Today I want to add some points to ongoing discussion about localized spins in Fedora Project. As a member of translation team, I think there is no need to create such spins. In translation of any packages there are two surfaces that every localizers can work on. The first one is the translation of GUI. As my experience shows that great majority of users prefer to use the source language rather than their localized GUI. This case is very special for RTL languages, like Persian and Balochi. Because the users based on their previous experience with LTR interface are not familiar with this new interface. He can spend sometimes on RTL environment for example here, then when he surfs in internet, he uses LTR web pages. So the translated UI is not as favorable as we expect. For example, I myself get confused as I use a translated UI, because I don’t get accustomed a RTL interface at all.
The next item is translation of documentations. This is the point we should emphasis in localizing. A lot of Fedora users are not English native speakers or have difficulties in understanding English. As they get into trouble in using Fedora they are unable to find any local help, I mean in their computers. Localized docs can be useful here.
Shall we have a localized spin? frankly I should say that NO. As I check in Transifex in Fedora I find that we have 89 languages. If the Board authorized creation of a language-based spin for one language or two, the other languages that don’t have any spin up to that time may request a new spin. One day we have a lot of spins. I think it’s better to have a plan in development process to have some sort of modifications in coding for example in Anaconda. As a user select his region, for example, Anaconda suggest usage of localized version. developers can think about installing of selective packages based on users locale, for example.