Belarusian either, at the same time its function is comparable to the Russian ъ, to say nothing about historical usage of this for Ukrainian. Is ъ a letter? 🤷🏼‍♂️. Usage on the internet is of course a bad guide in Unicode matters, it always depends on what is at hand technically. I can show many sites using " " for German quotation marks but this gets corrected to „ “, you get your exams red if you use it in school. Fay Freak (talk) 15:44, 14 November 2018 (UTC)
"they have no meaning" is a claim made towards the two components U+2019 ends up separating. liu’uttaa means something; liu and uttaa meanwhile really don't. As to "Usage on the internet is of course a bad guide in Unicode matters", indeed it is since most people just use the ASCII apostrophe, which is incorrect; the apostrophe as used in Finnish looks like U+2019 and is curved. It is also worth noting that Unicode has preferred U+2019 over U+02BC when used as an apostrophe, as per this (look for RIGHT SINGLE QUOTATION MARK). SURJECTION ·talk·contr·log· 19:02, 14 November 2018 (UTC)

Make insertTags work for a bookmarkletEdit

The following bookmarklet inserting certain text in current cursor position used to work:

javascript:insertTags('===Further%20reading===\n*%20{{R:PSJC}}\n*%20{{R:SSJC}}','','');document.editform.wpSave.focus()

I tried to replace insertTags with mw.toolbar.insertTags, but I get "TypeError: mw.toolbar is undefined"; I have the editing toolbar disabled in preferences but I get that error even when I enable the editing toolbar.

Does anyone know how to change it to make it still work? Thank you. --Dan Polansky (talk) 13:00, 10 November 2018 (UTC)

futsal#Etymology_2Edit

How can this be dones without the + [term?] showing up. Should Template:blend of perhaps be changed so that an idiom can be ”blended”, which is the case here.Jonteemil (talk) 16:40, 10 November 2018 (UTC)

@Jonteemil: I tried to fix it. Are you happy with the result? --Dan Polansky (talk) 17:15, 10 November 2018 (UTC)

Changes to Abuse FiltersEdit

If anyone is wondering why I just edited most of the Abuse Filters: apparently they've come up with new names for a lot of the built-in variables and deprecated the old ones, so I replaced them in all of the filters that had them. FWIW, I like the new ones much better: I always thought it was confusing to refer to a page's name as "article_text".

I also rearranged a few of them to move high-overhead things like searching wikitext to the end. It's easy to forget that every enabled abuse filter is executed for every edit and they all keep going until a failed condition makes them stop. The idea is to put the easy stuff like namespace and user-group checks at the beginning so the slow stuff gets executed on as few edits as possible. I have no idea how much time is actually used by the abuse filters, but there's no reason to waste any of it over something as simple as this. Thanks! Chuck Entz (talk) 02:21, 12 November 2018 (UTC)

Thanks for this, Chuck. - TheDaveRoss 14:32, 12 November 2018 (UTC)
I have created a handful of these. It strikes me that the UI is not very good (especially having to click the big arrow to see the newest filters, and only a few per page): ideally perhaps we would want a hierarchical view so that we can easily find anti-spam vs. anti-vandalism and so on. Equinox 17:31, 12 November 2018 (UTC)
I really wish we had a way to write tests for these. DTLHS (talk) 17:31, 12 November 2018 (UTC)

Inflexible inflection tablesEdit

Some inflection tables, so the ones for German, as on Schnalle, needs take the whole width of the article body, in such a fashion that images can only be above or below the table and not run alongside it. On the same page the picture aligns smoothly and the inflection table recoils if only the inflection table is the Arabic one, that is: use {{ar-decl-noun}} instead of {{de-decl-noun-f}} and the layout is as it should be. Since the former leaves ugly gaps between the inflection tables and their headings it needs some alignment change. My stylesheet practice is a bit too long ago to find out what’s going on here. Fay Freak (talk) 23:22, 12 November 2018 (UTC)

Try removing "style="width:100%"" from Template:de-decl-noun-table-full. DTLHS (talk) 23:29, 12 November 2018 (UTC)
Yes, that has fixed it. I thought so complicated. Fay Freak (talk) 23:39, 12 November 2018 (UTC)
A better remedy in the long term is to use the vsSwitcher type of collapsible table, like the Finnish and Sami tables. There is no longer a need for a wrapper div in this type of collapsible table, it works with just a plain table and you have the option to collapse each individual table row as well. —Rua (mew) 20:21, 14 November 2018 (UTC)

Cleanup suggestions for some badly attested Semitic languages, needing admin actionEdit

Discussion moved to Wiktionary:Requests for moves, mergers and splits#Cleanup suggestions for some badly attested Semitic languages, needing admin action.

Help Adding Translit. Parameters to Syriac TemplatesEdit

Hello there!

