User talk:Erutuon

Archives:
2009 – 2010 – 2011 – 2015 – 2016 – 2017 – 2018 – 2019 – 2020

Quarrying

Also, it would be fun to see the league table for the most one-sided thank relationship (an unrequited-love list, if you will), like this where we can all see that I'm so meta even this acronym (talk • contribs) is totally stalking JohnC5 (talk • contribs) (951 thanks), but with a part where JohnC5's number of thanks to ISMETA (341) is deducted from that total. --Elvinrust (talk) 22:49, 4 May 2020 (UTC)

Which makes me think of another fun list - most affectionate thank-couples (combined thank-totals...John as ISMETA will win that, hands down) --Elvinrust (talk) 22:51, 4 May 2020 (UTC)

Ooh, and a list of the most aggressive relationships, coz I'm gonna make a documentary about it called Users who Revert Other Users. --Elvinrust (talk) 22:53, 4 May 2020 (UTC)

Header

For some hiragana entries I can’t remove the header, therefore you see some weird empty header which is my attempt of removing it. Thanks for the correction! Shen233 (talk) 03:52, 7 May 2020 (UTC)

You may not be very familiar with Japanese wiktionary, but for most non-lemma hiragana entries, no header is needed for the current "｛｛ja-see｝｝" redirect mechanism. There maybe a noun and a verb which share the same hiragana, then we separate them by etymologies, such as 五日 (いつか) a noun and いつか, an adverb. In older practices they put PoS and then "｛｛ja-def ｝｝" though.　Shen233 (talk) 04:03, 7 May 2020 (UTC)

Thanks for the explanation. I did a little editing of Japanese, mostly updating {{ja-readings}} and fixing ruby, in the past, before {{ja-see}} was renovated. Here I was just removing the empty header === ===: it isn't officially allowed in WT:EL and shows up in my cleanup list, User:Erutuon/mainspace headers/possibly incorrect. — Eru·tuon 17:48, 7 May 2020 (UTC)

Module:es-pronunc

Hi, Erutuon. Could you change a part of Module:es-pronunc? In Template:es-IPA ll is shown as a consonant with different pronounciation in Castile and Latin America but that is not accurate. yeísmo and lleísmo exist in both regions, yeísmo is preferred in both too. Words like llamar should say "(yeísmo) IPA(key): /ʎaˈmaɾ/ (lleísmo) IPA(key): /ɟ͡ʝaˈmaɾ/. Thanks in advance. 181.226.219.122 20:04, 16 May 2020 (UTC)

I agree that this should be changed, but it's more complicated because seseo is also involved. So with this change, there might have to be four pronunciations, distinción and lleísmo, distinción and yeísmo, seseo and lleísmo, seseo and yeísmo. I don't know if all of these exist. It would be better to discuss this at Module talk:es-pronunc. — Eru·tuon 16:26, 21 May 2020 (UTC)

Thank you. Yes, all of them exist, being distinción + lleísmo the less common way. Someone posted a message in 2018 and it remains unanswered. I'll repeat my request there. Regards. Lin linao (talk) 18:57, 7 June 2020 (UTC)

Missing Spanish idioms

Hey. Can you make me a list of all the entries in this Spanish cat that are not in en.wikt? Let's put it at Wiktionary:Todo/Missing Spanish verb idioms --Spanishlearner574 (talk) 21:50, 23 May 2020 (UTC)

@Spanishlearner574: Okay, made a Quarry query and pasted the results there. — Eru·tuon 22:57, 23 May 2020 (UTC)

Sweet. There's more entries there than I was expecting. --Spanishlearner574 (talk) 23:16, 23 May 2020 (UTC)

Could you make that list even better by including links to es.wikt, like below - I started doing it manually offline but found no quick way to make the changes. --Undurbjáni (talk) 10:35, 28 May 2020 (UTC)

@Undurbjáni: Ah, yeah, makes sense. Added it. You can change the look of it by editing the same part that I edited. — Eru·tuon 15:54, 28 May 2020 (UTC)

1. abrir cancha (es:abrir cancha)
2. abrir el tarro (es:abrir el tarro)

Category:Han script

This has had a mostly unnoticed module error for quite some time. As far as I can tell, it's a disagreement between Module:category tree/script cat/blocks and Module:Unicode data/blocks about where the end of a block is, and it seems to have been triggered by this edit. Could you take a look at it? It's definitely not high priority, but it's mildly annoying... Chuck Entz (talk) 01:31, 24 May 2020 (UTC)

@Chuck Entz: Thanks! Fixed. That was because Module:scripts/data was assigning a range of unassigned code points ending in U+2FFFF to Hani, and that one doesn't have a block assigned to it. It would be handy to include the whole Supplementary Ideographic Plane (U+20000-U+2FFFF) because I guess it will only ever include Han characters, but Module:category tree/script cat/blocks requires the first and last code points of the ranges to be assigned. — Eru·tuon 01:58, 24 May 2020 (UTC)

fixing excessive width of Hungarian-language number boxes

(Antecedents.) Would you please change the display of "Adverbial ordinal:" to A.o. (preferably with this tool tip) in Module:number list? See e.g. tizenkilenc. Currently this seems to be the only way to avoid its double entries excessively widening the table, without any side-effects. Thank you in advance.

Another way I could imagine is inserting a string length check possibly before table.concat(form, ", ")), so that a line break should be inserted instead of a space after the comma if a given value is longer than e.g. 15 characters. However, it would affect lots of other tables as well, so I understand if you'd rather avoid it, although it might be some improvement nevertheless.

I've deleted one out of the three values given at Adverbial ordinals from Module:number list/data/hu, because the formatting of its values simply didn't allow the number box to be inserted into másodszor; it produced an error message. A similar solution could be considered for the distributive, possibly abbreviating it to "Dist.:", because we're bound to have the same problem, see e.g. száz. (@Panda10, do you have any suggestion?) Adam78 (talk) 22:55, 2 June 2020 (UTC)

Sorry, I don't have any new suggestions. Panda10 (talk) 16:24, 3 June 2020 (UTC)

Unfortunately string length is complicated. There is a function to count the number of code points, but it doesn't correspond to the number of visible characters when you have, for instance, combining accents (a + ́ = á). This doesn't come up with many European languages, but would with various Indian and Southeast Asian scripts. The number of Unicode graphemes is better (á is a single grapheme) but we don't a grapheme-counting function here on Wiktionary; I'd have to find or write one. Graphemes aren't exactly proportional to font length, but they are closer.

I think these tweaks are not the final answer to the problem of the number box layout. Some sort of redesigning would be better, but I have no good ideas at the moment and I'm just discouraged about the whole thing. I might implement your suggestions as a temporary measure. — Eru·tuon 00:02, 4 June 2020 (UTC)

All right, no problem. In this case, forget about string length. All I'd like to ask you is modify these two names in the list:

• A.o.:
• Dist.:

I was also thinking about the current term "number of people", which is named "adverb of number" in a grammar book, but it sounds too unspecific to me (many of these terms are some kinds of adverb of number anyway), so it's better kept as it is, unless you suggest otherwise. Thanks a lot in advance. Adam78 (talk) 18:35, 4 June 2020 (UTC)

@Adam78: I changed the display of "adverbial ordinal" but am not sure what to do about distributive because it's also used by other languages. Maybe there needs to be a way to customize the label for each number type for each language. — Eru·tuon 18:32, 5 June 2020 (UTC)

OK, thank you. We're still one step ahead. Now the width looks considerably better, if not the best. Adam78 (talk) 23:18, 5 June 2020 (UTC)

Arabic new entry templates

Hello, I created these templates: Template:ar-nogomatch

• Template:ar-new-verb
• Template:ar-new-noun
• Template:ar-new-adj
• Template:ar-new-adv

and then I realized that it should be added to MediaWiki:Searchmenu-new. Can you please add it there to language picker the dropdown? LinguisticMystic (talk) 09:45, 22 June 2020 (UTC)

@LinguisticMystic: I've added the buttons from Template:ar-nogomatch, but I modified the style to match the rest of the languages in MediaWiki:Searchmenu-new. — Eru·tuon 18:18, 22 June 2020 (UTC)

Thanks! Actually it's still not working, when I try to create a new entry, only English, American Sign Language, Spanish, Swedish pop up in the dropdown menu, and they dont seem to be working either. When I select Swedish, for example nothing happens. LinguisticMystic (talk) 18:28, 22 June 2020 (UTC)

I'm wondering what could be the problem. LinguisticMystic (talk) 18:40, 22 June 2020 (UTC)

@LinguisticMystic: MediaWiki:Searchmenu-new doesn't have a dropdown menu, at least with my settings. You must be seeing a different MediaWiki message. — Eru·tuon 18:51, 22 June 2020 (UTC)

Below the table, below VERB, it says English, Select a different language. If you click English, the others appear, except for Arabic. LinguisticMystic (talk) 18:55, 22 June 2020 (UTC)

@LinguisticMystic: Ah, it looks like I'd disabled MediaWiki:Gadget-SpecialSearch.js, which generates the dropdown menu in MediaWiki:Searchmenu-new. I might be able to figure out why the gadget isn't picking up Arabic. — Eru·tuon 19:07, 22 June 2020 (UTC)

Unfortunately, the other options don't work for me either, only the default English option, so please check if the code is okay. LinguisticMystic (talk) 19:11, 22 June 2020 (UTC)

Great. It is working indeed. Thanks a lot. You made my work much easier and faster. LinguisticMystic (talk) 20:10, 22 June 2020 (UTC)

Hello @LinguisticMystic and @Erutuon, I just got interested in the new entry templates. I've run into the exact same problems described by you above though (logged in and logged out):

1. "when I try to create a new entry, only English, American Sign Language, Spanish, Swedish pop up in the dropdown menu"
2. "and they dont seem to be working either. When I select Swedish, for example nothing happens"

It seems that it had worked in the meantime. Do you have an idea what has broken since then? Thanks a lot in advance, --Marsupium (talk) 10:55, 18 September 2022 (UTC)

Module errors due to removing items from Module:unsupported titles/data

{{unsupported|://}} at Wiktionary:Beer_parlour/2016/October#Possible future vote about deleting all programming language_symbols, {{unsupported|ideographic space}} at ideographic space, and a whole Finnish-declension-table full of unsupported inflected forms at Unsupported titles/n:s. Chuck Entz (talk) 04:41, 25 June 2020 (UTC)

@Chuck Entz: Ouch. Reverted. I noticed the one (þ), but didn't go looking for more. Later I might try to figure out which titles were added in the edit and restore them, or maybe User:J3133 will be kind enough to do it. — Eru·tuon 04:47, 25 June 2020 (UTC)

lots of rfdateks

Hi. It's been a while since I've bugged you for a random list. Would you be able to cook up a list of the entries which contain the most occurrences of {{rfdatek}} and {{rfdate}}? My bet is the "winner" will have around 25. --Nueva normalidad (talk) 07:42, 1 July 2020 (UTC)

@Nueva normalidad: Shoot, forgot about this. But lists are fun. I made User:Erutuon/lists/rfdate and rfdatek leaderboard but unfortunately you and others have been so diligent that the most winning one has only half that. — Eru·tuon 07:47, 6 July 2020 (UTC)

Nueva’s been blocked, but I suppose he’ll be back in another incarnation before long. — SGconlaw (talk) 12:29, 6 July 2020 (UTC)

Oh right, I forgot that too. — Eru·tuon 19:03, 6 July 2020 (UTC)

Yeah, (s)he is probably gonna be back soon, but will probably be going anonymous. I, OTOH, as a completely new user, happen to be interested in this too, and give thanks for the creation of the list. --Dada por viva (talk) 00:00, 7 July 2020 (UTC)

Curly quotes in CFI

I would like to ask you to undo adding curly quotes to CFI since they are in fact quite controversial, as per vote that I will try to find; "Any substantial or contested changes require a VOTE." The vote is Wiktionary:Votes/pl-2008-12/curly quotes in WT:ELE; it pertains to EL specifically, but it would be the same story for CFI. This should only be done via a vote. --Dan Polansky (talk) 09:24, 11 July 2020 (UTC)

@Dan Polansky: The edit was undertaken by request of User:J3133 because WT:STYLE#Quotations seems to sort of favor curly quotes, in entries in particular. I don't consider this a substantial change, but it's contested since you at least contest it. I see there are some slightly newer polls about this too, which didn't show strong favor for curly quotes (Wiktionary:Votes/pl-2013-02/Disallow typographic punctuation in policies, Wiktionary:Votes/2013-02/Typographic vs ASCII punctuation in policies). I favor curly quotes myself and don't see why WT:CFI should be different from entries, which already use curly quotes to a great extent because common modules or templates like Module:links and {{rfc}} use them, but I'll revert since I'm not interested in drawing controversy. — Eru·tuon 17:31, 11 July 2020 (UTC)

Thank you. I reverted recent changes made by User:J3133 in Wiktionary:Style guide in which they introduced curly braces without any trace to a discussion or a vote. --Dan Polansky (talk) 18:31, 11 July 2020 (UTC)

@Dan Polansky: Now you're pissing me off with your edit warring over quotes in User:AryamanA/Wonderfool. Holding extended discussions in edit summaries in an edit war is an idiotic practice that I refuse to engage in. Since when are user pages prohibited from having curly quotes unless they are prescribed for policy pages by a vote? If that were a policy, then you should go around annoying more people by changing their curly quotes and apostrophes to straight ones. — Eru·tuon 18:54, 11 July 2020 (UTC)

