Wiktionary:Grease pit/2020/March
Bot cleanup request
[edit]We have 120 Swedish entries using ad hoc tables from 2013. Can someone replace them with {{col3}}
? Ultimateria (talk) 02:25, 2 March 2020 (UTC)
This was NOT vandalism, like the filter told me. — This unsigned comment was added by Victionarier (talk • contribs) at 22:00, 2 March 2020 (UTC).
- Yeah, the entry looked correctly formatted to me. However, the correct page for this content is Reconstruction:Proto-West Germanic/wīpan because Frankish has been merged into Proto-West Germanic by Wiktionary:Votes/2020-01/Make Frankish an etymology-only variant of Proto-West Germanic. — Eru·tuon 02:57, 3 March 2020 (UTC)
Permission error trying to create page "أغروم"
[edit]Trying to create the page أغروم (Algerian Arabic for bread), I get this weird permission error:
- You do not have permission to create this page, for the following reason:
- The title "أﻏﺮوم" has been banned from creation. It matches the following blacklist entry: .*[\x{FB50}-\x{FDC7}\x{FE70}-\x{FEFC}].*
Seems like something went wrong with the regular expression... פֿינצטערניש (Fintsternish), she/her (talk) 12:09, 7 March 2020 (UTC)
- @פֿינצטערניש: You've got two different sequences of code points there, though they look identical when rendered in a font and font engine with Arabic letter joining. أﻏﺮوم is bad (it contains the Arabic Presentation Forms that are blacklisted, U+FECF ARABIC LETTER GHAIN INITIAL FORM ﻏ and U+FEAE ARABIC LETTER REH FINAL FORM ﺮ), أغروم is good (it contains U+063A ARABIC LETTER GHAIN غ and U+0631 ARABIC LETTER REH ر) and you can edit it without getting in trouble with the title blacklist. — Eru·tuon 16:26, 7 March 2020 (UTC)
Not enough memory
[edit]The page a, o and mi seems not to have enough memory to store the word in all languages. StrongestStrike (talk) 02:31, 11 March 2020 (UTC)
- It's not memory for storage, but memory for programming done by templates. Unfortunately, there are too many languages with "o" and "a" and the like as a term (a has 114 of them)- we only have 50 MB for the whole page, and it can take a MB just for the headword template. It seems like at least a year that we've had these un-fixable module errors. See CAT:E for the rest of them. Chuck Entz (talk) 03:07, 11 March 2020 (UTC)
- I've done a bit more research and left new comment on phab:T188492. I think at this point we don't have any other choice than to have the memory limit raised. – Jberkel 22:08, 14 March 2020 (UTC)
I am an old lady
[edit]I am an old lady, who knows nothing about computers, scripts and Lua, and who does some greek edits here and at el.wiktionary, where I have been trying to unify and update some declension modules. The two Lua creators of el.wiktionary are gone for some years now. Erutuon has already helped with crucial corrections and I wouldn't wish to bother him once more -thank you Eru!!-. I can see how busy you all are but if you ever have time: I cannot make one stem (stem["3"]) work at el:Module:zz. It is a patchwork of copypastes and I am very embarassed to show it to experts. Please do not bother too much. If it is something easy, just a little tip to the right direction would be very helpufl. If it is tricky, I can always use the old modules. Thank you. ‑‑Sarri.greek ♫ | 17:41, 13 March 2020 (UTC)
- Done, thank you! ‑‑Sarri.greek ♫ | 01:41, 21 March 2020 (UTC)
Status of “definition popups”?
[edit]Can we get Navigation Popups to show the definitions of a word, as at it:abbaside, where I, with Navigation Popups enabled, get definitions when hovering over the links Abbasidi, pittura (multiple definitions and even a picture) and sillibazione (while, incidentally, aggetivo and relativo merely show “m sing”)? I for one would find this very helpful indeed, and I imagine I am far from alone.
There was a Grease pit discussion in April 2012 about this; it seems that back then it could be done with pretty heavy hacking. Our documentation of our version of Navigation Popups says nothing about this (or about anything, really). This would ideally just happen when Navigation Popups were enabled, and also in page previews, if they exist apart from Navigation Popups.