I need to add transliteration parameters for plurals in all the noun and adjective templates in Category:Classical Syriac headword-line templates. Please look to Arabic templates as a guide (e.g. بيضة, note the various plurals and their transliterations), though Syriac transliterations can't be automatically generated. Something like pltr=, pltr2=, pltr3=, etc. would be great. Any help would be much appreciated. :) --334a (talk) 22:04, 14 November 2018 (UTC)

I have added them, these parameters work now. But note that the way {{head}} works |ftr= or |mtr= cannot be added because like plurals masculine and feminine counterparts use positional parameters in the backbone template. You can only fill the transcriptions for them with the same parameters, i. e. in this case by abusing |pltr2= if there is one plural and one gender counterpart given.
Sooner or later not only the Aramaic headword templates must be governed by modules but also {{head}} must be amended to allow explicit (non-positional) plural and gender statements, literally {{head}} should have |plN= and |pltrN=, since, remember, lexicalized plurals do not only exist in Semitic languages, even in Germanic, for example German, one must learn plurals separately. And |f= and |m= should be added because there are only two genders, in what matters for language, so regularly many languages have nouns paired by masculine and feminine counterparts. Or maybe only a wrapper or alternative general headword-line template {{head-noun}} or {{head-nominal}} should have such parameters, for simplicity, if you say that’s noun-specific and hence does not belong into {{head}} (I don’t know). Maybe then the collective-singulative stuff (when there is not singular and plural but collective and singulative, like Arabic نَخْل (naḵl, palm, palms)) could be handled with such a template too. @334a, Benwing2. Fay Freak (talk) 23:21, 14 November 2018 (UTC)
Thank you, @Fay Freak, that's exactly what I needed! I'm not too tech-savvy to understand everything you said, though; I tried to add a transliteration for the feminine counterpart in ܝܠܘܦܐ using |pltr2= and then |ftr=, both to no avail. --334a (talk) 05:31, 16 November 2018 (UTC)
Indeed, this does not work, and I now do not know why. @334a I suggest you use |mtr= and |ftr= anyway even though they do not display yet – later if someone makes a module this information will be useful. Fay Freak (talk) 12:03, 16 November 2018 (UTC)
Will do. Thanks again. :) --334a (talk) 17:35, 16 November 2018 (UTC)

No wikidata interwiki on Reconstruction:Proto-Indo-European/gʷḗnEdit

Hello, Reconstruction:Proto-Indo-European/gʷḗn is linked to other Wiktionary pages via Wikidata. However, no interwiki links are displayed while it works on the French Wiktionary. Do you have any idea why it is not working here? Pamputt (talk) 06:49, 15 November 2018 (UTC)

German verb conjugation with spacesEdit

In German, there are a few verbs that always have a space between it and the "seperable prefix" (Rad fahren, Schluss machen, etc.). The current solution for making a conjugation table for them seems to be placing a &#32; after the "seperable prefix" in Template:de-conj-strong/Template:de-conj-weak. This gives the proper form in all cases except the zu-infinitive, where you get *Rad zufahren instead of Rad zu fahren. How would one fix this? Mofvanes (talk) 00:28, 16 November 2018 (UTC)

Problem with Template:zh-x in quotation on 隔音符號 pageEdit

The example sentence I added to 隔音符號 today is basically the origin point of the term '隔音符號' (syllable-diving apostrophe) as it is used in Mandarin Chinese's Hanyu Pinyin scheme; it is a sentence that is important enough that it may eventually need to be moved from the quotation section to the etymology section for this crucial concept of Mandarin Hanyu Pinyin. This is not some nonsense, see-me-now-forget-me-later sentence: it has been printed in every dictionary and every book related to pinyin in Mainland China since 1960. The only problem for me is, if you look at the pinyin given for in the Mandarin Hanyu Pinyin pronunciation guide for this sentence (note: the first line in the quotation is given in traditional characters, the second line is in simplified characters, the third line is a citation of the source of the sentence, the fourth line is a Mandarin Hanyu Pinyin pronunciation guide and the fifth line is my English translation), you will see that it the pinyin starts with the capital letter 'A'. I believe that capitalizing this letter is a grave error; this sentence may be one of the very few sentences used in a zh-x context that should definitely not start off with a capital letter. But I searched the Template:zh-x page and couldn't figure out how to turn off the first-letter-capitalization function for the Mandarin Hanyu Pinyin pronunciation guide. If you can't understand me, please ask me some questions and I will try to answer them and make clear my problem. To lay it out more concretely, what we see on the page right now is this:

A, o, e kāitóu de yīnjié liánjiē zài qítā yīnjié hòumiàn de shíhou, rúguǒ yīnjié de jièxiàn fāshēng hùnxiáo, yòng géyīn fúhào (’) gékāi, lìrú: pi’ ao (pí'ǎo). [Pinyin]

But what I want to see is this:

a, o, e kāitóu de yīnjié liánjiē zài qítā yīnjié hòumiàn de shíhou, rúguǒ yīnjié de jièxiàn fāshēng hùnxiáo, yòng géyīn fúhào (’) gékāi, lìrú: pi’ ao (pí'ǎo). [Pinyin]

Can you help me?

--Geographyinitiative (talk) 01:53, 16 November 2018 (UTC)