I don't like to piss you off. Why does your preference in User:AryamanA/Wonderfool prevail over my preference? Why did you revert me the first time knowing there is no consensus for curly braces, not even 55% majority? The status quo ante in that page is no curly braces; what makes the status quo ante not prevail? --Dan Polansky (talk) 18:57, 11 July 2020 (UTC)

@Dan Polansky: I suppose it's reasonable to assert that it's a status quo, and that User:AryamanA's Wonderfool page is more like a Wiktionary-namespace page than a user page at this point. So, yes I regret reverting you now. However, I wish that in reverting you would restore the indisputably legitimate changes that do not involve this apparently controversial typographical issue. — Eru·tuon 19:02, 11 July 2020 (UTC)

An issue that User:J3133 raised in a private message with me is that Wiktionary:Style guide already has some curly quotes. His edit was an improvement in consistency. — Eru·tuon 19:06, 11 July 2020 (UTC)

I wanted to only revert the quotes; I apologize for reverting more than that, a mistake. J3133 seems to be Wonderfool, who likes to stir controversy. Wiktionary:Style guide was edited by J3133. To me, all private messages are a red flag, but maybe I am a hyper-transparentist. --Dan Polansky (talk) 19:17, 11 July 2020 (UTC)

@Dan Polansky: From my personal experience I can assure you that J3133 is not Wonderfool, nor is there any significant resemblance between them once you actually get to know them. In publicly visible behavior there's the fact that J3133 doesn't engage in the same goofy stunts as Wonderfool, and hence has been granted autopatroller status. I do prefer that discussions that require input from others or consensus be on-wiki so that they're publicly accessible, but find Discord useful for dealing with relatively uncontroversial things. I thought that curly quotes would be an uncontroversial issue because WT:STYLE says they're preferred and they are used in many common templates, but apparently at least you disagree, so there should be some on-wiki discussion or yet another tedious poll. — Eru·tuon 19:33, 11 July 2020 (UTC)

Thank you. Wiktionary:Style guide is not a policy. It would be better do delete these kinds of non-vote-controlled quasi-policy pages where non-consensual material accumulates, or at least change them to "kept for historical interest". --Dan Polansky (talk) 19:39, 11 July 2020 (UTC)

Module Errors from Whitespace in Language codes

I've seen two module errors in the past 24 hours where Module:redlink category called Module:links.getLinkPage for a link and the template in the entry had "|da |" as the language code. The only recent change I could find was your edits to Module:links. They were easy to fix, so they probably weren't around long enough for you to see them. Here and here are the diffs from before I fixed them. Please take a look. Thanks! Chuck Entz (talk) 06:54, 12 July 2020 (UTC)

@Chuck Entz: Huh. Odd little bug. This fixes it, because named parameters have whitespace stripped from them. — Eru·tuon 07:12, 12 July 2020 (UTC)

perfection

Hi. I can't edit the page perfection. Please can you add the quote to the entry? It's from Philip Sidney's Arcadia - at Talk:perfection --Kriss Barnes (talk) 18:13, 23 July 2020 (UTC)

@Kriss Barnes: There you go! — Eru·tuon 18:23, 23 July 2020 (UTC)

Cheers. Could you do the Bible quote at cool too? #* '''1611''', Bible ({{w|King James Version}}) --Kriss Barnes (talk) 18:27, 23 July 2020 (UTC)

@Kriss Barnes: Done as well! — Eru·tuon 18:36, 23 July 2020 (UTC)

And the quote by William Woty at [[perfection]] is from 1770, the poem A Mock Invocation to Genius --Kriss Barnes (talk) 20:49, 23 July 2020 (UTC)

@Kriss Barnes: Filled out the citation, hopefully correctly. — Eru·tuon 19:48, 24 July 2020 (UTC)

Joseph Hall's quotes

Thanks again for tracking down the previous request. The next mini-project was the 17th century bishop Joseph Hall's quotes - I dated all of them here except for 3 - at occecation, cramming and waning, which I couldn't figure out - maybe you can fill them in for me. Anyway, I declare the Joseph Hall dating project complete, and will move on to something else. --Kriss Barnes (talk) 09:56, 28 July 2020 (UTC)

Ugh, I did a brief search but I'm not clever enough to find the Bishop's occecations and crammings and wanings. — Eru·tuon 08:16, 31 July 2020 (UTC)

"no duplicate parameters as of the latest dump, so safe to throw error here"

See CAT:E: 9 templates, 18 module errors- all of them with their html showing. It's not something you see everyday... Chuck Entz (talk) 07:26, 30 July 2020 (UTC)

@Chuck Entz: Ouch. Thanks for the ping, I've reverted my misinformed edit. (I didn't think and looked for |tr1= instead of |tr= when parameter 1 was present, so I missed all the {{ar-IPA}} with both Arabic and transliteration.) Apparently strange stuff happens when a module error includes an erroring template. — Eru·tuon 09:08, 30 July 2020 (UTC)

Leaderboard

Hey. Can you regenerate User:Erutuon/lists/rfdate and rfdatek leaderboard after (I guess) the next dump? Just to see how little progress I have actually made and make me super-demotivated. --Kriss Barnes (talk) 08:50, 31 July 2020 (UTC)

@Kriss Barnes: There you go! — Eru·tuon 00:09, 8 August 2020 (UTC)

Awesome. I think I have sliced off about 15% of that list, which is just small enough to still count as a mere drop in the ocean and, indeed, to de-motivate me. --Kriss Barnes (talk) 00:16, 8 August 2020 (UTC)

Did it really take over an hour to make the list? If so, double-thanks. --Kriss Barnes (talk) 00:17, 8 August 2020 (UTC)

@Kriss Barnes: Oh, no, it takes about 80 seconds to run the script over the 6-GiB dump. But then I go away from the computer to wait for it and come back later to post it, and then come up with a change and rerun it, and so on, so the whole process is drawn out. — Eru·tuon 00:26, 8 August 2020 (UTC)

Ah, OK, so it takes over an hour to make the list... It takes about that time for me to date 50 quotes (with an estimated accuracy rating of 96%). Sometimes I wonder why I spend soooo much time on this task. --Kriss Barnes (talk) 00:31, 8 August 2020 (UTC)

Scripts in Module:languages

We now have script data for about two-thirds of the languages we have codes for (and with effort, could bring that higher). In the past, "canonical name", "Wikidata item" and "family" were made positional rather than explicitly named parameters/fields, which I seem to recall was to make the modules more efficient in their use of memory. Would it make sense to also make scripts "positional", or would handling cases where languages have multiple scripts make this not worth the bother? - -sche (discuss) 22:06, 1 August 2020 (UTC)

@-sche: Yeah, it would be a good idea now. Just have to track down all the modules that access the "scripts" field in the language data modules, change them to check for field 4 as well, move all the "scripts" fields to field 4, and finally change all the modules to access field 4. — Eru·tuon 19:13, 2 August 2020 (UTC)

Template:cite-meta spacing problems

As exhibited by Template:R:sga:GOI, cite-meta is refusing to put a space between the translators and the next citation element in line. I wonder if other parameters suffer the same problem. mellohi! (僕の乖離) 18:41, 8 August 2020 (UTC)

@Mellohi!: Well, I'm not familiar with {{cite-meta}}, but I made this edit, which fixes the problem in {{R:sga:GOI}}, though it might cause other problems elsewhere and need to be reverted. — Eru·tuon 19:08, 8 August 2020 (UTC)

Do you know anyone else who is familiar with it? mellohi! (僕の乖離) 19:11, 8 August 2020 (UTC)

@Mellohi!: It looks like User:Sgconlaw has edited it the most; Sgconlaw, could you look over my edit? It unconditionally adds a space after the parenthesis that contains the year or date, but I don't know if that's going to always be correct. — Eru·tuon 19:20, 8 August 2020 (UTC)

Actually, I was thinking of revising this meta-template so that it no longer displays the date or year of publication between the author's name and the title. I think this might be simpler and eliminate the problem you describe. However, it's not going to happen immediately as I'll need to find some time to work on it. — SGconlaw (talk) 19:35, 8 August 2020 (UTC)

Old Irish empty inflection tables

I remember you brought up these empty declension tables off-wiki, where there are a bunch of Old Irish entries using {{sga-decl-noun}} and stating supposed stem classes but with no forms inputted in. I would like to ask, is it possible to configure the template or its backend module so that if no parameters other than stem class are filled, what happened here, the page ends up in a maintenance category like "Old Irish nouns with empty inflection tables". Could Module:sga-verbs also be configured the same way to catch verb conjugation tables with no forms into "Old Irish verbs with empty inflection tables"? mellohi! (僕の乖離) 20:23, 8 August 2020 (UTC)

@Mellohi!: I've made Module:sga-nouns add Category:Old Irish nouns with empty inflection tables to pages where {{sga-decl-noun}} has no parameter besides |class=, if that's what you mean. I don't know under what condition Module:sga-verbs displays absolutely no forms in the table though. — Eru·tuon 21:59, 8 August 2020 (UTC)

Lists

I hear you're the person to go to for lists, so if you're down I have some requests for lists to connect entries by etymology. These are the 3 types that I want to tackle, with one each to start:

1. Latin entries that list Italian descendants that aren't linked to from the Italian descendant: User:Ultimateria/oneway/la-it
2. Category:Armenian terms borrowed from Russian that aren't linked to from the Russian etymon: User:Ultimateria/oneway/hy-ru
3. Category:Catalan words suffixed with -ment that aren't linked to from |2= of the suffix/affix template: User:Ultimateria/oneway/-ment ca

I'm slowly running out of tasks like this that I can find with regex searches, so I'd appreciate your help. Ultimateria (talk) 01:31, 14 August 2020 (UTC)

I've created the la-it page with an example to show you what I'm picturing. Ultimateria (talk) 16:29, 14 August 2020 (UTC)

Hmm, it's an interesting task. It's somewhat more complex than some of the other lists I've done, because it requires looking at two sets of pages in the dump (the first set of pages, and the pages that should be linking to them), but I'll give it a try today. — Eru·tuon 19:19, 15 August 2020 (UTC)

Okay, created a first version of the la-it list, but it isn't completely accurate. I'll post a new version when I fix it. — Eru·tuon 03:14, 16 August 2020 (UTC)

Fixed. Note that I only checked for links in the Descendants section of the Latin entry and an Etymology section of the Italian entry, and in particular templates ({{l}}, {{m}}, {{desc}}, {{inh}}, {{der}}, {{bor}}). If you think of any additional places or templates, let me know. — Eru·tuon 08:39, 16 August 2020 (UTC)

Perfect, thank you! :D Ultimateria (talk) 16:15, 16 August 2020 (UTC)

Finished the other two as well. — Eru·tuon 01:28, 17 August 2020 (UTC)

Excellent, thanks so much! Ultimateria (talk) 16:38, 17 August 2020 (UTC)

Technical Request: Mass "touch" request to purge caches of transclusions where page was updated.

Many of the entries in : https://en.wiktionary.org/w/index.php?title=Special:LintErrors/html5-misnesting&offset=11120075&namespace=0

where fixed by updates to an uderlying template.

However, in order for those changes to propogate and clear the lint-errors, each page on the list, appears to need to be purged or a null edit conducted to 'force' the updated version to be used.

Doing this manually for > 10,000 pages isn't feasible, hence a bot or script would be better. ShakespeareFan00 (talk) 09:02, 14 August 2020 (UTC)

Done! Now it's down to half or so of the original ~50,000. — Eru·tuon 05:18, 15 August 2020 (UTC)

Doing another run should clear-out a few more - https://en.wiktionary.org/wiki/Special:LintErrors/html5-misnesting?namespace=0 :) ShakespeareFan00 (talk) 23:09, 15 August 2020 (UTC)

Ran touch.py again, and that lint error has 8,516 pages now. (It took 48,430 seconds at about 1 edit per second, whew.) — Eru·tuon 23:28, 16 August 2020 (UTC)

Deletion of Category:grc:Greek letter names in Ancient Greek

Hi Erutuon, I replied on my talk page. The thing missing with the deletion is the link to the page on the French wiktionary.--OpenNotes1 (talk) 21:42, 20 August 2020 (UTC)

Advice needed

Hello @Erutuon, hello teacher! Hope you are having a nice summer! A piece of advice, please. I did mycenaeanLinb.to.traslit here -in my clumsy way-, but I do not know how to think, which Module I could start studying, to manage the opposite: translit.to.mycLinb. I have seen your typing aids but I do not know what to do... I have tried many silly ways to split the syllables (1letter, 2letters, 3ltters with "-" as boundary). Could you give me some directions? Where to look at? This and that module...
P.S. Question2: Also, I have made a page with all data. Can this be used for specific conversions? Or, I need to do one pattern for each kind of conversion, like this one? Thanks ‑‑Sarri.greek ^♫ | 00:04, 21 August 2020 (UTC)

@Sarri.greek: Hello again! I looked at your modules. At first I was thinking it would be possible to write a Lua pattern (regex) for the syllables, or several patterns that would be searched in order, but actually there would be some ambiguities (rai versus ra-i for instance). So your method of separating the codes for the Linear B characters is the easiest to implement. latin_to_symbol looks the closest to the method I would use. But you can iterate over the hyphen-separated bits with mw.text.gsplit(latin, "%-") (instead of string.gmatch(text, "."), which iterates over bytes) and just return table.concat(ret) at the end of the function because the loop will not look at the hyphens any more. This requires hyphens to be removed from el:Module:my/translitToUni.