PJTraill (talk) 19:44, 14 March 2020 (UTC)
- It broke many years ago and nobody technically knowledgeable has gotten around to fixing it. @Yair rand might know how. —Μετάknowledgediscuss/deeds 19:51, 14 March 2020 (UTC)
- Thanks, maybe I shall take a look myself, as it is rather a sad state of affairs – but it would be a lot faster if someone knowledgable were to do it. PJTraill (talk) 21:23, 14 March 2020 (UTC)
- So, the version on Wikipedia that the gadget imports has since been changed to encapsulate its variables, so that other scripts can't modify them. (Normally good practice, but since we're trying to irresponsibly mess around with things, it's inconvenient.) If one uses the old versions of the WP gadget, and then imports Connel MacKenzie's script that messes with it, then it works. Any objections to forking the old version (or a slightly modified version of the current) onto MediaWiki:Gadget-popups.js and re-adding the other script? Pinging @Dixtosa:, who removed it in 2017 after things started breaking. --Yair rand (talk) 06:18, 17 March 2020 (UTC)
- I have no objections to anything that actually works! —Μετάknowledgediscuss/deeds 23:16, 18 March 2020 (UTC)
- +1. - -sche (discuss) 06:07, 19 March 2020 (UTC)
- @Yair rand: Thanks for your action, which seems to have made this function far more useful. I am not sure when it works (not always! e.g. bathing suit but not birthday suit, which starts with Alternative forms, Etymology, Pronunciation), but I am seeing much more relevant content than before. PJTraill (talk) 10:49, 6 April 2020 (UTC)
- I have no objections to anything that actually works! —Μετάknowledgediscuss/deeds 23:16, 18 March 2020 (UTC)
Lua error: not enough memory at "se" page
[edit]Apparently Lua is running out of memory on the se page (already at the Slovene inflections) which prevents me from making use of the Swedish entry. --RayZa (talk) 00:23, 15 March 2020 (UTC)
- A very kludgy workaround is to click Edit and then Show preview for just the Swedish section, which only uses a fraction of the memory. Chuck Entz (talk) 03:19, 15 March 2020 (UTC)
A new Bulgarian reference template
[edit]Can someone make a new new Bulgarian reference template so that a parameter is passed to the end of this URL: [1], e.g. to look up the word лъх (lǎh): [2]. Calling @Benwing2, this dictionary is more informative, IMO, than [3]. --Anatoli T. (обсудить/вклад) 23:44, 15 March 2020 (UTC)
- I've made one myself -
{{R:IBE}}
. --Anatoli T. (обсудить/вклад) 06:34, 17 March 2020 (UTC)
There's nothing really remarkable about this entry: no translation table, only two language sections, and a descendants list with 14 languages. Previewing the Ancient Greek section shows 22.99 MB of memory used, while the Greek shows 8.67 MB. Commenting out both sections to leave only {{also|βαρβάρα|barbaros|bárbaros|Barbaros}}
shows 870 KB.
And yet, after I added the entry to the {{redlink category}}
exclusion list to clear an out of memory error, previewing the whole entry shows 47.27 MB. Commenting out {{also|βαρβάρα|barbaros|bárbaros|Barbaros}}
at the top causes that to jump to 49.08 MB!
As I always do, I checked for changes in any of the transcluded modules, and these are the only ones with revisions in the past few days: [4], [5], and [6]- all very boring, routine edits that don't seem to be related. The only wildcard I can see is User:Isomorphyc's inscrutable sharded data modules, but those are in most Ancient Greek entries, and I've never seen an out of memory error in an Ancient Greek entry before.
