Just a couple of comments.

First, Have you guys been thinking about how multiple transifex implementations can work together to enhance the same translation codebases?

Let’s say that a relatively large project runs its own implementation and has its own translation community, and then a distributor like Fedora integrates that project. Do you have a best practises suggestion on how two translation communities can use separate instances of transifex to work on the same upstream codebase?

Second, I’d like to see some sort of high level vision document about how the transifex development teams sees how the translation community fits in in the rest of the open source ecosystem. In my mind, translation as a broad global community overlays over the project space in a completely different way than how developers interact.

It seems to me, collaborative code development processes breaks up traditional lines on a map and draws new lines based on learned skillsets, like computer language comprehension. Each computer language is its own little nation of sorts, so there is a break down in traditional cultural boundaries but replaced by new virtual boundaries. Boundaries with open borders, but easily discernible boundaries none-the-less. I would imagine that the nature of the translation work is somewhat different in this regard and aligns far more with traditional map boundaries in terms of team building (as the nature of human language diversification is tied to regional culture. ) And at the same time translation teams need to interact with developers from pretty much all the little virtual development “nations”. How do we overlay the map of the translation communities over the map of the development communities as efficiently as possible? But that’s me rambling. I have no idea what I’m talking about, I’m not doing the translations. But you guys are, which is why I’d like to see what you think about that.

-jef