As for the el:Module:Linb-translit/data, I'd keep it as a master data module, but you can have other data modules that require it, loop through it, and generate other formats, such as a code-to-character table for your latin_to_symbol to use, or a character-to-image table. With my suggested changes to latin_to_symbol above, el:Module:my/translitToUni would need to be something like:

local code_to_char = {}
for _, v in pairs(require("Module:Linb-translit/data")) do
	local latin, code, uni = v.latin, v.code, v.uni
	code_to_char[code] = uni
	code_to_char[latin:gsub("[%*%?]", "")] = uni
end
return code_to_char

So then all the data modules would be built based on the master data module, and you would only have to edit in one place.

I don't know if this is quite enough information; if not let me know and I can try explaining further or edit the modules myself. — Eru·tuon 18:48, 21 August 2020 (UTC)

ΤΗΑΝΚ ΥΟΥ Erutuon. Very much. I will try ‑‑Sarri.greek ^♫ | 19:47, 21 August 2020 (UTC)

@Erutuon! @Erutuon! Everything works like a wonder. And I do not need all the subpages now, with your «_, v in pairs». Everything comes from the data page, just as you wrote. And your blue links in the modules with the common.js! One more gift. You have no idea how precious your help is. Thanks ‑‑Sarri.greek ^♫ | 00:48, 22 August 2020 (UTC)

@Sarri.greek: Glad I could help! About the CodeLinks gadget, if you haven't, you can also enable it in your Preferences here as well. — Eru·tuon 16:57, 23 August 2020 (UTC)

extracting data from dumps

What technology would you suggest for extracting data from dumps? (For example, Chinese pronunciations of words.) —Suzukaze-c (talk) 02:57, 21 August 2020 (UTC)

@Suzukaze-c: Well, I mostly know of the tools that I wrote myself. It depends what data exactly. I have a program with a template-dumping subcommand, and another program in the same repository for running Lua scripts (yeah, I even use Lua off-wiki) over various aspects of pages. They're both badly documented because nobody's asked me to document them so far, but I use both of them regularly. For Chinese pronunciations specifically there's a dump of {{zh-pron}} on Toolforge. It's in CBOR format, but I could also dump in JSONL format, which is a bit less niche. So you could take that and write a script in your favorite language to iterate over it, and, say, get parameter |c= on page x. If any of this sounds useful, I can give more information.

(I feel like another likely question would be "how do I run modules off-wiki?", but I haven't really worked that out, beyond making Module:languages and a few other modules work. It would be useful for my IPA searcher because I could include more pronunciation templates, but it feels like such a complicated task that I haven't even tried yet.) — Eru·tuon 19:31, 21 August 2020 (UTC)

:D

Thinking about it, I think the {{zh-pron}} dump would be good enough for what I'm thinking of. (I was thinking about doing extra work to combine pronunciations with all the forms from {{zh-forms}}, but maybe it's not really necessary, possibly even undesirable.) —Suzukaze-c (talk) 08:05, 22 August 2020 (UTC)

Is the processing manually started? I notice that the latest update is 2020-08-20 (last month). —Suzukaze-c (talk) 05:32, 23 September 2020 (UTC)

@Suzukaze-c: Yeah, kind of. I have to log in to Toolforge to start the Makefile recipe for the template dumps, and I also have to manually update the HTML page with month links. Not ideal. I actually ran the 2020-09-01 and 2020-09-01 template dumps but didn't write them into the page. Done. Should have a little script that writes in the month links. — Eru·tuon 05:46, 23 September 2020 (UTC)

Cool :)

(To take away the mystery, I'm using the dumps here.)

—Suzukaze-c (talk) 05:52, 23 September 2020 (UTC)

Yay! I'm glad someone else has found a use for them. So far it's only been User:Jberkel/lists/wanted, which I haven't figured out how to update for a while though I have access on Toolforge. — Eru·tuon 09:13, 23 September 2020 (UTC)

Requesting an update to 2020-11-20. _(：3 」∠ )_ —Suzukaze-c (talk) 23:52, 23 November 2020 (UTC)

Done. I'd run the process for 20201120 but not updated the "latest" link. — Eru·tuon 00:00, 24 November 2020 (UTC)

Most used RQ template - listy fun!

Hey, List-Man. I was wondering what our most used Quotation Template is. Could you make a list of the most widely used ones? I'm looking for Templates that start with RQ: My money is on either Template:RQ:King James Version or Template:RQ:Spenser Faerie Queene. One tricky thing is that some templates redirect to others (like Template:RQ:KJV and Template:RQ:Spenser FQ , so they'll need to be counted together, I guess. It'd probably be useful to get a list of all unused RQ templates too - they might want to be deleted. --Java Beauty (talk) 10:50, 3 September 2020 (UTC)

@Java Beauty: Well, here's the lazy version – the number of pages transcluding each RQ template (filtering out redirects because when a redirect is transcluded its target is as well). Lazy because this counts 50 KJV templates on a single page as 1. Have to do wikitext parsing to actually figure out the total number of template instances. But given the numbers here, there's gotta be no way any template beats the KJV one. — Eru·tuon 21:03, 3 September 2020 (UTC)

I'm happy with a lazy list. According to this research, it seems that we're an extremely religious website - I wasn't aware. The Muslims won't be happy that their book was beaten by the Jewish one, either... --Java Beauty (talk) 21:21, 3 September 2020 (UTC)

Another useful thing: It turns out that Wonderfool's Wiktionary:Easter Competition 2010 bore much fruit. EP using Template:RQ:Schuster Hepaticae, Daniel. using Template:RQ:mul:Rowling Harry Potter, and Widsith's Template:RQ:Spenser Faerie Queene. And, funnily, there are still a handful of Who Let The Dogs Out? quotes left from Wonderfool himself. --Java Beauty (talk) 21:39, 3 September 2020 (UTC)

We'll all reunite again in 2030 to analyse something else and pat ourselves of the back. --Java Beauty (talk) 21:46, 3 September 2020 (UTC)

Easy dates

Can you please make a list at User:Erutuon/lists/cheap dates (pun intended) of all entries containing {{rfdate}} or {{rfdatek}}, the author AND the name of the work? For example the one at chrematistic which looks #* {{rfdate|en}}, {{w|Ludwig von Mises}}, ''{{w|Human Action}}''. These should in theory be really easy to date, so easy in fact that even n00bs can do it. --Java Beauty (talk) 16:14, 9 September 2020 (UTC)

Behold, it is done. I used a very broad search, just looking for rfdate template comma something comma something in italics. — Eru·tuon 21:41, 12 September 2020 (UTC)

A broad search for a broad church. It's great, thanks! --Java Beauty (talk) 21:47, 12 September 2020 (UTC)

I think I checked all the English ones from that page. Any chance of an update? Maybe you can include pages containing rfdate template comma something comma quotation mark too, as that's a common format used. Daleusher (talk) 17:39, 18 October 2020 (UTC)

Yup, after the 20th when they start the next dumping process. I'll add to the search pattern. — Eru·tuon 01:20, 19 October 2020 (UTC)

a request for help with Lua

Could you please help me with Module:hu-nominals? I've been trying to fix it for several hours but one little bit is still missing. I wanted to add a new parameter option called "isg", that is, i-type singular, so that multiple-possession forms (which have an -i instead of the normal plural -k) can be inflected like singular terms (without adding -k) but should be displayed in the plural column (since it is a kind of plural). It works all right, the values are correct, with the sole exception that the value of the parameter "esm_sg" (for whether the essive-modal singular should be displayed) should be copied into "esm_pl", so if it is set, esm_sg should be displayed in the plural column just like all other singular forms, and if it is not set, it should be indicated with a dash. During my various different attempts, sometimes it was displayed even if it was not set and sometimes it was not displayed even if it was set. For example in the case of ablakai, if I enter {{hu-infl-nom|ablakai|o|n=isg|esm_sg=1}}, it should be displayed, but if I enter {{hu-infl-nom|ablakai|o|n=isg}}, it should not be displayed. Could you possibly help me? Adam78 (talk) 20:08, 14 September 2020 (UTC)

@Adam78: I've implemented the logic you intended (I think). However, it seems counterintuitive to me that |esm_sg=1 actually causes the esm_pl form to be displayed in the table if |n=isg is present. It may also have made it harder for you to implement the logic in Lua. I'd suggest using |esm_pl=1 plus |n=isg for the same effect, but maybe the existing logic makes more sense to someone who actually speaks Hungarian. — Eru·tuon 05:06, 15 September 2020 (UTC)

Thank you very much! I admit it may be counterintuitive, but I think it's consistent with the other singular forms being displayed in the plural column, and it also makes it easier for us to change the existing code, having to insert only a letter "i". (We'll need to update a few thousand entries, although I hope we can get some help from a bot admin.) Thank you once again. Adam78 (talk) 10:17, 15 September 2020 (UTC)

A space lost in my Lua

@Erutuon, hate to bother you again. Everyone asks for your help... If you ever have a little time, could you check why is it that i get my Lua templates stuck togerher, without a space between them: it is explained at el:Module talk:labels. I have tired everything... I just cannot understand it at all :( If you do not have time, it is no problem: I can always go back to my old Module linking to lemmata and not to Cateogries. Thank you. ‑‑Sarri.greek ^♫ | 05:29, 15 September 2020 (UTC)
Thank you, @Erutuon so much for your immediate response! I do not know why this happens. We use |0=xx very often to take away parenthesis. I was wondering if it had to do something with this Legacy Vector and various wikis. Thank you ‑‑Sarri.greek ^♫ | 07:14, 15 September 2020 (UTC)

@Sarri.greek: Well, turns out I was wrong about |0=! It works both ways. But see the other thing I figured out on el:Module talk:labels. — Eru·tuon 07:25, 15 September 2020 (UTC)

Your magic! again. I thank you @Erutuon for helping el.wikt! We needed these labels so much! We use the zeros for parameters noparenthesis 0=- nocat 00=- and nodisplay is 000=- so that we will not have to change keyboards all the time. So, you saved us! ‑‑Sarri.greek ^♫ | 08:15, 15 September 2020 (UTC)

Working with the graph module

Hi erutuon,

I'm trying to use the mw:Module:Graph, but I'm getting Lua error in Module:Graph at line 1406: attempt to call method 'getParent' (a nil value). Thanks to your response on this this talk page, I know that the error is beacuse I'm not using a frame object, which has a function called getParent().

My intention is to use this module inside another module, and the below code is what I have tested (that returns the described error). I was wondering if you could help me out, and tell me how to "recreate" a frame object.

 graph = require("Module:Graph")
 function p.test()
	return graph.chartWrapper{type="line", width=1000, height=500, linewidth=4, x=2, y=22}
 end

Thank you,

Ajuanca (talk) 15:31, 22 September 2020 (UTC)

Hi, it looks like the function you should be calling in a module is graph.chart. The graph.chartWrapper function is for use in templates and it's possible but massively inconvenient to use it in a module because you have to write lots of nested tables and functions. For graph.chart you need a table with a field args containing the table of arguments in your example (graph.chart { args = { --[[ args here ]] } }). Inconvenient still, but less so as it's only one level of nesting. 17:47, 22 September 2020 (UTC)

hi! thank you so much for your fast response (and sorry for my late response). With:

return graph.chart { args= {type="line", width=1000, height=500, linewidth=4, x=2, y=22} }

I got the raw otuput on the wiki:

{"legends":[],"scales":[{"type":"linear","name":"x","zero":false,"domain":{"data":"chart","field":"x"},"range":"width","nice":true},{"type":"linear","name":"y","domain":{"data":"chart","field":"y"},"zero":false,"range":"height","nice":true},{"domain":{"data":"chart","field":"series"},"type":"ordinal","name":"color","range":"category10"}],"version":2,"marks":[{"type":"line","properties":{"hover":{"stroke":{"value":"red"}},"update":{"stroke":{"scale":"color","field":"series"}},"enter":{"y":{"scale":"y","field":"y"},"x":{"scale":"x","field":"x"},"stroke":{"scale":"color","field":"series"},"strokeWidth":{"value":4}}},"from":{"data":"chart"}}],"height":500,"axes":[{"type":"x","scale":"x","format":"d","properties":{"title":{"fill":{"value":"#54595d"}},"grid":{"stroke":{"value":"#54595d"}},"ticks":{"stroke":{"value":"#54595d"}},"axis":{"strokeWidth":{"value":2},"stroke":{"value":"#54595d"}},"labels":{"fill":{"value":"#54595d"}}},"grid":false},{"type":"y","scale":"y","format":"d","properties":{"title":{"fill":{"value":"#54595d"}},"grid":{"stroke":{"value":"#54595d"}},"ticks":{"stroke":{"value":"#54595d"}},"axis":{"strokeWidth":{"value":2},"stroke":{"value":"#54595d"}},"labels":{"fill":{"value":"#54595d"}}},"grid":false}],"data":[{"format":{"parse":{"y":"integer","x":"integer"},"type":"json"},"name":"chart","values":[{"y":2,"series":"y","x":2}]}],"width":1000}

and not the graph, as I pretend. How could I solve this? --Ajuanca (talk) 13:04, 23 September 2020 (UTC)

@Ajuanca: I'm sorry, I don't know. I haven't used the graph module or the graph extension. — Eru·tuon 19:52, 23 September 2020 (UTC)

Thank you anyway Erutuon! I apreciate your help. stay safe, Ajuanca (talk) 07:49, 24 September 2020 (UTC)

We sent you an e-mail

Hello Erutuon,

Really sorry for the inconvenience. This is a gentle note to request that you check your email. We sent you a message titled "The Community Insights survey is coming!". If you have questions, email surveys@wikimedia.org.