Something very strange is going on here. There's absolutely nothing to explain the 14+ MB difference between the memory usage of all the parts as previewed and that of the entry as a whole. I'm worried that this kind of thing may cause more 32MB entries to suddenly run out of memory- and we've got lots of those. Chuck Entz (talk) 03:11, 16 March 2020 (UTC)
Alternative spelling not alternative spellings
[edit]Is it possible to generate a page of all entries containing "alternative spelling of" which down use {{alternative spelling of}}
? To be put at Wiktionary:Todo/untemplatized alternative spellings. I'd imagine we have a few dozen. 150 tops. --Alsowalks (talk) 18:26, 16 March 2020 (UTC)
Automatic elsewhere
[edit]I remember we have a page somewhere, possibly off-site, which shows the entries which appear in the most other-language Wiktionaries but not en.wiktionary, like User:DTLHS/elsewhere but better, as it is autoupdated. Where was that list? It should appear at User talk:DTLHS/elsewhere too, so I don't forget about it. --Alsowalks (talk) 18:35, 16 March 2020 (UTC)
- I miss you, not literally, that's actually its name. --Vriullop (talk) 07:51, 17 March 2020 (UTC)
Templates seem to generate more-or-less random tracking categories
[edit]E.g. {{trans-top}}
at Orwellian. The current revision of this page has translations in a little over a dozen languages and of them, Arabic, Chinese/Mandarin, Danish, Dutch, Greek, Norwegian (Bokmål and Nynorsk), Occitan, Persian, Swedish, and Turkish are all redlinks but the only hidden categories generated by the template are Category:Mandarin redlinks and Category:Mandarin redlinks/t. Why not Category:Occitan redlinks? Is this deliberate or an oversite? If it's deliberate, what is the rationale? I have noticed similar behavior in other instances but this one caught my attention today. Thanks. —Justin (koavf)❤T☮C☺M☯ 19:03, 18 March 2020 (UTC)
- The rationale is that it's expensive and only certain languages are enabled. DTLHS (talk) 19:06, 18 March 2020 (UTC)
- I understand it's not all that hard to turn the tracking on (or off) if one desires to work on blueing redlinks. DCDuring (talk) 19:13, 18 March 2020 (UTC)
- I believe there's now a bot that periodically generates a list of red links. So tracking them live may no longer be needed. —Rua (mew) 19:16, 18 March 2020 (UTC)
- @Rua: Do you propose 1.) removing all these tracking categories, 2.) including all of them, or 3.) something else? Seems like if a bot makes the lists, we could easily remove the categories. —Justin (koavf)❤T☮C☺M☯ 05:44, 20 March 2020 (UTC)
- I believe there's now a bot that periodically generates a list of red links. So tracking them live may no longer be needed. —Rua (mew) 19:16, 18 March 2020 (UTC)
- I understand it's not all that hard to turn the tracking on (or off) if one desires to work on blueing redlinks. DCDuring (talk) 19:13, 18 March 2020 (UTC)
size of etyms
[edit]Looks like some Anglo-Norman (and Old French) mentions are in a font bigger than the rest. See reprisal#English or abash#English. Sobreira ►〓 (parlez) 21:30, 18 March 2020 (UTC)
- In my browser the Anglo-Norman and Old French words in both etymology sections are the same size as the rest of the text. — Eru·tuon 22:14, 18 March 2020 (UTC)
- @Sobreira: I have the same problem for terms in many language when they use whatever underlies
{{l}}
(also shared by many other templates). I have the problem in Firefox, but not in Chrome. What browser do you use? DCDuring (talk) 03:05, 20 March 2020 (UTC)
templates: prefix, pre
[edit]It seems that the templates prefix and pre are not categorising terms under categories of words that use that prefix. An example is foresight, which uses {{prefix|en|fore|sight}} in the etymology, but it doesn't place the term in Category:English words prefixed with fore-. Can someone please take a look and correct this ? Leasnam (talk) 06:44, 19 March 2020 (UTC)
- @Leasnam: I can't find any examples of this problem, and foresight is categorising correctly. In any case, you should probably be using
{{affix}}
instead. —Μετάknowledgediscuss/deeds 16:40, 19 March 2020 (UTC)- It appears to be working now. I noticed it first at offspring, and yes, I swapped prefix/pre for af/affix and got it working. Now I've switch it back to pre and it's working fine. Leasnam (talk) 18:45, 19 March 2020 (UTC)
- @Leasnam: The category reappeared after User:DTLHS reverted some edits by User:Benwing2, which somehow had removed the "prefixed with" category at foresight. — Eru·tuon 20:21, 19 March 2020 (UTC)
- Tracked the problem to the line
local first_category = table.remove(categories, 1)
. That removed the "words prefixed with fore-" category fromcategories
butfirst_category
was then not used. I've restored User:Benwing2's edits but with the offending line deleted. — Eru·tuon 20:35, 19 March 2020 (UTC)
