I think any decision to split wikis should not be made without consensus within the affected communities. From a technical standpoint, it would be much easier to decide in favour of a split, but I'm worried that the social ramifications of such a decision would be overlooked in a discussion among developers. It might be a good idea to present the challenges of continuing to maintain Language Converter to the communities, but if, for example, it turns out that having a single Chinese language is better for our free knowledge goals, then it might still be worth investing the time needed to overcome the technical hurdles.
Asaf, wctaiwan: I totally understand the social and cultural implications -- that is indeed one of the key things I wish to discuss during this session so that people understand that language variants are not *just* a technical issue.
For example, I believe that Urdu and Hindi are also amendable to LanguageConverter, but the political issues in that case make it highly unlikely that the two languages would wish to share a wiki.
I think Asaf's question was well-posed: let's see what the technical issues are, so that the community can have an informed discussion. --scott
On 27 January 2015 at 10:32, C. Scott Ananian cananian@wikimedia.org wrote:
Asaf, wctaiwan: I totally understand the social and cultural implications -- that is indeed one of the key things I wish to discuss during this session so that people understand that language variants are not *just* a technical issue.
For precisely this reason, I don't think it would be an appropriate conversation to have at a technical event like MWDS.
J.
Well, I'll give James the opposite answer then:
Language Variants are not *just* a social and cultural issue, either. There are real technical issues to address, having to do with how we manage "slightly forked" wikis, how we maintain code which is not used by enwiki, and the technical limitations of the three basic approaches which have been attempted to date (Content Translation tools, Language Converter, forked wikis with manual synchronization).
So, let's talk about the technical issues at the Developer summit. I'll show you the code to support LanguageConverter in Parsoid, let's talk about what it would take to make this work in Visual Editor and/or for HTML page views (targeting mobile performance).
And then we can discuss the highly technical RFC regarding "Glossary" support in core mediawiki: https://www.mediawiki.org/wiki/Requests_for_comment/Scoped_language_converte...
I think there is plenty of technical content to discuss. And non-technical issues to keep in mind, and should inform the technical discussion.
But if you don't think it's worth discussing, then don't come. ;) There are three other different talks at the same time... --scott
On 27 January 2015 at 10:49, C. Scott Ananian cananian@wikimedia.org wrote:
Well, I'll give James the opposite answer then:
Language Variants are not *just* a social and cultural issue, either. There are real technical issues to address, having to do with how we manage "slightly forked" wikis, how we maintain code which is not used by enwiki, and the technical limitations of the three basic approaches which have been attempted to date (Content Translation tools, Language Converter, forked wikis with manual synchronization).
So, let's talk about the technical issues at the Developer summit. I'll show you the code to support LanguageConverter in Parsoid, let's talk about what it would take to make this work in Visual Editor and/or for HTML page views (targeting mobile performance).
And then we can discuss the highly technical RFC regarding "Glossary" support in core mediawiki:
https://www.mediawiki.org/wiki/Requests_for_comment/Scoped_language_converte...
I think there is plenty of technical content to discuss. And non-technical issues to keep in mind, and should inform the technical discussion.
Absolutely. I didn't say anything about the technical issues, which do need some serious discussion (though I think that an ad-hoc session might be too short notice).
But if you don't think it's worth discussing, then don't come. ;) There are three other different talks at the same time...
I think it's totally worth discussing, but indeed, unfortunately I committed to another session. :-(
J.
I would love to go to this but the timing is not good. Like James, already committed to another session.
There seem to be a lot of language-related issues that are being worked on independently by different groups of people who don’t necessarily communicate. Hard for a newbie like me to wrap my head around it all. What could we do to be a bit more coordinated in our efforts? (Seriously, I’m asking.)
Joel Sahleen, Software Engineer Language Engineering Wikimedia Foundation jsahleen@wikimedia.org
On Jan 27, 2015, at 11:07 AM, James Forrester jforrester@wikimedia.org wrote:
On 27 January 2015 at 10:49, C. Scott Ananian cananian@wikimedia.org wrote:
Well, I'll give James the opposite answer then:
Language Variants are not *just* a social and cultural issue, either. There are real technical issues to address, having to do with how we manage "slightly forked" wikis, how we maintain code which is not used by enwiki, and the technical limitations of the three basic approaches which have been attempted to date (Content Translation tools, Language Converter, forked wikis with manual synchronization).
So, let's talk about the technical issues at the Developer summit. I'll show you the code to support LanguageConverter in Parsoid, let's talk about what it would take to make this work in Visual Editor and/or for HTML page views (targeting mobile performance).
And then we can discuss the highly technical RFC regarding "Glossary" support in core mediawiki:
https://www.mediawiki.org/wiki/Requests_for_comment/Scoped_language_converte...
I think there is plenty of technical content to discuss. And non-technical issues to keep in mind, and should inform the technical discussion.
Absolutely. I didn't say anything about the technical issues, which do need some serious discussion (though I think that an ad-hoc session might be too short notice).
But if you don't think it's worth discussing, then don't come. ;) There are three other different talks at the same time...
I think it's totally worth discussing, but indeed, unfortunately I committed to another session. :-(
J.
James D. Forrester Product Manager, Editing Wikimedia Foundation, Inc.
jforrester@wikimedia.org | @jdforrester _______________________________________________ Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
wikitech-l@lists.wikimedia.org