You can see my explanation here.

MediaWiki message delivery (talk) 18:48, 25 September 2020 (UTC)

Wanted templates

Hi again. Can you generate a list of Wanted templates? One just like this but without all the Tracking crap. --Java Beauty (talk) 14:44, 26 September 2020 (UTC)

Well, here it is. Was surprised it took all of 8 or 9 minutes to generate, but the templatelinks file is huge so it makes sense.

You can also remove tracking templates from Special:WantedTemplates by adding importScript('User:Erutuon/scripts/hideUnwantedCategories.js'); to your common.js. (Use this link to see all 5000 entries, though still it doesn't have all the more obscure templates on my list.) — Eru·tuon 07:39, 9 October 2020 (UTC)

Looks good! I trimmed the list to suit my own needs and put it at WT:Todo/RQ Crap. You can probably delete the page you made, as I'm not gonna use it. --Daleusher (talk) 12:48, 9 October 2020 (UTC)

I'm going to keep it (don't know if I'll update it often though). It's a fun museum of the weird templates people try to use. — Eru·tuon 17:55, 9 October 2020 (UTC)

Museum, indeed. My favourite artefact is Template:利用者:Lemonsquash/ふりがな. I can but wonder what it is for. Daleusher (talk) 21:01, 10 October 2020 (UTC)

• Hey. Any chance on an update? I added your script to my common, but anything used less than 3 times is still invisible. Alexfromiowa (talk) 23:11, 18 January 2021 (UTC)

  @Alexfromiowa: Oh yeah, there you go. I can update it in two or three days as well when the next dump comes out. — Eru·tuon 19:59, 19 January 2021 (UTC)

User:Erutuon/lists/rfdate and rfdatek leaderboard update 2

Hi there. I've been chopping down the trees at User:Erutuon/lists/rfdate and rfdatek leaderboard. At the next update, can you regen this? --Java Beauty (talk) 21:35, 26 September 2020 (UTC)

Sure thing. (I can update the cheap dates too, but if you edit that page between October 1 and when I update the page, this'll probably happen again, because the dump file will reflect the state of the wiki on October 1 but takes a few days to generate.) — Eru·tuon 22:13, 26 September 2020 (UTC)

Sounds good. I'm also getting close to my monthyl limit. Time for me to be blocked, I'm afraid. --Java Beauty (talk) 22:14, 26 September 2020 (UTC)

Swahili version of {{WOTD}}

Hi! I was asked by an editor at the Swahili Wiktionary to create a version of {{WOTD}} there, and I think I mostly managed to do it (see "sw:Kigezo:NLS"), but I can't seem to get the equivalent of {{WOTD/previous or next day}} (which is called "sw:Kigezo:NLS/siku iliyotangulia au inayofuata") to work – it is throwing an invalid time error. Any idea what's wrong? — SGconlaw (talk) 16:53, 4 October 2020 (UTC)

Oh, I figured it out. It seems {{#time}} only accepts a month name in English, so I updated "sw:Kigezo:NLS" to convert the names in Swahili to English and the error disappeared. Not sure if there is a tidier way to do it. — SGconlaw (talk) 17:22, 4 October 2020 (UTC)

Hmmm, I think I still need help. The "yesterday" and "tomorrow" links at "sw:Kigezo:NLS" are not displaying the year, and I'm not sure why. See the draft Main Page at "sw:Mwanzo/sandbox". The links display properly on pages with the format "sw:Wiktionary:Neno la siku/2020/Oktoba 5". — SGconlaw (talk) 18:14, 4 October 2020 (UTC)

@Sgconlaw: Ahh, the template was only adding the year to the link if the word-of-the-day template with /year/month day existed. Changed so that it adds it always. — Eru·tuon 18:54, 4 October 2020 (UTC)

Excellent, thanks! (I guess we added that feature here because we have some pages that use the year and some that don’t?) — SGconlaw (talk) 18:56, 4 October 2020 (UTC)

Crappy date format

Thanks for this correction. It seems that's a common mistake by the sloppier editors (i.e. Wonderfool). Could you generate yet another cleanup list of all entries with similar cases containing #* ''year''' or #*''year''? You can put it at User:Erutuon/lists/ugly dates (like your mom) --Daleusher (talk) 23:26, 7 October 2020 (UTC)

@Daleusher: Okay, got most of the plain years at least, though it's not the full list because there are badly formatted dates in other formats judging from this other version of the list. — Eru·tuon 08:58, 8 October 2020 (UTC)

Cool stuff. I think I corrected all with too few or too many apostrophes. I suppose I might do the unapostrophed ones one day (but as they weren't WF's fuckupperies, maybe not). I guess a bot could fix all of them instead. --Daleusher (talk) 12:36, 9 October 2020 (UTC)

Combination of previous cleanups

Hi. To finish off the cleanup set of dates, I was wondering if there are any templates that start by Template:RQ which contain crappy date format like in your previous list. I imagine there might be around half a dozen, judging by Wonderfool's current sloppiness rating. Any chance of whipping up another list? I can't think of a humorous name for the page, unfortunately. WT:Todo/Sloppy seconds will do. --Daleusher (talk) 13:23, 9 October 2020 (UTC)

Lua parents

If it is not too much trouble... i was trying to make a very simple auto cat at el.wikt (an impudence, but, just for a few categories). The pagetitles are like:
Category:Title (languaname). All I need is to extract the language iso from that languagename. So, I tried at the last fucntion at el:Module:lang, applied at el:Module:yy. tested at el:Κατηγορία:Ετυμολογία (αγγλικά). Nothing works at Module.lang function name_to_iso. Something with parents and args. I will never understand what Lua wants... I have repeatedly asked for help at meta in vain. I shall write copypaste versions at my help page, hopefully not to bother you again. ‑‑Sarri.greek ^♫ | 19:19, 11 October 2020 (UTC)

:) Thank you. Please, consider writing a Lua textbook! you are a good teacher. ‑‑Sarri.greek ^♫ | 19:33, 11 October 2020 (UTC)

@sarri.greek: (edit conflict) Fixed! The error was because the function expected a frame object (the object that functionname in {{#invoke:modulename|functionname}} receives as a argument) rather than a string (language name). So I switched to the function that wants a string.

You can continue asking me questions, but perhaps a better place then Meta to ask questions is w:Wikipedia talk:Lua or w:Wikipedia:Village pump (technical), because there are quite a few module writers at Wikipedia who seem to pay attention to those pages. — Eru·tuon 19:37, 11 October 2020 (UTC)

A! @Erutuon that is great advice. I wish there were a something like Lua models for copypaste. I shall try there for support. Thanks a million, again. ‑‑Sarri.greek ^♫ | 19:41, 11 October 2020 (UTC)

a few questions

Hi. I have a few questions:

1. At one point you expressed interest in having me do a bot run over Ancient Greek non-lemma forms to merge {{inflection of}} calls, so that e.g. on ὀξεῖαι, you'd get {{inflection of|grc|ὀξῠ́ς||nom//voc|f|p}} instead of separate calls to nominative feminine plural of ὀξῠ́ς (oxús) and vocative feminine plural of ὀξῠ́ς (oxús). Are you still interested in having that done?
2. I would like to get rid of Category:Ancient Greek common nouns. There are only two entries in there, and both of them look to actually be adjectives. Can you fix them?
3. I would like to fix MediaWiki:Gadget-TranslationAdder-Data.js to support {{multitrans}} and {{tt}}/{{tt+}}. Conceptually it isn't hard but I haven't ever worked with Javascript much, and not at all here on Wiktionary. Can you give me a few pointers as to how you test out and debug changes to a gadget? With Lua, for example, I am used to (a) creating private copies of modules, (b) previewing changes using the "Preview page with this template" box at the bottom of the page, (c) debugging using error() as a kludgy but effective way of inserting print statements into modules. How do you do the equivalent for Javascript gadgets?

Thanks! Benwing2 (talk) 05:44, 15 October 2020 (UTC)

1. I would like to merge {{inflection of}} instances, but ideally while doing that you'd also reorder the order of inflectional categories, and I'm not sure there's agreement on which order to use among Ancient Greek editors. Perhaps if you can merge without changing the orders?
2. I've fixed those entries; they are indeed just adjectives that are only attested in the masculine or feminine. I've also deleted the category and made Module:grc-headword no longer support common gender, which doesn't really exist in Ancient Greek.
3. So the trouble with editing gadgets is they have dependencies, which are automatically handled by ResourceLoader, but have to be manually loaded when you are writing a modified version of the gadget and loading it in your common.js. MediaWiki:Gadgets-definition has the dependencies of the gadget. (You can use this script to format the page and add links.) I think you can just plug the dependencies into mw.loader.using like this in your common.js:

   mw.loader.using("ext.gadget.Editor, ext.gadget.LegacyScriptsNewNode, jquery.cookie, ext.gadget.LanguageUtils, mediawiki.util, ext.gadget.TranslationAdder-Data".split(", "), function() {
   	importScript("your version of translation adder in a userpage");
   });
   

   This isn't the most convenient method because it requires you to edit your copy of the gadget and save it on your user page, then reload the page where you're testing the gadget. Firefox used to let me edit the gadget in a built-in editor window, which was pretty convenient for testing gadgets, but they replaced that with a multiline editor for the JavaScript console and I haven't tried to use it yet. If you do have a convenient way to run multi-line JavaScript in your browser, you can take the snipped above and put the gadget's source in place of the importScript call and run that to update the gadget. Hope that's helpful; I haven't done serious gadget editing for a while. — Eru·tuon 22:36, 15 October 2020 (UTC)

Old pages

So, I was sweeping up in the old, cold, possibly haunted, parts of Wiktionary, as I like do to every few years. I found Wiktionary talk:Entry Layout explained, which was last edited in December 24 2004 (even before Wonderfool was here). And it made me think - could you get a list like this one but for other namespaces, where nothing's been editted for ages? I guess User:Erutuon/lists/really old stuff can be the name Daleusher (talk) 23:09, 18 October 2020 (UTC)

@Daleusher: Well, here's a database query for it. Kind of annoying because it uses namespace numbers, but it has URLs. — Eru·tuon 03:22, 19 October 2020 (UTC)

Chuvash character ҫ

Hi,

Could you please add handling for the Chuvash character ҫ in the translation adder? Roman look-alike ç should be replaced with Cyrillic ҫ in translations but I can't remember, which module does it! --Anatoli T. ^{(обсудить}/^вклад) 00:08, 20 October 2020 (UTC)

E.g if a user enters "çул палли", it should insert the normalised form ҫул палли (śul p̬alli, “traffic sign”). --Anatoli T. ^{(обсудить}/^вклад)

I have fixed it, now that I remembered where to fix :) --Anatoli T. ^{(обсудить}/^вклад) 03:50, 20 October 2020 (UTC)

Wiktionary:Todo/Undated English quote-templates

Hey, lad. Wiktionary:Todo/Undated English quote-templates needs updating at the next dump. I've run through all of them on the list, picking off the easy ones. Lots were non-durably-archived quotes which I mercilessly deleted, and some I couldn't figure out, will need to crank up my lexi-skillz once again. Cheers! Candle-holding servant (talk) 22:08, 25 October 2020 (UTC)

Done but I bet you've already cleared off a lot of the ones from the October 20th dump. I'll update it after the November 1st and it'll probably be significantly shorter. (For some reason with the latest dump my program that extracts all quote templates from the XML dump is taking about 10 times as long as before. Super weird. Hah, I ran the unoptimized version of the program.) — Eru·tuon 02:08, 26 October 2020 (UTC)

VisualEditor and quote toggling

I'm currently adding TemplateData to a bunch of quotation templates (for Citoid) and noticed that it's not possible to edit quotes in the visual editor (they're hidden, and clicking on the toggle option in the left navigation menu doesn't do anything). Is there an easy fix for this? Perhaps the visual editor should automatically show hidden elements? – Jberkel 00:58, 4 November 2020 (UTC)

@Jberkel: I'm pretty unfamiliar with the visual editor (both as an editor and a JavaScript tinkerer), but perhaps it's because the visibility toggle JavaScript (MediaWiki:Gadget-defaultVisibilityToggles.js) fails to run at the correct time during the loading of the visual editor. If so, this might be fixable by adding the function that enables toggling in another mw.hook('some hook related to the visual editor here').add call. Or if the visibility-toggled elements are shown when the JavaScript does not run, perhaps the function could simply exclude the elements within the visual editor from visibility toggling so that they would always be displayed within the visual editor. Those are some ideas, but you should probably talk to someone who knows more about the visual editor than me before trying them out, if you'd like to pursue this. I might be persuadable to try to fix this, but I wouldn't be able to thoroughly test it because I don't edit with the visual editor. — Eru·tuon 21:59, 6 November 2020 (UTC)

I found the problem–I changed my local CSS to not hide quotes when the visual editor is active, see Special:Diff/51371908/61048492. Perhaps we could change the global MediaWiki:common.css to something like:

.client-js:not(.ve-active) .ns-0 ol>li>ul { display: none; } (not sure about the CSS and the difference between ve-active and ve-activated, basically skip this rule if the VE is active) – Jberkel 22:19, 6 November 2020 (UTC)

The other alternative would be to run the toggle code as you suggested, on mw.hook( 've.activationComplete' ).add( function () { }. (Documented here: mw:VisualEditor/Gadgets#Code_snippets). I think the CSS solution is simpler and cleaner though. – Jberkel 23:38, 6 November 2020 (UTC)

Could be fun to know...