- It appears to be working now. I noticed it first at offspring, and yes, I swapped prefix/pre for af/affix and got it working. Now I've switch it back to pre and it's working fine. Leasnam (talk) 18:45, 19 March 2020 (UTC)
Hello,
Are all those transliteration removals justified? I think that at least some of them are not, since a standardized way to transliterate do exist (for Indian scripts in particular). 176.147.224.55 14:16, 19 March 2020 (UTC)
- Some may not be. I think @Benwing2 was handling this. —Μετάknowledgediscuss/deeds 16:41, 19 March 2020 (UTC)
- @Metaknowledge Yuck. This guy has reappeared after we went through and cleaned up his previous changes. I can write a script to blanket revert all his changes, but the problem historically has been that he mixes good changes with bad ones. We need someone who knows the languages in question to review some of the changes and let us know if they all need to be reverted. Benwing2 (talk) 04:04, 20 March 2020 (UTC)
- @Benwing2: The good changes I see are unnecessary, so I think a blanket revert is the way to go. —Μετάknowledgediscuss/deeds 04:17, 20 March 2020 (UTC)
- @Metaknowledge Yuck. This guy has reappeared after we went through and cleaned up his previous changes. I can write a script to blanket revert all his changes, but the problem historically has been that he mixes good changes with bad ones. We need someone who knows the languages in question to review some of the changes and let us know if they all need to be reverted. Benwing2 (talk) 04:04, 20 March 2020 (UTC)
- (edit conflict) Are all those transliteration removals justified? It depends. I support removal of Burmese translations. The current standard is MLCTS but there could be whatever. There are very few cases where the manual is justified, if the spelling is abnormal, as with ယောက်ျား. Perhaps such cases where manual transliteration is required (even if an automated exists) for certain languages should be marked in comments. For Indic languages, like any Dravidian no manual is required.
- Hindi and other north Indian - should be case by case, the majority should be automated.
- Thai and Khmer - if terms are defined and use the pronunciation templates - the automatic transliteration is more accurate. Otherwise, the manual may be justified, e.g. Khmer ឆែកកូស្លូវ៉ាគី (chaekkouslouvaakii, “Czechoslovakia”) currently showing "chaekkouhluuvaakii" (auotomated) but should be "chaekkouslouvaakii". --Anatoli T. (обсудить/вклад) 04:28, 20 March 2020 (UTC)
- As for the Russian transliterations, as in стани́чник (staníčnik, staníšnik), the manual transliteration is required. --Anatoli T. (обсудить/вклад) 04:35, 20 March 2020 (UTC)
Reference error
[edit]Tried to cite the pronunciation I found for the page bolete and added the reference in the pronunciation section for the English entry, for some reason the URL popped up down in the Latin entry. I'm not sure how to fix this or what went wrong in the first place. Could someone let me know what to do do avoid this in the future? Thanks 2WR1 (talk) 14:23, 19 March 2020 (UTC)
- @2WR1: Fixed. You need to create a References section in the English section; it won't appear magically. Take a look at the source of bolete to see what I did. Also, please don't give references as bare URLs — we have lots of reference templates to format them, in this case
{{R:Merriam-Webster Online}}
. —Μετάknowledgediscuss/deeds 16:45, 19 March 2020 (UTC)- @metaknowledge Thanks so much for the help, I'll keep it in mind for the future! 2WR1 (talk) 04:02, 25 March 2020 (UTC)
I wrote up the article 'hīo' to create an entry for the variant of the Old English word 'hēo'. At the time I did not know how to add the article to the general list of headings under 'hio'. Having figured out how to do so, I have added it under said general listing and was attempting to get rid of my now spurious and superfluous creation. — This unsigned comment was added by 2602:306:CF6F:520:B931:1AE0:C075:58F2 (talk) at 20:20, 21 March 2020 (UTC).
- Deleted hīo. In future, you can just put
{{delete|reason}}
on the page to request a deletion in uncontroversial cases like this. — Eru·tuon 20:39, 21 March 2020 (UTC)
Bot Request- frk
[edit]15 hours ago, User:Mahagaja switched the language code frk (Frankish) from a full language to an etymology-only language. He spent 7 hours fixing the resulting module errors, and I've spent about 9 hours. After all of that, there are about 200 entries in CAT:E- and there are lots more that haven't shown up yet.
I would like to request for someone with a bot to fix a subset of these that are straightforward and don't need human input.
- To start with, change all instances of:
{{etyl|frk|en}} {{m|frk|
to:
{{der|en|frk|
You'll also need to add an * to the term linked to if it doesn't already have one
The same goes for the codes ca, enm, es, frm, frp, gl, it, nrf, oc, pt and wa, though the most important ones are en, fr, enm and frm.
These are all languages that cannot have either borrowed or inherited from Frankish.
