Anthere-
- activity shows interest, so avoid missed launches such as the french
wikinews one
Actually, it doesn't. French Wikinews would have passed the edit count requirements (haven't checked user duration); as you yourself said, Greudin is a very active user on fr.wikipedia and has pledged support for the French Wikinews, yet he has only made a handful of edits there.
Translating/creating policies seems to be a much better test of actual interest in doing work. Once you do that, that shows a commitment to the project.
- past activity of at least 2 editors on a wikipedia (for example)
indicates that at least 2 editors are aware of our basic principles and in particular NPOV requirement.
What better test could there be for people understanding a policy like NPOV than requiring them to translate it?
Again, the is a security measure. If 5 people, not even one oldby on one of our project, decide to launch a wikinews with no experience at all, there is rather high risk that some of our principles are not respected;
I don't see it that way. Just because someone has been on Wikipedia for months doesn't mean that they respect policies at all. Quantity is not quality, and measuring quality is almost impossible while keeping the process scalable and fair. One could even argue that malicious trolls or otherwise harmful users who know how to manipulate policies in their interest are more likely to come from our existing user base. In fact, Wikinews will especially attract people who are fed up with Wikipedia and want to work on another wiki.
and since it is not in a language we necessarily manage, it might go on for a long time.
This is more likely if key policies like NPOV are *not* translated. If we can agree on which parts of our policies are not negotiable, we can make sure that they are in place. One of these policies can even include instructions on what to do if your wiki doesn't follow the Wikimedia spirit (contact stewards etc.).
Future projects don't necessarily match our current userbase. To tie the process for creating new language editions directly to that userbase seems needlessly restrictive. Building a small community on Meta and writing key pages before launching the project is also simply good planning -- exactly the kind of thing that could have helped to prevent the current fr.wikinews.org situation, much more so than algorithmic requirements whose actual predictive value is very low, as that experience has shown.
Regards,
Erik