I know my requests are getting lamer every week, this one is especially boring, but still. Could you whip up a list of the users with the most subpages? I have a bet that Visviva (talk • contribs) is top and DCDuring (talk • contribs) is in the top 5. Darren X. Thorsson (talk) 22:46, 6 November 2020 (UTC)

I think Darren X. Thorsson is in a tie for the fewest. DCDuring (talk) 23:27, 6 November 2020 (UTC)

Apparently so because he's not on the full list, but here are the user page creation leaders. Visvisa's number 4 actually. User:Jberkel's the second because of his wanted lists. — Eru·tuon 01:01, 7 November 2020 (UTC)

Thanks as always,E. Eventually I'll be looking to delete baches of the user subpages, especially those belonging to deceased users. Interwicket, for example, is a decent candidate. My next request will be something meta, like User:Erutuon/lists/coolest requests ever. Darren X. Thorsson (talk) 01:17, 7 November 2020 (UTC)

Bible update

Any chance of regenning Wiktionary:Todo/Undated Bible? Most of the actual Bible quotes have been done, so the list in more to find quotes with crappy formats. Darren X. Thorsson (talk) 00:04, 14 November 2020 (UTC)

@Darren X. Thorsson: I updated it... but there are so many more that it might not be the same script I used last time. Or maybe you just removed that many non-Bible quotes. — Eru·tuon 01:21, 14 November 2020 (UTC)

Looks pretty good, boss. I'll eventually get through them, or if I don't I'll slap {{rfdatek}} on them Darren X. Thorsson (talk) 10:54, 14 November 2020 (UTC)

More crappy quotes please

I keep finding crappy quotes in various formats. Any chance you could paste the results of the hunt for things like in this old version of mure at Wiktionary:Todo/Undated or crappily formatted quotes? Returning2stadia (talk) 22:27, 23 November 2020 (UTC)

@Returning2stadia: I didn't find any other mdash- (or ndash-)marked quotations like #: — Shakespeare, Henry IV, Part II, [IV, 4], line 2870. Just searched for any #: or #* followed by a dash. Looks like that was a one-off thing. — Eru·tuon 21:12, 25 November 2020 (UTC)

How about anything else formatted like the quote from the Bowie song at freakiest? --Returning2stadia (talk) 20:45, 30 November 2020 (UTC)

@Returning2stadia: Well, here's my attempt though it doesn't yield many quotes aside from another Bowie song. Looked for the list syntax followed by italics followed by a link. It did turn up a lot of weird non-quote things that probably don't belong in the definition section. — Eru·tuon 04:59, 1 December 2020 (UTC)

Cool, I checked all of them. Returning2stadia (talk) 10:44, 1 December 2020 (UTC)

• Next one - any other matching the format at worst comes to worst? Returning2stadia (talk) 12:20, 3 December 2020 (UTC)

  Oof, looked up all potential quotes starting with a Wikipedia link and it's a pretty long list. (Note: while there are a lot of non-English quotes, these are only from pages that have an English section.) — Eru·tuon 21:42, 4 December 2020 (UTC)

• Yeah, the list was long, but a satisfying, finishable size. They're mostly done now - maybe a regen could be done after the next dump or whatever, without the Filipino surnames. Also, I came across a few quotes starting with a Wikisource link, maybe they can be added too. La más guay (talk) 01:02, 12 December 2020 (UTC)
• Another common thing is things like wel-begone. Sometimes the mdash is a -- too La más guay (talk) 22:59, 13 December 2020 (UTC)

Wiktionary:Todo/multiword Spanish lemmas not idiom or proverb again

Hi. Can you get an update of Wiktionary:Todo/multiword Spanish lemmas not idiom or proverb? Again, without anything that has previously been there La más guay (talk) 09:22, 19 December 2020 (UTC)

Rhyme redirects

I've moved several rhyme pages and I'd like away from the new redirects. Can you use your bot to update the mainspace pages that link to them? Here's the list:

• Rhymes:Spanish/að to Rhymes:Spanish/ad
• Rhymes:Spanish/eð to Rhymes:Spanish/ed
• Rhymes:Spanish/eɾβo to Rhymes:Spanish/eɾbo
• Rhymes:Spanish/eβa to Rhymes:Spanish/eba
• Rhymes:Spanish/eβɾe to Rhymes:Spanish/ebɾe
• Rhymes:Spanish/iðjo to Rhymes:Spanish/idjo
• Rhymes:Spanish/iŋko to Rhymes:Spanish/inko
• Rhymes:Spanish/iŋɡo to Rhymes:Spanish/inɡo
• Rhymes:Spanish/iɣo to Rhymes:Spanish/igo
• Rhymes:Spanish/iβulo to Rhymes:Spanish/ibulo
• Rhymes:Spanish/izmo to Rhymes:Spanish/ismo
• Rhymes:Spanish/oi̯ to Rhymes:Spanish/oi
• Rhymes:Spanish/oðo to Rhymes:Spanish/odo
• Rhymes:Spanish/oŋko to Rhymes:Spanish/onko
• Rhymes:Spanish/oŋxa to Rhymes:Spanish/onxa
• Rhymes:Spanish/uð to Rhymes:Spanish/ud
• Rhymes:Spanish/uβia to Rhymes:Spanish/ubja

Thanks, Ultimateria (talk) 19:54, 31 December 2020 (UTC)

Please let me know if you don't plan to change these links. I apologize if I gave the impression of voluntelling you; my intention was to make the request and describe the exact task in a single message, but I can see that it comes across as presumptuous. Ultimateria (talk) 18:42, 23 January 2021 (UTC)

@Ultimateria: Sorry, I was willing to do it at some point, but forgot about it. I've made a list of pages and will edit them soon. — Eru·tuon 07:22, 25 January 2021 (UTC)

@Ultimateria: I tried to go about this systematically and made a log showing how most of the changes or removals of symbols in your list above could be applied to all Spanish rhymes. However, some of these are debatable or problematic changes so I will rewrite the script to just use your list. — Eru·tuon 21:59, 27 January 2021 (UTC)

Okay, done at last: User:ToilBot/edit logs/2021-01/Spanish rhymes. I can quickly do any other rhyme changes now that the script is written. — Eru·tuon 22:24, 27 January 2021 (UTC)

Great, thank you! I don't expect to move anymore but if I do I'll let you know. Ultimateria (talk) 23:42, 27 January 2021 (UTC)

Templatehoard '21

Please let me know once it's available, I'd like to produce some new wanted lists. – Jberkel 15:27, 5 January 2021 (UTC)

@Jberkel: Oh yeah, forgot to do that. I've just started the first job and the second one will be done as well within 4 hours if I don't forget. — Eru·tuon 21:13, 5 January 2021 (UTC)

Started the second job, which generally takes about 25 minutes. — Eru·tuon 21:22, 5 January 2021 (UTC)

Perfect, thanks. Has anything changed how titles are handled? User:Jberkel/lists/wanted/20210101/en now contains some Appendix links (links w/ underscores). – Jberkel 23:54, 5 January 2021 (UTC)

@Jberkel: That is strange. I checked a few of the Appendix links and they were there on 2020-11-01 and would therefore be in the template dump, yet didn't appear in the list on that date. Those Appendix pages haven't appeared in the augmented entry index for either date. So I don't see how either the template dumps or the entry index could be causing this. Has your code changed at all since the previous lists were generated? — Eru·tuon 00:45, 6 January 2021 (UTC)

I thought it hadn't, but maybe I overlooked something, I'll check. – Jberkel 02:18, 6 January 2021 (UTC)

I checked the index file (augmented_entry_index/20210101.txt) and it's missing entries for some Appendix pages (:Animals, :Colors etc), which explains why they get included in the wanted pages. – Jberkel 21:38, 6 January 2021 (UTC)

I don't think it does, because I searched the 2020-11-01 index and it didn't have Appendix:Countries of the world or Appendix:Colors there either. They aren't entries so they don't have a language code and can't be put in the index. — Eru·tuon 21:41, 6 January 2021 (UTC)

Only links to appendix pages that start in a language name and a slash should be noticed by the wanted entries script, like Appendix:Lojban/bajbakni; others, like the previously mentioned ones, should be ignored. — Eru·tuon 21:45, 6 January 2021 (UTC)

Ok that makes sense, but this logic never existed, so why didn't this happen in the previous runs? I also don't understand why some editors use forms like {{l|en|Appendix:Colors}} when there's no need for it. – Jberkel 22:52, 6 January 2021 (UTC)

Fortunately, at least in American Samoa, it was an unaccounted-for condition in a script of User:MewBot. — Eru·tuon 23:14, 6 January 2021 (UTC)

Blue link, target missing

Dear @Erutuon, hope 2021 will be good for you! I was wondering. When there is no lemma for a link, we get red (as in Wanted Pages. When linking to a language that has not been written in this same page yet, is there a way to mark it? perhaps with a different colour, and create a Category:Wanted lemmata? ‑‑Sarri.greek ^♫ | 00:13, 8 February 2021 (UTC)

@sarri.greek: To color for instance a link to the Greek section on a page that only has an Ancient Greek section, you can install the OrangeLinks gadget in the gadgets tab of your preferences. There isn't a way to create a category because it would be expensive for link templates to get the wikitext of the page and check for a language section and add a category, but User:Jberkel/lists/wanted/latest provides lists made by analyzing templates in the dump. It includes both nonexistent pages and pages that don't have a section for the language. — Eru·tuon 18:34, 8 February 2021 (UTC)

So, would it expensive to look in one page? -Ι was thinking for el.wikt, where we have only 2 modules with this linking- something like

{{l|xx|theword}}
if {{xx}} lookinthepage [[theword]] == nil, then write {{l|xx|theword|font color:orange theword}} &  write at page [[theword]] [[Category:Wanted lemmata (xx)]]
else do as usual

:) sorry for the naive question. ‑‑Sarri.greek ^♫ | 07:49, 10 February 2021 (UTC)

@sarri.greek: It depends how many links are on the page. It will not be too expensive if there are only a few linking templates on the page. You could try it and see if it causes Lua memory errors in larger pages. It is only too expensive here because some pages (particularly pages with lots of links, in translations and derived terms and inflection tables) are already running into the Lua memory limit even without checking whether linked-to entries exist, and others are close behind them. There might be fewer links per page in Greek Wiktionary entries. — Eru·tuon 09:56, 17 February 2021 (UTC)

Related discussion

• @Erutuon, sorry for intruding here, but I had one asking: I have enabled OrangeLinks in my Preferences, but I edit using mobile & therefor cannot see orange links. Have I no way to have them? Thank you. -_⸘- inqilābī ^{‹inqilāb·zinda·bād›} 12:13, 6 April 2021 (UTC)

  @Inqilābī: MediaWiki:Gadgets-definition doesn't enable OrangeLinks on mobile. I'm not convinced it should be run on mobile by default; it loads the lists of categories for all entries linked from the current page, which could use too much memory and CPU on some mobile devices. But you can enable it for yourself by pasting the following code to your common.js:

  mw.loader.using(["mediawiki.Uri", "mediawiki.Title", "mediawiki.util"], function () {
  	mw.loader.load("/w/index.php?title=MediaWiki:Gadget-OrangeLinks.js&action=raw&ctype=text/javascript");
  }).then(function() { console.log("Manually loaded OrangeLinks") }, function() { console.log("Failed to manually load OrangeLinks") });
  

  This just manually replicates what the MediaWiki server does to load the OrangeLinks gadget when it is enabled in preferences. (This will run the gadget only once if you have the gadget enabled in preferences and visit the desktop site.) Then if you visit a page with many entry links, you can see how it performs on your device and decide whether to keep it. — Eru·tuon 01:26, 7 April 2021 (UTC)

  @Erutuon: Thanks for all the explanation! So, do you mean that OrangeLinks is going to work only in the desktop view? Actually I can already see orange links when I switch over to the desktop view, but editing in the desktop view is cumbersome, that is why I prefer editing in the mobile view. So basically it’s a forlorn hope for me, right? (Also I know nothing about programming, so until I learn ’bout those stuffs, your technical suggestions would be of no help to me.) But again thanks anyway. -_⸘- inqilābī ^{‹inqilāb·zinda·bād›} 11:30, 7 April 2021 (UTC)

  @Inqilābī: Yes, the only way to enable OrangeLinks is to add the code as described above. I have no plans to get the gadget in preferences to work on mobile. — Eru·tuon 18:46, 7 April 2021 (UTC)

Page history showing blocked user name

Hello, Erutuon. Thank you for rapidly taking care of recent troll edits. On the page history for Jewish Question, a removed user name is still visible in a description of my edit undoing that editor's contribution. I don't know if you want to remove my edit from the history or something, but whatever is necessary is fine with me. Happy editing, Cnilep (talk) 03:47, 2 April 2021 (UTC)

Triggering |alts=

Hello. {{alt}} was created lately as a shortcut of {{alter}}. I have noticed that |alts= (used in {{desc}}) does not work when {{alt}} is used in the lemma entry instead of {{alter}}. Could you please fix the problem? Thanks in advance! -_⸘- inqilābī ^{‹inqilāb·zinda·bād›} 12:25, 6 April 2021 (UTC)

@Inqilābī: Hopefully done. — Eru·tuon 00:53, 7 April 2021 (UTC)

Still not working, I think. See here for the Bengali descendant. -_⸘- dictātor·mundī 13:47, 12 May 2021 (UTC)

@Inqilābī: That entry's working now. I didn't have an example page to test so I just assumed my module edit worked. — Eru·tuon 17:00, 12 May 2021 (UTC)

Thanks! -_⸘- dictātor·mundī 17:32, 12 May 2021 (UTC)

Rusyn Flag Update

Notice concerning Wiktionary:Language flags list: CanadianToast (talk • contribs) updated the flag for the Rusyn language. --Apisite (talk) 14:24, 15 April 2021 (UTC)

@Apisite: Updated. The Rusyn change was reverted, but some other flags were added since I last updated. — Eru·tuon 19:27, 15 April 2021 (UTC)

User:Erutuon/scripts/gadgets-definition.js

Hi, thanks for creating this very useful script. In fact, this is also quite useful on non-English language wikis. But at the moment, I need to paste this into the console and tweak the entry condition to get it to work on non-English wikis.