- The other part is changing frk to gmw-pro in all instances of
{{m}}
and{{l}}
that aren't preceded by an etymology template with which they could be merged (and adding * as needed).
Since this is just for cutting the task down to size, you don't have to worry about edge cases- when in doubt, skip it.
Also see the discussion of this on Mahagaja's talk page. Chuck Entz (talk) 06:22, 23 March 2020 (UTC)
- @Chuck Entz I'll see about doing this. Benwing2 (talk) 01:04, 24 March 2020 (UTC)
- @Chuck Entz I take it that Old French *can* borrow from Frankish? If so I will fix up all the Old French uses of
{{etyl}}
(there are a lot of them), using my quick "manually-assisted" bot process (i.e. I am checking them manually to make sure the use of{{bor}}
is correct). Benwing2 (talk) 01:29, 24 March 2020 (UTC)- Also, some guidance on
{{bor}}
? Presumably an Old French verb cannot{{bor}}
from a Frankish noun (e.g. Frankish *fōdar -> Old French forragier)? What about cases like Frankish *sporo -> Old French esperon, or Frankish *frumjan -> Old French fornir, or Frankish *wala + *hlaupan -> galoper, or Frankish *haifst -> Old French hastif? I assume these should use{{der}}
as there is some reshaping or morphological addition going on. Benwing2 (talk) 01:40, 24 March 2020 (UTC)- I left Old French out of the list because I honestly don't know, and I didn't want to delay work on the obvious stuff during the time it took to find out. Chuck Entz (talk) 01:45, 24 March 2020 (UTC)
- (edit conflict) I wonder if we should rethink the restriction that disallows etymology-only languages like 'frk' in
{{m}}
and{{l}}
. Consider this, for example (from estoffer): Borrowed from {{bor|fro|goh|stoffōn}} or {{bor|fro|frk|*stopfōn}}, {{m|frk|*stoppōn}}, both from {{der|fro|gem-pro|*stuppōną||to stop, close, push, prick}}.
{{m|frk|*stoppōn}}
causes an error but is a very natural use of this template with this language code. Benwing2 (talk) 01:47, 24 March 2020 (UTC)- @Chuck Entz BTW I reduced the number from 429 -> 276 after applying the first part of your proposed fix (except for adding * where necessary, I'll get to that next). Benwing2 (talk) 01:49, 24 March 2020 (UTC)
- OK, I went ahead and modified
{{m}}
,{{l}}
,{{ll}}
and{{m+}}
to allow etymology languages and link appropriately to their non-etymology parent. Hope there are no objections to this. Benwing2 (talk) 02:46, 24 March 2020 (UTC)- If people agree with this, I can fix certain other templates like
{{t}}
and variants and{{affix}}
/{{compound}}
/etc. to work similarly. Benwing2 (talk) 02:56, 24 March 2020 (UTC)- I'm assuming this is temporary? —Rua (mew) 14:26, 24 March 2020 (UTC)
- I would be hesitant about allowing etymology-only languages in translations. At the very least we I think we should modify the
{{t}}
backend to disallow translations in unattested languages or anything else outside of mainspace. Chuck Entz (talk) 15:00, 24 March 2020 (UTC)- I agree that
{{t}}
shouldn't accept etymology-only languages, or reconstructed languages or non-mainspace conlangs either, for that matter. —Mahāgaja · talk 15:22, 24 March 2020 (UTC)- I haven't touched
{{t}}
or friends, only{{m}}
,{{l}}
,{{ll}}
and{{m+}}
. I'm not sure why we should disallow etymology-only languages for these basic linking templates; the average user won't know which languages are etymology-only and which ones aren't, and won't understand why a given etymology-only language is disallowed in these templates. We could "fix" this by issuing an error that says something like "'frk' is an etymology-only language and disallowed, please use 'gmw-pro' instead", but why bother when we can just as easily accept the etymology-only language? It also seems to me there's useful info in specifying an etymology-only language in{{l}}
or{{m}}
, it says the term is particular to that lect. Benwing2 (talk) 01:25, 25 March 2020 (UTC)- Interesting idea. This adds another step to the processing of these templates, including the generation of User:Jberkel/lists/wanted (pinging User:Jberkel), but that's easy to do if you already have code that handles etymology templates like
{{der}}
. In theory it will ease splitting etymology languages into full languages, in the rare cases when that happens: templates would not have to be updated because they would already use the etymology language code, if people use the feature. At the moment using an etymology language code has no effect on the template output, except in{{m+}}
. — Eru·tuon 23:03, 7 April 2020 (UTC)
- Interesting idea. This adds another step to the processing of these templates, including the generation of User:Jberkel/lists/wanted (pinging User:Jberkel), but that's easy to do if you already have code that handles etymology templates like
- I haven't touched
- I agree that
- If people agree with this, I can fix certain other templates like
- OK, I went ahead and modified
- (edit conflict) I wonder if we should rethink the restriction that disallows etymology-only languages like 'frk' in
- I left Old French out of the list because I honestly don't know, and I didn't want to delay work on the obvious stuff during the time it took to find out. Chuck Entz (talk) 01:45, 24 March 2020 (UTC)
- Also, some guidance on
- @Chuck Entz I take it that Old French *can* borrow from Frankish? If so I will fix up all the Old French uses of
{{t-simple}}
is not working as documented. terms with diacritics are linking to wrong entries. The 2nd parameter is actually gender not "The word to link to." What's the point of having all these complex modules and templates if a translation is replaced with this and it's wrong: Template:t-simple or Template:t-simple. Of course любо́ў and حُبّ don't exist but любоў and حب do. Take a look at love#Translations.
