[Foundation-l] IYL'08: Moratorium on deleting language projects?

Fri Jan 4 20:46:01 UTC 2008

Hoi Gerard,

I think there are two issues to address. First, different groups of small
language projects obviously have very different problems, and I am not
qualified to discuss all of them. I know better about the problems of the
WPs of the smaller languages of former Soviet Union. They typically have
several hundred thousand to several millions native speakers, and all
these speakers are also fluent in Russian. Part of the motivation of
editors to create these WPs is to save the language from extinction and
encourage the native speakers who are living in big cities and do not have
contact with other speakers to preserve the language and to edit
Wikipedia. In some cases, the collection of texts in WP is the biggest in
the language everywhere in the internet. It always works, sometimes
quicker, sometimes slower. But even for projects with several thousand
articles it is usually difficult to find more than a dozen regular
editors. It is particularly very unlikely that a project would grow in
such a way in the Incubator - no potential editor is going to find it, it
does not get media attention etc. The question is whether we would like to
encourage such use of Wikipedia, or we barely put red tape by setting some
restriction: minimally five different editors every month, several hundred
articles etc. And as far as I am concerned this is the question to be
answered by WMF, possibly with a preceding discussion in the community.
Some policy on the issue must be established by WMF.

Another issue, whatever policy on encouraging/discouraging smaller
language projects is established, is transparency. I think you guys in the
Language Subcommittee are doing very good job, and I had pleasant
interactions with some of you on Meta and Incubator, but the decisions
come (or not come) as a bolt from the blue. Is the committee elected? When
you say that projects with less than 100 articles are dead, is it your own
opinion, opinion of Language Subcommittee, policy of Language Subcommittee
or policy of WMF? Why are the policies for opening new projects not
written anywhere down, change without notice and are applied
retroactively? How does it come that LS has the entire responsibility for
opening new projects but has no relation whatsoever to existing projects,
even their closure? How could it be that at the same time as the closure
of Lak WP is discussed, a temporary admin status is not granted since the
project is active enough to have a permanent admin? And I guess many
people who are regulars on Meta and Incubator can ask dozens of questions
like this. As I said I really respect what you guys are doing, and I
understand that it is absolutely necessary, but I am pretty sure everybody
would welcome more transparency and more written rules concerning project
localization.

Cheers,
Yaroslav

> Hoi,
> The projects that are being closed are not being closed to any particular
> policy. That is bad in and of itself. What is good is that we at least
> have
> an understanding that projects can go to the Incubator to revive them.
> Project with no localisation, with less then 100 articles are hardly
> relevant on any scale. The language committee has no dealing with the
> pre-existing projects, this is in many ways a mixed blessing.
>
> The new projects for a new language have at least minimal localisation,
> what
> is considered minimal is determined by the people at BetaWiki; they are
> the
> most relevant messages for our READERS. For subsequent projects we insist
> on
> a full localisation of MediaWiki and this policy is bearing fruit; the
> MediaWiki languages is particularly good for languages that want to start
> a
> new project. A good example is Sranan Tongo that only recently was given
> conditional approval and started localisation and is doing well.
>
> When people vote for a particular project to start, they do not realise
> that
> their vote is quite meaningless. People are in favour or against often for
> political reasons. The only thing they do is create a stir. People who
> walk
> the walk and talk the talk make the difference. People who create credible
> articles in the Incubator, people that do the localisation. People that
> create a presence for their language once the project is approved.
>
> Projects are approved and they are sometimes for languages that are quite
> small. It takes a few dedicated people to start a new language and, it
> takes
> dedication, prolonged dedication to make a successful project. We are
> HAPPY
> to approve new languages and projects and, I do want to make the
> International Year of Languages a success by making sure that the
> localisation of MediaWiki is a cornerstone to what makes a Wikipedia
> relevant combined with a minimum of well written articles.
>
> If you want to have Wikipedias for African languages, then people that are
> literate in those languages have to show their interest. Wikipedia is a
> written project. When people want localisation, we can now help them by
> creating .po files for MediaWiki. This allows for the use of tools like
> Computer Aided Translation tools. Open Progress has financially supported
> the Wikiread functionality for OmegaT (a GPL licensed CAT tool). We hope
> to
> get you Wikiwrite as well so that you can both read and write MediaWiki
> articles from withing OmegaT.
>
> Stopping the closure of WMF projects is currently not in the cards. There
> are people who do not appreciate the relevance of supporting under
> resourced
> languages and are really aggressive. The only sane thing is to be prudent
> in
> approving new projects. As to the differentiation of incubator projects,
> this is effectively already the case for Wikibooks. However, for a
> separate
> language version of Wikibooks, approval is required.
>
> Where you suggest stronger requirements when a language already exists,
> this
> is already the case. For a follow up project the localisation has to be
> complete. This means that the Turkish request for a Wikiversity will wait
> until this requirement is met. (The Turkish language projects are lively
> and
> not problematic). In essence the suggestions are already in place.
>
> Thanks,
>     GerardM
>