A change like tweaking line 15 from

if (!(mw.config.get("wgPageName") === "MediaWiki:Gadgets-definition"

to

if (!(mw.config.get("wgNamespaceNumber") === 8 && mw.config.get("wgPageName").indexOf(":Gadgets-definition") !== -1

would work. The namespace name is localised in different wikis, but the page name ("Gadgets-definition") is English everywhere. Best, SD0001 (talk) 09:42, 17 April 2021 (UTC)

@SD0001: Good idea. Done, though in a slightly different way. Happy to hear somebody likes my script. — Eru·tuon 23:45, 17 April 2021 (UTC)

Hi again Erutuon, with https://gerrit.wikimedia.org/r/c/mediawiki/extensions/Gadgets/+/731284 merged, can you update your script to wiklink JSON pages in gadget definition as well? Thanks! SD0001 (talk) 17:05, 5 January 2022 (UTC)

Question about Module:string

I have two questions about "Module:string":

• Is it possible to use {{#invoke:string|replace}} to find a string of text enclosed in straight double quotation marks (like "Introduction") and replace it with curly single quotation marks (like ‘Introduction’)? If so, what should I put as the pattern string and the replacement string?
• Again, if the above is possible, can I at the same time invoke {{#invoke:string|replace}} to also replace straight apostrophes (') with curly ones (’)?

— SGconlaw (talk) 17:25, 2 May 2021 (UTC)

@Sgconlaw: I've mostly used {{#invoke:string|gsub}} for replacing because it is minimally different from the Lua function by the same name, but it looks like using replace is as simple as {{#invoke:string|replace|source=text with "Introduction"|pattern="Introduction"|replace=‘Introduction’}}. — Eru·tuon 18:05, 2 May 2021 (UTC)

I was not clear enough; what I meant was that the text in between the quotation marks will change depending on what is input by editors. I tried using "." as the pattern string and ‘.’ as the replacement string (since a full stop refers to any characters) but it didn’t work. — SGconlaw (talk) 18:09, 2 May 2021 (UTC)

Oh, okay. That requires more thought and might need multiple replacements or a dedicated Lua function to get it right, particularly if it has to distinguish single quotation marks from apostrophes (which is probably impossible to do infallibly). — Eru·tuon 18:11, 2 May 2021 (UTC)

At the moment I was thinking only of paired straight double quotation marks, but if paired straight single quotation marks could also be dealt with that would of course be great. Essentially I suppose the question is whether there is an easy way to match a string in the form "*" or '*' where * is a varying string. Replacing characters that don’t change seems quite straightforward. — SGconlaw (talk) 18:47, 2 May 2021 (UTC)

@Sgconlaw: Yes, there is. You can try {{#invoke:string|gsub|text with "quotation marks"|"(.-)"|‘%1’}}. However, single quotation marks are more complicated because they're ambiguous with apostrophes, at least when word-initial, and disambiguating in some cases requires understanding the text. For instance, any simple regex replacement for paired single quotation marks would match 'Tis the kings' and replace it with ‛Tis the kings’, but based on the most likely interpretation it should be ’Tis the kings’. Technically there is a word tis, so even knowing that 'tis is a word, the module can't reliably determine whether 'tis has an apostrophe or opening single quotation mark because it can't understand English. So a simple regex replacement would work for many single quotation marks because initial apostrophes are relatively rare, but there would be some cases in which it would convert an apostrophe to ‘ instead of ’, and there is no way to correct this without human intervention. — Eru·tuon 19:38, 2 May 2021 (UTC)

Thanks. Yes, that works; I implemented it at {{RQ:Stevenson Osbourne Wrecker}} where the issue with single quotation marks doesn't occur. — SGconlaw (talk) 16:28, 3 May 2021 (UTC)

Some updates for cleanup pages

Hi E. It#s been a long time since I've been bothering you. Any chance you can regen Wiktionary:Todo/Undated or crappily formatted quotes, Wiktionary:Todo/Undated English quote-templates, User:Erutuon/lists/wanted templates not tracking, User:Erutuon/lists/ugly dates and User:Erutuon/lists/cheap dates? Indian subcontinent (talk) 23:29, 25 May 2021 (UTC)

@Indian subcontinent: I will try to at some point. Some of the code that I used to make those lists has broken and I've had trouble fixing it... — Eru·tuon 06:16, 27 May 2021 (UTC)

@Indian subcontinent: Well! I did all of them (took me ages because I'm so disorganized), and it looks like you made a lot of progress in six or seven months... good job! — Eru·tuon 05:07, 18 June 2021 (UTC)

Translations — target languages

I posted this on Grease Pit [here] and wonder if you can help, or may know who can!

For some weeks now (on my screen at least) the "select target languages" option appears not to be working. However, on closer inspection today, it is working but the choice buttons are invisible. So you can make a choice, but only by guessing/clicking where the button might be.

Thanks — Saltmarsh. 07:48, 28 May 2021 (UTC)

Special:AbuseFilter/115

Hello. Can you edit the filter to skip edits such as Special:Diff/62622446? --Minorax (talk) 09:01, 29 May 2021 (UTC)

Looked at other edits that tripped the filter and most of them are similar to the diff provided above. --Minorax (talk) 09:03, 29 May 2021 (UTC)

That's the only situation in which there can be a blank line before the first heading, and it's what the filter is meant to detect. There isn't supposed to be a blank line there according to WT:NORM § Headings. — Eru·tuon 17:42, 29 May 2021 (UTC)

Listing multiple pages in a reference templet

Is that possible within a single reference templet?— like in the reference in this entry. If possible, could you please fix the templet so that ‘pages’ is displayed instead of ‘page’? Thank you so much. -_⸘- dictātor·mundī 10:23, 30 May 2021 (UTC)

adding a feature to {{R:ErtSz}} and its variant {{R:TotfalusiNagyEty 2001}}

Hello again,

we're repurposing the module you were kind enough to create for us ({{R:ErtSz}}) at {{R:TotfalusiNagyEty 2001}} (an etymological dictionary hosted at the same website with the same structure). I'm trying to add a feature that it should be able to look up terms that are relevant but not exactly identical to PAGENAME. It would be supplied as the first parameter, so if it's not a number (i.e. a homonym), the module should try to interpret it as an alternative term to look up in the database and there would be error only when this attempt too fails.

For example, the data relevant for sípol is available in the entry of síp, so I'd like to be able to use the link {{R:TotfalusiNagyEty 2001|síp}} in the entry of sípol. – In fact, this feature would be useful for {{R:ErtSz}} as well: we've been managing it by using the earlier version {{R:BarcziOrszagh 1962}}, which still requires the code to be manually entered, but that's the only way we've been able to link to something other than PAGENAME.

I tried to modify Module:R:TotfEty accordingly but when I insert {{R:TotfalusiNagyEty 2001|síp}} in sípol, I get the error "Lua error in Module:R:TotfEty/data at line 13: attempt to concatenate local 'key' (a nil value)", so the problem is manifested in the "data" module. Could you please help us resolve it? Adam78 (talk) 20:23, 2 June 2021 (UTC)

@Adam78: Okay, I think I've done what you've requested, but give it a try and let me know. — Eru·tuon 22:04, 2 June 2021 (UTC)

Thank you very much! Based on what I've tried it works perfectly. The only place where I noticed a problem was where I tried to link to a term that is homonymous, namely minap, where mi² should be linked. Something like {{R:TotfalusiNagyEty 2001|mi|2}} would be too complicated to implement, I suppose? – If it is, never mind. I don't mean to take advantage of your helpfulness; we'll manage those cases in some other way. Adam78 (talk) 22:20, 2 June 2021 (UTC)

@Adam78: That sounds like a convenient way for the template to work. I think I've done it. — Eru·tuon 23:18, 2 June 2021 (UTC)

Thank you, you're amazing! 🎊 🎉 👑 🎈 💐 🍰 🍷 Adam78 (talk) 23:34, 2 June 2021 (UTC)

Oops! Just noticed that argument 2 is treated as the sense even when it's a number. Should the sense be the argument immediately after the number then? — Eru·tuon 00:24, 3 June 2021 (UTC)

Customized edittools

Hello, I would like to customize the edittools box below the editing space, and you seem to have one. I tried to look in the various help pages, and get inspiration from users, but I didn't seem to find something that works. It's a combination of common.js / monobook.js / edittools.js / edittools subpages, but I can't seem to find a working combination of content. Could you please help me out on the minimal setup to have this achieved? Thank you! Sitaron (talk) 08:57, 7 June 2021 (UTC)

I was actually able to make it work reusing User:Eirikr version of the scripts. Seems this had been broken for a while...

Thx. Sitaron (talk) 12:24, 7 June 2021 (UTC)

|pcal=

We do not have a parameter |pcal= (for partial calques) analogous to |cal= (for calques) for use in {{desc}}. I think we can have it? ·~ dictátor·mundꟾ 23:38, 12 June 2021 (UTC)

@Inqilābī: Sorry for the slow response. I took a look at the source code of the template and there's already a |pclq= parameter, so I just documented it in Template:descendant/documentation. — Eru·tuon

Oh, I did not know the parameter was there; I should have checked it out myself. Thanks! ·~ dictátor·mundꟾ 15:08, 18 June 2021 (UTC)

Lua local environment

What's your local environment for running Lua? My local environment involves many hacks. Thanks. EdwardAlexanderCrowley (talk) 02:48, 26 June 2021 (UTC)

@EdwardAlexanderCrowley: Hi, I've got a setup for running some English Wiktionary modules in Lua 5.3. (It would probably work under 5.4 too, but I haven't tried.) It runs in Linux Mint (old computer) and Ubuntu (Windows Subsystem for Linux 2). It does some of what yours does. All the English Wiktionary stuff is nested under the directory for wikimedia.enwiktionary in package.path. To use it I require "wikimedia.enwiktionary". This sets up the environment so that it can run the English Wiktionary modules that I use, without any changes in code. It defines some mw stuff and adds a function to package.searchers so that require("Module:languages") works. I manually run a script once in a while to download Module:languages and some other modules.

I use a C-based library for much of mw.ustring, which is hopefully faster than the Lua and PHP versions of ustring, though I didn't benchmark it. (Worked on that today, so it's a bit easier to build. HOpe to upload it to Luarocks eventually.) So far I haven't needed to install mw.title or mw.text or mw.html for any of the modules that I've used, but I could add those. (Preferably with a script to just download mw.text or mw.html from GitHub and install them.)

If you would like to try setting up the environment, I could work on making it easier to do: make a GitHub repository with scripts? (Only one person asked me before, Allahverdi Verdizade, over Discord, and it was hard to remember everything that I did to set up the environment.) And any modules besides Module:th-translit that you particularly want to use? — Eru·tuon 07:04, 26 June 2021 (UTC)

I see. I've also found [1]. I'm also using Linux, so I can hack by myself. Thanks for information. EdwardAlexanderCrowley (talk) 07:44, 26 June 2021 (UTC)

*klajjaz

Hi there ! I saw that you deleted PGmc *klajjaz, but although it is only inherited in PWGmc, doesn't a PGmc need to exist to connect it to PIE ? (i.e. if it existed in West Germanic and the root in PIE, then at least a base should have been present at PGmc, like *klajj-, right ? The etymology at PWGmc states from *gloy-(y)ó-s though, so... Leasnam (talk) 05:39, 4 July 2021 (UTC)

@Leasnam: I'm operating on the understanding that the consensus is that we won't reconstruct a Proto-Germanic term when there are no reflexes outside the West Germanic branch. I believe this was discussed and agreed on at some point, but I can't remember where. I've been taking page merging requests from Victar, and he or Mahagaja may be able to give you more information. — Eru·tuon 18:49, 4 July 2021 (UTC)

1. With no entra-WG cognates, many PWG words could have multiple conceivable PG reconstructions; beside *klajjaz, *klajaz and *klajijaz are also possible, not to mention readjustments in case endings. 2. WG-only PG entries really should be moved instead of recreated as PWG entries to better attribute their edit histories; note that *klajjaz was merged into *klaij, not strictly deleted. --{{victar|talk}} 20:02, 4 July 2021 (UTC)

Ok Thank you. Makes sense. Leasnam (talk) 20:46, 4 July 2021 (UTC)

@Victar, Eruton I'm assuming this does not also apply to Norse-only and Gothic-only reconstructs, which are not merged (?) Leasnam (talk) 20:48, 4 July 2021 (UTC)

IMO, yes, but that's not something I want to spearhead at this time. --{{victar|talk}} 03:10, 5 July 2021 (UTC)

@Victar, Erutuon: There's Icelandic kleggi, Norwegian klegg, Swedish klägg, and Danish klæg though, which can either mean “gunk” or “horsefly < the sticky one”. See de Vries, Jan (1977) Altnordisches etymologisches Wörterbuch [Old Norse Etymological Dictionary] (in German), 2nd revised edition, Leiden: Brill, page 315 —caoimhinoc (talk) 17:59, 10 September 2021 (UTC)

Wiktionary:Todo/multiword French lemmas

How about an update on Wiktionary:Todo/multiword French lemmas and Wiktionary:Todo/multiword Spanish lemmas not idiom or proverb? — This unsigned comment was added by Roger the Rodger (talk • contribs) at 15:31, 8 July 2021 (UTC).

