Le 10/08/2014 10:27, svetlana a écrit :
I feel that having development carried out by "employees" hinders programming the same software as a hobby: for instance, they work in a single language, and don't need localised documentation
We have developers across all continents with a wide range of native languages. We just fallback to a common language which happens to be English.
If the computer industry started in France, we would probably all be using french to communicate and that would become a de facto requirement. Keep in mind that all languages are based on English and most of their documentation / source code comments are also in English.
I definitely understand how it is a barrier of entry for a hobbyist developer, but lack of English skill is a different problem which is out of scope of our mission. The way we mitigate it is that we have developers fluent in a bunch of languages and, I at least, enjoy helping/replying/guiding developers in our native language.
As for our documentation, the user/sysadmin guides are on mediawiki.org. That has been done for a reason: let folks document MediaWiki instead of the overbusy developers. That also opens possibilities to at least translate the user/sysadmin guides.
Regarding the documentation shipped inside Mediawiki core, it is based on source code comments and I don't see us maintaining a translation for them. We don't even properly maintain our doc as-is!
there are other (which i haven't shaped properly yet) differences of architecture of comminity-run tech projects and tech projects run by employees, which make getting involved as a hobby harder
for instance, i could not make a difference to a big linux distro run by a corporation (or using one as an upstream)
which is why I'm not very supportive of any plans that involve more employees at WMF Engineering either
I hope this way to put it is slightly more clear than it was before
I am not sure how it applies to MediaWiki development process. We have a fair share of volunteers involved, and more than a handful of them that have very deep knowledge about wiki and propose very useful code. There are some area of codes I would ask volunteers to review, I trust them more than myself or other employees to handle the review and make sure it is not going to break the Wikimedia cluster.
For beginners developers, the barrier entry is much higher. I agree. But I think it is more related to how complicated MediaWiki is than some localization issues or WMF having paid employees.