Dans T210114#8646160, @Nikki a écrit :

Namespace 118 is excluded because it's the "Draft" namespace in some Wikipedias, but looking at the namespace definitions, namespace 118 is the "Reconstruction" namespace for enwiktionary, urwiktionary and zhwiktionary, and the "2014" namespace on wikimaniawiki. The places where 118 is the "Draft" namespace are all Wikipedias, except for hiwikiversity (and testwiki), so I think the config needs to check whether the current wiki is a Wikipedia or Wikiversity before excluding namespace 118.

Could you explain waht it is marked as declined?

I would add one more argument in favour of Vega: the conversion of existing graph code would be much easier using Vega than a new library because all existing code already use Vega syntax.

Dans T348203#9287126, @SCherukuwada a écrit :

Partnerships just told me they will reach out to Google soon. We'll post updates here.

Today, the French Wiktionary link uses Wikipedia as title. The French Wikipedia link uses a weird logo. And the English Wikipedia looks correct...

Lingua Libre Bot has gotten the global bot status.

I submitted a feedback to Google 10 days ago but nothing has changed since then.

A verification has been added in Lingua Libre Bot for the Odia Wiktionary (point 2 in the list).

I am not sure this is related to this ticket but a Lingua Libre user is not able to translate the interface into the Pa'O language because this language is not available in the list of languages. See https://lingualibre.org/wiki/User_talk:%E5%92%BD%E9%A0%AD%E3%81%B9%E3%81%95#Add_Pa.27O_language_and_Tai_Ahom_language

You cannot do this yourself. It is not currently supported by the Record Wizard. As Lucas said, this feature needs to be developed in the Record Wizard. If you want to contribute, you can propose a patch on Github :)

Brahmi script looks also to be blacklisted.
Please, allow filename using the Brahmi script.

The same for the Tyap Wiktionary: a̱sa̱khwot shows interwiki links but the same entry on the French Wiktionary does not show the link to the Tyap Wiktionary.

OK, I will check Lingua Libre Bot and once the "new language code system" is supported I will run the bot on the French Wiktionary only for thos languages.

@XANA000 does Lingua Libre Bot now add pronunciations in the languages you have listed above?

@LucasWerkmeister and @Nikki have made a lot of work on this topic recently. A bit more information on Commons.

The newly created Wiktionary in Sorani (code ckb) has the same behaviour: Cognate is working on the Sorani Wiktionary but the other Wiktionaries do not see the Wiktionary in Sorani. For example water links to the other Wiktionaries but the French Wiktionary does not list the entry on the Wiktionary in Sorani.

I reverted because one should list only the "main" language code. Aliases are for Wiktionary internal purpose and must not be used by LLbot. If we really want to add them in the Lingua Libre Wikibase, then we should mark these values as deprecated.

So I go ahead and I have created a new property. See an example here. I think we can use "Wikimedia language code" property to avoid creating a new one.

Not using the Lingua Libre wikibase anymore? Why not, but I don't think it's going to happen tomorrow. By the way, are there any public discussions (a Phabricator ticket or something) where you can read about the state of thinking on the subject?
So in the meantime, I think adding a property to the Lingua Libre wikibase is not such a bad idea.

Dans T329568#8613730, @Lepticed7 a écrit :

On the wikidata element of the calabrais, maybe we can add a property like "identifier in Wiktionary", with the value "calabrais centro-méridional" and a qualifier "wiki" with the value "French Wiktionary".

This query gives currently 46 Lingua Libre languages that do not have ISO 639-3 code.

Hmmm, not sure if the Wikidata community will accept such a property that is very specific. I mean I do not know how to write a property proposal because the only need for now is for Lingua Libre Bot.

Dans T329568#8613533, @Lepticed7 a écrit :

Hi. Is it not possible on wikidata, with a specific property?

This is rather important topic because currently minor languages without ISO 639-3 code are not supported by Lingua Libre Bot and thus recordings are not added into Wikimedia projects.

The WAV specification indicates (page 23) in the INFO List Chunk section that metadata can be stored in a WAV file among predefined chunks ("keywords"). The specification says also "New chunks may be defined, but an application should ignore any chunk it doesn't understand". So in principle, we can define our own metadata and add them in the WAV file. Obviously, we need to document the new chunks somewhere. See alo https://www.robotplanet.dk/audio/wav_meta_data/

I have written a bash script that loops over all languages supported in Lingua Libre and request all recordings for each language. Then, Lingua LIbre Bot runs on a given wiki project to add all missing recordings. For languages with a large number of recordings (for example French), I split the resquest year by year.

Actually, Cognate links work on the Gungbe Wiktionary. For example osin on the English Wiktionary proposes a link to the Gungbe Wiktionary. And osin on the Gungbe Wiktionary also proposes links to the other Wiktionary versions.