@Roger the Rodger: Done at last! I dreaded it so much because I haven't made a proper script for this... — Eru·tuon 06:43, 17 July 2021 (UTC)

• Nice! Thanks, I´ll make some headway with them Roger the Rodger (talk) 21:28, 17 July 2021 (UTC)
  • One thing though, instead of removing those on the original list, you should remove those on this more recent list Roger the Rodger (talk) 22:03, 17 July 2021 (UTC)
    • @Roger the Rodger: Yep, removed those as well. — Eru·tuon 05:36, 18 July 2021 (UTC)

Props

Props for locking the module; I'd thought about doing that myself for several days, but didn't want to 'endorse' one revision over another. (But if I had realized, as you did, the terrible load it was putting on the servers, I would've.) - -sche (discuss) 07:25, 23 July 2021 (UTC)

aa-IPA

Hi, I was wondering if you could help out with a bot operation. @Malku H₂n̥rés was so kind to create Module:aa-IPA, and to effectively implement it into the Afar entries, we'd need something like what you've done with Hausa: To take the headword of entries and place them into the template, while replacing the earlier {{IPA}} template. If it's possible, I would not touch multiword terms, since some of them (like taban kee affara) have secondary stress, while others (like baaxo danan) will need respelling whithin the template. Could you help me out? Thanks in advance. Thadh (talk) 11:52, 5 August 2021 (UTC)

Two more things: unlike {{ha-IPA}}, this {{aa-IPA}} doesn't take an asterisk before the template (since it generates it), and in the existing entries, pronunciation and hyphenation should be replaced by the single template {{aa-IPA}} which automates both. @Thadh, are there Afar entries with multiple headwords or multiple pronunciation section? The answer to these questions may be helpful for Erutuon, like with Hausa. Malku H₂n̥rés (talk) 19:47, 5 August 2021 (UTC)

Sure, there are many that distinguish different accentuation, like áwka vs awká and ála vs alá. There shouldn't be too many of them, so if that's a problem, you could generate a list and I'll fix them by hand. Thadh (talk) 19:50, 5 August 2021 (UTC)

@Malku H₂n̥rés, Thadh: I'll modify my Hausa scripts to do this when the page dump finally comes out. I can generate the list of pages with multiple headwords pretty easily. — Eru·tuon 20:14, 5 August 2021 (UTC)

The dump finally came out yesterday (my time) so I will work on this soon. — Eru·tuon 06:55, 11 August 2021 (UTC)

@Malku H₂n̥rés, Thadh: Made a list of Afar headwords without spaces from the dump. The 46 pages that two or more distinct headwords probably have to be done manually because I don't have bot code to handle headers yet, but I can add {{aa-IPA}} to the 1090 pages with just one headword or two or more identical headwords. That will considerably reduce your workload! — Eru·tuon 23:09, 11 August 2021 (UTC)

@Malku H₂n̥rés, Thadh: I think I'm ready to go. But could one of you could quickly check these planned changes and let me know if it looks okay? (I skipped -i#Afar because it has "masculine" and "feminine" labels.) — Eru·tuon 03:39, 12 August 2021 (UTC)

@Malku H₂n̥rés Does the script exclude hyphenation with suffixes and prefixes? We've forgotten to discuss this. If it does, then we're ready to go, if not, we should probably fix that. Also, the hyphen should be given in the IPA (e.g. -í would be /ˈ-i/). For the other (non-morpheme) entries the changes look okay. Thadh (talk) 05:56, 12 August 2021 (UTC)

Also, evidently there is a problem with the hyphenation in words with more than two syllables (abaareyna, abeyna, absuma all don't work). Thadh (talk) 06:02, 12 August 2021 (UTC)

Here is the list without the headwords with hyphens, including prefixes and suffixes but also words with internal hyphens, like bacra-taban. Or will {{aa-IPA}} correctly handle the headwords with internal hyphens? (I've fixed where the script was inserting the full list of headwords found on each page into {{aa-IPA}}....) — Eru·tuon 07:11, 12 August 2021 (UTC)

As I've stated above, the script can't possibly handle multiword terms (and those including hyphens), since they have unpredictable pronunciations. Thanks for taking that out. Thadh (talk) 07:49, 12 August 2021 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ Lots of messages. If I sum up I've to fix these bugs:

1. remove hyphenation if the pagename begins or ends with a hyphen.
2. for -í, it's just because of the shape of the lemma, there's no consonant before the stressed vowel, just a hyphen, I'll fix that.
3. fix a hyphenation problem related to Lua patterns. When it hyphenates "alalalalala" it returns "a•lala•lala•la", so I'll use the %f thing.

Malku H₂n̥rés (talk) 08:21, 12 August 2021 (UTC)

Okay, forgot to post an update. I've run the bot so more than 1000 pages have {{aa-IPA}} now. This list shows pages that have headwords to which {{aa-IPA}} will probably need to be added manually. — Eru·tuon 04:23, 13 August 2021 (UTC)

Problem 2 didn't exist actually. I fixed problem 3 first so hyphenation appeared as perfectly working, but I tried to fix problem 1 and now hyphenation doesn't appear at all. It's only due to the line 74. Malku H₂n̥rés (talk) 08:11, 13 August 2021 (UTC)

Curly quotes

Hi, Sorry for the trouble. I seem to have remembered uncorrectly that curly quotes in titles are permitted. Too bad. When will that error be fixed? Cheers, H. (talk) 18:31, 19 August 2021 (UTC)

@Hamaryns: I doubt we'll ever use curly quotes and apostrophes in entry titles. It would be a lot of work to move all the pages. (There are 24522 mainspace titles with ' as of 2021-08-01.) There was a vote on using curlies in policy pages (Wiktionary:Votes/2020-07/Converting policy and guide pages as for quotes and apostrophes), but it failed; I'd guess that the same vote for mainspace titles would be even less likely to pass. We've had the rule of straight apostrophes in at least English and French entry titles for as long as I can remember, but in the text of entries people often use curlies (even for links to entries) because the vote didn't prohibit that. — Eru·tuon 18:44, 19 August 2021 (UTC)

Module:zh/data/och-pron-ZS/documentation

Module:zh/data/och-pron-ZS/documentation, Module:zh/data/och-pron-BS/documentation and Module:zh/data/ltc-pron/documentation shouldn't be protected. EdwardAlexanderCrowley (talk) 10:04, 31 August 2021 (UTC)

@EdwardAlexanderCrowley: I reduced their protection levels. When I protected all the modules I forgot to exclude documentation pages. — Eru·tuon 16:32, 31 August 2021 (UTC)

Thanks, all updated. EdwardAlexanderCrowley (talk) 16:50, 31 August 2021 (UTC)

Help with Template:Han char

Discussion moved to Template talk:Han char#Three or more stroke parameters.

User:Erutuon/misplaced_categories

Hi, I'm moving Reconstruction:Proto-Celtic/emban to Reconstruction:Proto-Celtic/amben because it was misspelled and I came across your list. I'm wondering what it is and how did you find them and what do you do with them? —caoimhinoc (talk) 15:39, 10 September 2021 (UTC)

@Caoimhin ceallach: I generate this list with a program that has three tasks that it performs using files from the dump: get the file to convert language codes to names from the pages-article.xml dump file, makes a list of reconstruction pages and the categories they have from the page.sql and categorylinks.sql dump files, and then looks through the categories of each page that are language-specific (like Category:Proto-Celtic lemmas and Category:sla-pro:Mythological creatures and Category:Requests for etymologies in Proto-Indo-European entries) and keeps the ones where the language in the category title doesn't match the language in the reconstruction title. (The last step took the most time to perfect.) I made this because of a request at Wiktionary:Grease pit/2021/May § Having modules check language name against language code in Reconstruction namespace. I describe more about the process there. I don't have specific plans to deal with the list myself, but anyone is welcome to look at the list and edit any of the entries on it to put them in the right categories. Though I should probably update it. — Eru·tuon 16:46, 10 September 2021 (UTC)

I see. Thanks. —caoimhinoc (talk) 17:37, 10 September 2021 (UTC)

Unicode 14.0

Please update Unicode-related modules to Unicode 14.0 if you have automatic scripts for them. (I am not rushing you.) --Octahedron80 (talk) 02:25, 15 September 2021 (UTC)

Module:Unicode data --Apisite (talk) 02:40, 27 September 2021 (UTC)

Ccv2020 (talk • contribs) made the addition at the Vietnamese edition. --Apisite (talk) 03:56, 29 September 2021 (UTC)

@Octahedron80, Apisite: Done. The instructions for updating are at Module:Unicode data/documentation#Updating. Not as convenient as they could be. Would be nice to have a Pywikibot script that would do all the work in one go. Looks like Vietnamese Wikipedia is missing Module:Unicode data/names/01A and Module:Unicode data/names/01C, which were created with the update. — Eru·tuon 17:22, 29 September 2021 (UTC)

Undeletion

Hi. I noticed you recently undeleted a lot of my templates (thanks :)). Any chance you could restore all of these Wonderfool entries that Equinox brutally zapped back in June (edit summary Mass deletion of pages added by Indian subcontinent)? All except the categories, please, which are probably all empty. Roger the Rodger (talk) 13:50, 17 September 2021 (UTC)

musicke

What do you think about the quote at musicke? I doubt 2001 should be given as the year Roger the Rodger (talk) 18:58, 18 September 2021 (UTC)

Here is the 1904 reprint of the actual 1623 work. Chuck Entz (talk) 19:11, 18 September 2021 (UTC)

Accelerated Asturian

Hi again. Can you add an WT:ACCEL function to Template:ast-noun-mf? Roger the Rodger (talk) 08:56, 22 September 2021 (UTC)

Module:headword/data request

Hi, could you please add to Module:headword/data the following non-lemma POSes:

• excessive adjective forms
• excessive adjectives
• excessive adverbs

These POSes are only used in Basque (as far as I know) but for categorization purposes they should work in the same way as superlative/comparative adjectives/adverbs. --Santi2222 (talk) 21:13, 25 September 2021 (UTC)

Special:diff/64060705

ja-gv and ja-see-kango should be added to AbuseFilter/2. Crowley666 (talk) 08:44, 29 September 2021 (UTC)

@Crowley666: Done. — Eru·tuon 15:34, 29 September 2021 (UTC)

Various fixes requested for Module:la-pronunc

Hello, thank you for putting a stop to the edit-warring on Module:la-pronunc. Unfortunately, the final revision that was frozen in place when the page was protected included several outright errors. Could you edit to fix the following mistakes (or alternatively, unprotect the page now that it’s been over a month since the edit warring that prompted it)?

• Transcription of dark l is incorrectly omitted before consonants:
  • Fix: Add {"l", “ɫ̪"} as the new line 160 (after current line 159 and directly before the current line 160) and remove lines 168 {"l([eɛaoɔuʊ])", “ɫ̪%1"} and 169 {"l$", “ɫ̪"} which will be made redundant.
• /ts/ in etsi is incorrectly transcribed as [s̠ː] with assimilation:
  • Fix: Change line 138 back to {"[d]([.ˈ]?)s", "s%1s"} (as in the previous revision)

I confirmed with the author of the last revision, The Nicodene, that both of the above are mistakes. The following two cases are slightly less straightforward, but seem to clearly produce unintended results:

• /r/ is currently transcribed everywhere as as [ɾ]: The Nicodene agrees with me that the current transcription [ɾː] for geminate /rr/ is a dubious and unintended result. Aside from that, the general value of Latin /r/ is more commonly reconstructed as [r] than [ɾ], so for a transcription that uses one IPA letter unconditionally, [r] is preferable, especially since [r] can serve as a reasonable broad transcription of [ɾ], but not the reverse.
  • Fix: Remove line 124: {"r", "ɾ"}
• /t/ is currently shown as optional [(t̪)] in complex onset clusters like word-initial /str/ or word-medial /s.tr/. This appears to be an unintended result of code intended to show optional elision of coda /t/ when preceded by /s/ and followed by a heterosyllabic consonant in contexts like post.quam.
  • Fix: Edit line 139, {"st([.ˈ]?)([^aeɛiɪoɔuʊyʏe̯u̯])", "s(t)%1%2"}, to require a syllable break after the t: {"st([.ˈ])([^aeɛiɪoɔuʊyʏe̯u̯])", "s(t)%1%2"})

Again, you can see discussion on the linked page where The Nicodene assents to these two changes.