The regular templates: любо́ў f (ljubóŭ) and حُبّ (ar) m (ḥubb) work as expected. --Anatoli T. (обсудить/вклад) 07:49, 23 March 2020 (UTC)
- It is working as documented, but the point of
{{t-simple}}
is to reduce server load in various ways, including by not having diacritic stripping, automatic language name generation, or automatic transliteration. You type:
{{t-simple|be|любоў|alt=любо́ў|f|langname=Belarusian|tr=ljubóŭ}}
And here:
|1=
is the language code,be
|2=
is the word you want to link to, in its page name form. This should beлюбоў
without the diacritic.|alt=
is the form you want to display, with diacritics|3=
or|g=
is the gender|langname=
is the language name|tr=
is the transliteration
- So it works, but it's a lot less user-friendly than
{{t}}
. That's the price we pay for reducing server load and thus out-of-memory errors. —Mahāgaja · talk 08:39, 23 March 2020 (UTC)
- @Mahagaja: Thank you for the explanation. I have now moved translations at love#Translations for both noun and verb to love/translations. Not sure now if I can apply
|alt=
on all translations or use the regular{{t+}}
and{{t}}
templates. Whoever applied the{{t-simple}}
, did a poor a job by breaking all interwiki links and displaying existing entries in red. --00:24, 25 March 2020 (UTC)- @Atitarev: You didn’t get it. You are breaking the interwiki links yourself by not providing the page name in raw form to the template. The template is not supposed to recognize that diacritics have to be stripped, on the contrary it must not because the whole point is never to run these processes like deriving page names from spellings with diacritics because they cost computing resources. What you are right to observe however that this template is not simple for the user: it is the simple for the server. Fay Freak (talk) 00:41, 25 March 2020 (UTC)
- @Fay Freak: What? I did get it. Adding
{{t-simple}}
that way wasn't my doing. If someone ran a bot or did it manually, should have considered what the result was going to be. They should write Template:t-simple, instead of Template:t-simple.--Anatoli T. (обсудить/вклад) 00:47, 25 March 2020 (UTC)- There are still some messed-up links at school (Russian, Hebrew, Arabic …), so the IP in December failed the template too. Somebody might run a bot to do the following: In invocations of
{{t-simple}}
, check (with the typical suspect languages needing diacritics in dictionary forms) for links with diacritics and unspecified scripts, and fix or at least list the cases. There are not many pages using the template, otherwise I would suspect there are probably many IPs who do it wrongly and it would be a periodical bot job, but still by our own eyes we would not notice all and could make further wrong. Fay Freak (talk) 01:16, 25 March 2020 (UTC)- It's all over the place. I am fixing where I see where I can when I'm willing to. If the regular templates are so resource hungry, then maybe the
{{t-simple}}
should be used for new translations. A bot should be able to fix the current mess. BTW, don't jump to conclusions who breaks what next time. Hint: read the discussion carefully and check the history, like you did now. --Anatoli T. (обсудить/вклад) 01:34, 25 March 2020 (UTC)- I found it independently of what you wrote. So you actually may have pointed out – what you wrote is difficult to understand – that people use the template wrongly. But still the template works “as expected”, just that people do not work with it as expected. Fay Freak (talk) 01:56, 25 March 2020 (UTC)
- It's all over the place. I am fixing where I see where I can when I'm willing to. If the regular templates are so resource hungry, then maybe the
- There are still some messed-up links at school (Russian, Hebrew, Arabic …), so the IP in December failed the template too. Somebody might run a bot to do the following: In invocations of
- @Fay Freak: What? I did get it. Adding