Dans T321666#8504825, @Pamputt a écrit :

I just discover that there is an issue with Cognate and the newly created Wiktionary projects. Two examples:

• • On the Igbo Wiktionary: the entry agba links to all other Wiktionary version, but the same page on the other Wiktionary versions does not link to the Igbo page (for example, here is the French page that does not contain any link to the Igbo Wiktionary)
• On the Gorontalo Wiktionary: the page taluhu links to all other Wiktionary version, but the same page on the other Wiktionary versions does not link to the Gorontalo page (for example, here is the French page that does not contain any link to the Gorontalo Wiktionary)

I did not check if there is already a Phabricator ticket for this problem (if not I will create one) but it may be related to the problem described in this ticket.

I just discover that there is an issue with Cognate and the newly created Wiktionary projects. Two examples:

Is there any plan for the data engineers to work on that topic?
It is enable for Wikipedia since a while and [https://stats.wikimedia.org/#/fr.wiktionary.org/contributing/active-editors-by-country/normal|map|last-month|(activity-level)~5..99-edits|monthly stille nothing] for the Wiktionary projects.

@Yug could you create anew ticket to ask to implement that in the extension repository and tag it as feature rrequest. Otherwise, I fear it will be forgotten soon.

What I understand is to allow double licence, not to set default licences: CC by sa + CC0. If you want this dual licence to be the default, could you elaborate why it should be the default instead of other licence (CC0 only, CC by-sa only, etc.)?

See the Technical board.

Hello, it is said that the new page is updated automatically.
If so, how are chosen the languages of the Wiktionaries that are displayed at the bottom? For example, if I am not wrong, the Tacawit Wiktionary (shy.wiktionary.org) that has about 3500 entries currently is not listed.

The error message "Error saving translation: You have been blocked from editing." is indeed correct but does not explain at all why one is block from editing. I would expect a clearer message. But if you consider there is no issue and I am alone to complain about it, then no problem.

The French Wiktionary is also intested in such feature also to manage links in translation templates.

I maybe experience the same issue.

Dans T239272#6046005, @0x010C a écrit :

@Pamputt I'm not sure the "linked user" property would be usefull on Commons, imho the LL property P5 "speaker" is more relevant. A property P823 (speaker)" exists on Wikidata, but it only accepts items. Thus, is it relevant to create wikidata items for each speakers?

A new Wikidata property related to Lingua Libre audio files has been created. See https://www.wikidata.org/wiki/Wikidata:Property_proposal/Lingua_Libre_ID : P10369

See also https://github.com/lingua-libre/CommonsDownloadTool/issues/2

I tried to run create_datasets.sh. The problem occurs when the script is at the "Download and package ALL sounds..." step.

I do not know whether the server has already been updated but it seems the new syntax is still not known by MediaWiki. See https://lingualibre.org/index.php?diff=683554 for example. Maybe @Eihel knows a bit more. At https://www.mediawiki.org/wiki/Help:Extension:Translate/Page_translation_administration#Segmentation, it is written that the new syntax is supported since Translate version 2021.04 but on https://lingualibre.org/wiki/Special:Version, I see that Translate version 2020-04-20 is installed.

See also https://lingualibre.org/wiki/LinguaLibre:Chat_room#.5BFeature_Request.5D_Play_next_sound_automatically_while_checking_recordings

See also: https://lingualibre.org/wiki/LinguaLibre:Chat_room#commons%3Acommons%3Astructured_data

There was a bug in LLbot (introduced by myself) and that has been fixed. It should not happen anymore. But let this task opens until we have removed all the pronunciations in doublon.

It appears that CommonsDownloadTool needs to be fixed to solved this issue. If @Poslovitch is not wrong, the problem comes from the fact there are too many files to get from Wikimedia Commons in one time and there is a ratelimit that is currently not supported by CommonsDownloadTool and which prevents the download to success.

@Noe, @WikiLucas00 where can wwe find some documentation about the DDF SPARQL? For example, what are the prefixes and so on? In short, how to write a DDF query?

Because it concerns only the French language, would it be possible to develop a generator gadget instead so that it can be enable only by French contributors?

@Seb35 if this commit this commit is taken into account, do you know why the Vector skin is still available on Lingua Libre?

I opened a pull request concerning frwiktionary.

Dans T276013#7261602, @Eihel a écrit :

I also started a discussion between the place and the country in the "écouter" template, example: Paris (France). So, there would be consequences from LiLi. @Pamputt

Bon, bah je me suis lancé. J'ai ouvert une discussion sur la Wikidémie

I think the Wikidémie is good. If there are a lot of discussions, we will start a Prise de décision, but I think we can avoid it.

Agreeds with all what you propose. Do you want to start a discussion on the French Wiktionary to ask to add such parameters in the écouter template? LLbot should be able to add this information for each Wiktionary that is interested in; I do not see any issue about that.

I have opened a pull request on Github to select and display the place of learning when it exists instead of the place of residence. I let @Poslovitch review it and merge it if ok.