Thank you! --Urszag (talk) 02:55, 5 October 2021 (UTC)

@Urszag: Please make these changes in Module:la-pronunc/sandbox and I or someone else can transfer it over. — Eru·tuon 17:36, 5 October 2021 (UTC)

I don't want to be the go-to person for edit requests in the module and was annoyed. I reconsidered because of the work you did putting together the edit request, but unfortunately I think some of the edits will not work as intended (like the dark l edit). Editing a sandbox is a good idea to reduce the amount of work the servers have to do changing the Latin transcriptions across Wiktionary, and if you also edit its testcases, you can verify that the module works. — Eru·tuon 17:49, 5 October 2021 (UTC)

Thank you very much! I updated the test cases and implemented a non-buggy version of the changes mentioned above in the sandbox, as well as the following additional changes:

• fixed a bug where /sw/ displayed as [zv] in Ecclesiastical Latin
• fixed a bug where intervocalic /j/ displayed with inconsistent length in Ecclesiastical Latin
• disabled a contested rule that gave /u̯u/ and /i̯i/ special transcriptions as [u̯o] and [i̯e] (discussed on the Template talk:la-IPA talk page where the 2 participants were both in favor of removing)
• replaced a contested rule transcribing /a/ as [ɑ] in Ecclesiastical Latin (discussed on Template talk:la-IPA where the 4 participants were in consensus against [ɑ], with 3 in favor of replacing it with [ä] and one in favor of replacing it with [a])

There is one remaining bug I could not fix where the letter K is improperly displayed as an affricate/fricative in Ecclesiastical Latin, which causes two testcases to fail.--Urszag (talk) 03:56, 6 October 2021 (UTC)

I fixed the k thing in a hacky way, by replacing it with q in the "phonemes" stage and then replacing it with k again later. Copied the version from Module:la-pronunc/sandbox into Module:la-pronunc. — Eru·tuon 05:01, 6 October 2021 (UTC)

Crappy RQ: templates

Hi again. There's an error I've found a few times, undoubtedly always caused by my sloppy self, where an RQ: Template is used, including |passage=whatever, but the template doesn't have a |passage= bit in it, so that bit is hidden, much like my recent John Dryden quote in provoke. Could you whip up a list of such RQ: template errors at, say, User:Erutuon/WF is sloppy, and I'll get round to fixing them? Thanks in advance, as always. Your most humble servant etc. Roger the Rodger (talk) 21:29, 15 October 2021 (UTC)

Does this regex in the search box generate the list you seek?

template: intitle:/RQ\:/ -"passage"

HTH. DCDuring (talk) 00:56, 16 October 2021 (UTC)

Four More Flags

At the Language flags list, about four more flags were added. --Apisite (talk) 05:35, 5 December 2021 (UTC)

Now a flag for the Rakhine language (rki) has been to the list added. --Apisite (talk) 23:32, 10 December 2021 (UTC)

dashes

Hey list guy. Happy holidays and all. I made my Christmas wish list, and top of it was a list of any entries like ancestry with the classic Webster crap -Addison. at the end. I thought we'd caught them all before, but apparently not... Br00pVain (talk) 16:33, 24 December 2021 (UTC)

Perhaps a reboot of Wiktionary:Todo/Dash quotes would be doable? Br00pVain (talk) 15:41, 7 January 2022 (UTC)

How we will see unregistered users

Hi!

You get this message because you are an admin on a Wikimedia wiki.

When someone edits a Wikimedia wiki without being logged in today, we show their IP address. As you may already know, we will not be able to do this in the future. This is a decision by the Wikimedia Foundation Legal department, because norms and regulations for privacy online have changed.

Instead of the IP we will show a masked identity. You as an admin will still be able to access the IP. There will also be a new user right for those who need to see the full IPs of unregistered users to fight vandalism, harassment and spam without being admins. Patrollers will also see part of the IP even without this user right. We are also working on better tools to help.

If you have not seen it before, you can read more on Meta. If you want to make sure you don’t miss technical changes on the Wikimedia wikis, you can subscribe to the weekly technical newsletter.

We have two suggested ways this identity could work. We would appreciate your feedback on which way you think would work best for you and your wiki, now and in the future. You can let us know on the talk page. You can write in your language. The suggestions were posted in October and we will decide after 17 January.

Thank you. /Johan (WMF)

18:14, 4 January 2022 (UTC)

Happy 2022

Erutuon! Lots of wishes for a Happy -and healthy!- 2022. Many thanks from el.wikt for your help all these years! — This unsigned comment was added by Sarri.greek (talk • contribs).

delete + merge

I was going to speedy Template:R:bn:ODBL, a redirect of Template:R:ODBL that’s no longer used in any pages, but I discovered that the edit histories of both of these need to be merged— the edit histories are divided uglily due to a bad move in the past. Please fix the edit history and then delete Template:R:bn:ODBL! Thanks in advance. ·~ dictátor·mundꟾ 20:08, 12 January 2022 (UTC)

@Inqilābī: Only the first edit could be saved without rigmarole because the pages were created around the same time. But it's done. — Eru·tuon 20:23, 20 January 2022 (UTC)

French = English doesn't exist

Any chance you could get a list of all terms with an English section with a French translation that is exactly the same, but is missing a French section? To find things like conceptualisation, which is missing a French section. Br00pVain (talk) 00:45, 20 January 2022 (UTC)

ToilBot and IPA templates

Hey, I was wondering, would it be possible to include the automatic insertion of an asterix (if absent) before plain IPA templates in ToilBot's monthly routine? It seems that this is a pretty frequent mistake that is a pain to fix by hand if a user forgot to do this. Thadh (talk) 00:48, 20 January 2022 (UTC)

Problem with your "multitrans"

See diff. Apparently the translation adder uses "subst:" to create the language name, and your code doesn't know what to do with that. Chuck Entz (talk) 06:57, 20 January 2022 (UTC)

@Chuck Entz: I hadn't thought of that. Unfortunately there's no way to substitute a template inside of a nowiki tag, so the translation adder needs to get language names some other way. For now at least the module expands that module invocation correctly. — Eru·tuon 20:07, 20 January 2022 (UTC)

Hi Erutuon and Chuck Entz. I belong to a group of people who have run into another problem with that template. The wiktextract script that creates a machine-readable version of the Wiktionary data gets confused by it – especially since it wraps the <nowiki> tag around what's very clearly wiki data.

Another user had removed your template, thinking it not really needed in the first place, but Chuck Entz has reverted that change to wolf with the comment that the script prevents an "out of memory" error. I don't fully get that – as far as I can see, there was no such problem even on the version of the page that uses the standard multitrans template.

Can you explain what's going on here? Is your non-standard template really needed in the main namespace? If so, would it maybe be possible to find a solution that works without the `<nowiki>` tag? Krissie (talk) 18:55, 10 March 2022 (UTC)

@Krissie: Hi, very nice project. Unfortunately we can't get rid of the nowiki tag inside {{#invoke:User:Erutuon/multitrans|expand|...}}. {{#invoke:User:Erutuon/multitrans|expand|...}} is to reduce Lua memory usage. For every module invocation ({{#invoke:}}) a certain amount of extra memory is used to create the Lua environment, to re-execute required modules, to wrap data with mw.loadData, and so on. (This extra memory is because Scribunto can't allow some data to be shared between module invocations without transferring actual information that would change the wikitext generated by the module invocations.) There is an {{#invoke:}} transcluded inside many of the templates in translation sections (most commonly in {{t}} and {{t+}}). If these module invocations are executed in a single invoke, the extra memory usage is removed. {{#invoke:User:Erutuon/multitrans|expand|...}} replaces multiple instances of {{#invoke:}} with a single {{#invoke:}}. The only way to prevent the module invocations inside {{#invoke:User:Erutuon/multitrans|expand|...}} from being evaluated individually, and using the extra Lua memory, is by wrapping all of them in a nowiki tag and then un-nowikifying the tag in the Lua module that implements it. If we remove the nowiki tag, the templates are evaluated individually and the extra Lua memory is used. Another technique to reduce memory, in {{multitrans}}, was to use templates like {{tt}} and {{tt+}} that generate fake template syntax and pass that to a module function, but {{#invoke:User:Erutuon/multitrans|expand|...}} is easier to use (because you don't have to remember to use different template names) and ends up using less Lua memory (perhaps because it can reduce memory for all templates). These are both ugly workarounds, but they are essential on certain pages to remove Lua memory error messages.

I have a translation search website, enwikt-translations, that also has to deal with the nowiki tag inside of {{#invoke:User:Erutuon/multitrans|expand|...}}. When visiting the value of the first parameter of multitrans templates, the database generator finds nowiki tags and evaluates the templates inside the nowiki tag just like templates outside nowiki tags. I haven't looked at the wiktextract code in detail, but it could probably do something equivalent to ensure contents of the nowiki tag inside of multitrans are treated as if they were wikitext. — Eru·tuon 19:28, 10 March 2022 (UTC)

I see, thanks for the explanation. Though I admit I don't quite get it. The standard multitrans template is called with |data= and Wiktextract has no problems with that. It's only your specific version of the template, called with #invoke:User:Erutuon/multitrans|expand| followed by <nowiki> that causes the parser to fail, because it doesn't expect such usage of the nowiki tag. Why isn't the normal multitrans|data= invocation sufficient in these cases? (This specific syntax is or was used only in 4 or 5 pages in the main namespace, as far as we could see.)

(I'm not really involved in the Wiktextract project, BTW, except as a happy user. But I agree that it's very nice!) Krissie (talk) 12:58, 11 March 2022 (UTC)

Sorry, what I said above applies to {{#invoke:User:Erutuon/multitrans|expand|...}}, not to {{multitrans|data=...}}. I'd forgotten the difference and relied only on my memory. I edited my comment to correct it.

Yeah, I'd misunderstood at first and thought you were a maintainer of the project. I see you mentioned this conversation on the issue about this. The raw module invocation made it look like a temporary test, but we're planning to keep it in some form, I think. I should assign it a template name and fix the translation adder problem. — Eru·tuon 21:47, 11 March 2022 (UTC)

Now the template {{multitrans-nowiki}} replaces {{#invoke:User:Erutuon/multitrans|expand|...}}. — Eru·tuon 22:36, 11 March 2022 (UTC)

Category:unsigned with nonstandard timestamp

1. What is the point of having this category?
2. What makes the timestamp nonstandard?
3. How can I use {{unsigned}} without adding to the category?

DCDuring (talk) 17:44, 23 January 2022 (UTC)

@DCDuring: As I understand it, every use of {{unsigned}} adds a category. If your timestamp is in the standard format (HH:MM, Day Month YYYY (UTC)", with HH going from "00" to "24" and MM from 00 to 59, Day from 1 to 31, and Month from "January" to "December"- note that the only comma is after the time), it will add Category:unsigned with standard timestamp. The gadget that converts displayed dates and times to a user's timezone can only convert the standard format, but otherwise there's not much point in distinguishing between standard and nonstandard. I believe the "nonstandard" category has been nominated for deletion. Chuck Entz (talk) 20:06, 23 January 2022 (UTC)

So, to truly "correct" the timestamp, I would have to know the UTC differential, which I could not do for IPs and could conveniently do for registered users only if they provided that info on their user page (or if I were a checkuser)? DCDuring (talk) 20:12, 23 January 2022 (UTC)

@DCDuring: Actually, the information is there in the revision history. The dates and times there are all adjusted to one timezone- UTC without the time-adjustment gadget. If you have the time-adjustment gadget active like I do, you have to learn to mentally adjust the times given there. For me in PST, that means adding 8 hours (7 during DST). For you in EST, that would be 5/4. To illustrate: the wikitext for your message says "20:12, 23 January 2022 (UTC)", but on the page I see it as "12:12 pm, Today (UTC−8)". In the revision history, I see it as "12:12, January 23, 2022‎", but I know that the gadget has changed it from "20:12, 23 January 2022". If you have the gadget enabled, you would see it as "15:12, January 23, 2022‎", or "3:12 pm, Today (UTC−5)". If you have the gadget enabled, getting the format right will cause the date to magically change in preview to be consistant with what you see in the revision history. By the way, I fixed the unsigned templates in the Grease Pit, and the date you put was consistent with your timezone (and 5 hours off from the standard UTC), which means you do have the gadget enabled. Chuck Entz (talk) 21:45, 23 January 2022 (UTC)

We have 963 items in the category but only some 4,011 transclusions of {{unsigned}}. That is suggestive of a usability problem.

Am I correct in saying that all one has to do when adding {{unsigned}} is copy from the history, but reverse the username/IP address and the time/date? DCDuring (talk) 22:13, 23 January 2022 (UTC)

If you make sure that the time is in UTC and you follow every picky little detail of the format, yes. As I outlined above, both of those are rather tricky. If you forget and put a comma in the date, or you leave out the space before "(UTC)" or the parentheses, it won't accept it. If you use an adjusted date/time without correcting it, your time will be 5 hours early. In less than 2 hours, it will be tomorrow, so the date will be wrong, too, for the 5 hours it takes your timezone to catch up with UTC. It took me a while before I could constently get the timestamp right, even after I realized what was going on. Chuck Entz (talk) 22:36, 23 January 2022 (UTC)

• I'll play with it until I get it right. DCDuring (talk) 23:07, 23 January 2022 (UTC)
  • @DCDuring, Chuck Entz: As it sounds like you both don’t know it: You don’t have to mentally adjust anything. As the documentation of {{unsigned}} says, enable the gadget “⚓” in your preferences. The only thing one has to copy is preformatted and unveiled by clicking the pencil sign near the page title saying “difference between revisions” on any diff or near the page-title in any archived version. The “title” technically apparently being the h1 HTML tag with that gadget presumably adding the symbols via a CSS ::after Selector. Fay Freak (talk) 00:23, 24 January 2022 (UTC)

    If 24% of the uses of {{unsigned}} are wrong, despite the availability of the gadget and documentation, there IS a usability problem. DCDuring (talk) DCDuring (talk) 01:16, 24 January 2022 (UTC)

Help needed

Hello Erutuon! Simple English Wiktionary needs your help. I can't seem to fix the issue raise (last para) and would like you to take a look. Minorax (talk) 09:42, 15 February 2022 (UTC)

Wrong namespace

The entries Slava Ukraini! & Heroiam slava! should be merged with the non-exclamation forms. An anon has wrongly created both entries. Thanks. ·~ dictátor·