- @Atitarev: You didn’t get it. You are breaking the interwiki links yourself by not providing the page name in raw form to the template. The template is not supposed to recognize that diacritics have to be stripped, on the contrary it must not because the whole point is never to run these processes like deriving page names from spellings with diacritics because they cost computing resources. What you are right to observe however that this template is not simple for the user: it is the simple for the server. Fay Freak (talk) 00:41, 25 March 2020 (UTC)
- @Mahagaja: Thank you for the explanation. I have now moved translations at love#Translations for both noun and verb to love/translations. Not sure now if I can apply
Descendants header
[edit]At hi, the "Descendants" header is not visible, despite the fact that it is on the page. Can someone please look into this ? Leasnam (talk) 15:49, 23 March 2020 (UTC)
- The problem is with
{{dum-decl-ppron}}
, which is @Rua's creation. —Μετάknowledgediscuss/deeds 16:56, 23 March 2020 (UTC)- I don't immediately see the problem in the template. Does anyone else? —Rua (mew) 17:00, 23 March 2020 (UTC)
- Removing the
float: left
onclass="NavFrame"
fixed it, for what it's worth. --{{victar|talk}}
17:13, 23 March 2020 (UTC)- Thanks, All ! Leasnam (talk) 01:25, 24 March 2020 (UTC)
- Removing the
- I don't immediately see the problem in the template. Does anyone else? —Rua (mew) 17:00, 23 March 2020 (UTC)
Categories for other then japanese Hani
[edit]@Rua, Shāntián Tàiláng, Tooironic, Eirikr:I think, that on en.wikt should be such categories as e. g. Category:Chinese terms spelled with 定, Category:Korean terms spelled with 定, Category:Vietnamese terms spelled with 定, Category:Cantonese terms spelled with 定, not only Category:Japanese terms spelled with 定 (I mean for all corresponding (=if exists in that language) Hani, not only 定); followed also by categories such as Category:Chinese terms spelled with 定 read as ding4/Category:Chinese terms spelled with 定 read as dìng or/and Category:Chinese terms spelled with 定 read as ㄉㄧㄥˋ. The former category should be categorized not only as is (for Japanese) but also to commom category: Category:Terms spelled with 定, gathering info about all languages, that use Hani 定. Any tips for that? Thank you. --Kusurija (talk) 17:19, 23 March 2020 (UTC)
Proto-West Germanic irregular verbs
[edit]Should templates for Proto-West Germanic irregular verbs be created? StrongestStrike (talk) 15:52, 24 March 2020 (UTC)
Something wrong with WOTD?
[edit]The yesterday/tomorrow links seem to be red links due to a doubled-up forward slash in the URLs. Equinox ◑ 19:51, 24 March 2020 (UTC)
Overline above large Roman numerals
[edit]Module:foreign numerals does not add an overline above large Roman numerals, showing a million as MMMMMMMMMMMMMMMMMMMMMMMMMMMMMMM... instead of M. This is a problem in Module:number list/data/la for the New Latin numeral millio (“million”). J3133 (talk) 12:04, 25 March 2020 (UTC)
- @J3133: I'm not familiar with this convention, but I've made the module emit numbers 5000 and over with overline diacritic (U+0305) and numbers 5,000,000 and over with a double overline diacritic (U+033F). I didn't like the idea of using inline CSS, though it's used in
{{overline}}
and Wikipedia articles like w:1000000, and though the overline and double overline diacritics won't join together when multiple numerals that use the same diacritic are in sequence. — Eru·tuon 05:11, 31 March 2020 (UTC)
Should categorise
[edit]At cymotrichous, and elsewhere, there is a cool Pronunciation bit, where the IPA is embedded in the audio file. It looks lovely, but doesn't categorise the entry into Category:English terms with IPA pronunciation. I'm sure a little fiddle with Template:audio-IPA or Template:audio-pron could fix that. You know what...those templates can probably be merged (along with Template:audio too, i guess) --Guiri Falso (talk) 15:21, 25 March 2020 (UTC)
- Displaying the IPA in the caption for the audio file is a bad idea, not only because the entry isn't categorized, but also because it makes showing variant pronunciations unwieldly. I'm going to standardize that entry now. —Mahāgaja · talk 16:48, 25 March 2020 (UTC)
Yiddish verb conjugation template should allow for alternate forms
[edit]With the Yiddish verb זײַן there are two forms each for the 2nd plural and 1st/3rd plural: זענט/זײַט and זענען/זײַנען. Unfortunately Template:yi-conj doesn't seem to allow for listing alternate forms, only for overriding the standard conjugation with one form. I would appreciate if someone could change this when they have the chance. Sorry if this is the wrong place to ask. Thank you. פֿינצטערניש (Fintsternish), she/her (talk) 11:45, 30 March 2020 (UTC)
Non left-to-right writing systems in page title
[edit]Hi everyone, I’m from the French Wiktionary and I have a question regarding the writing direction in the title. We just added this Manchu word and we have a problem with the page’s title. As you can see, it is written from left to right but we want it written as in this page, that is verticaly. How did you achieve this? I already looked into the Common.js page but didn’t find anything related. Is it some kind of template? Thanks in advance Darmo117 (talk) 22:26, 30 March 2020 (UTC)
- @Darmo117: It's done through headword-line templates and MediaWiki:Common.css. Module:headword adds a display title (via
{{head}}
and other headword-line templates) that applies the script classMong
to the header, and the common CSS file applies the ruleto that class. The display title in ᠪᡳᡨ᠌ᡥᡝ (added by.Mong { /* ... */ -webkit-writing-mode: vertical-lr; -moz-writing-mode: vertical-lr; writing-mode: vertical-lr; layout-flow: vertical-ideographic; }
{{head|mnc|noun}}
) is equivalent to{{DISPLAYTITLE:<span class="Mong">ᠪᡳᡨ᠌ᡥᡝ</span>}}
so the CSS rule makes the top header display vertically. — Eru·tuon 22:44, 30 March 2020 (UTC)- @Erutuon, I was wondering: at el. we do it with el:Module:title. But in the Category they do not appear vertical. Is there a trick we can add to make them vertical in the Category without the other modules (head, etc) involved? ‑‑Sarri.greek ♫ | 07:25, 31 March 2020 (UTC)
- Thanks a lot to both of you, I’ll look into it
Darmo117 (talk) 10:29, 31 March 2020 (UTC)
- Thanks a lot to both of you, I’ll look into it
- @Erutuon, I was wondering: at el. we do it with el:Module:title. But in the Category they do not appear vertical. Is there a trick we can add to make them vertical in the Category without the other modules (head, etc) involved? ‑‑Sarri.greek ♫ | 07:25, 31 March 2020 (UTC)
New suggestion for category tree language pages
[edit]Hi! I am from Turkish Wiktionary, and recently I have adapted the view design of the language categories in a responsive way. Therefore, we now have a decent page when these categories are visited on the mobile website. You can try one by entering this page on your mobile phone: tr:Kategori:İngilizce. We could adapt the same styles here? ~ Z (m) 20:52, 31 March 2020 (UTC)
Correct usage of template:wikipedia
[edit]I was considering adding a wikipedia link to https://en.wikipedia.org/wiki/Oni on the page for oni, and possibly to 鬼#Japanese. In order to figure out the correct usage of the template, I looked at the source of 山, but I only ended up confused.
The Chinese section does this
{{zh-wp|zh|yue|lzh|gan|hak:sân|cdo:săng|nan:soaⁿ|wuu}}
which looks like this:
The Japanese section does this
<div style="float:right;"> {{wp|lang=ja}} {{wp|Mountain}} </div>
which looks like this:
and also puts it in the Etymology 1 subheading, which appears to be wrong.
Why is the Chinese version subtly different? The icon is smaller and multiple languages are listed in a single box, while the other has a separate box for each. Also, in all cases except the English, the link shows up on hover as https://en.wikipedia.org/wiki/{lang}:{term}. This redirects to https://{lang}.wikipedia.org/wiki/{term}, but why doesn't the link point there to begin with? 71.168.173.2 20:54, 1 April 2020 (UTC)
- Chinese is different because a user made a Wikipedia template tailored for inclusion of multiple Chinese lects. —Suzukaze-c◇◇ 08:27, 24 April 2020 (UTC)