On Fri, Jul 11, 2008 at 1:10 PM, Ankitkumar Rameshchandra Patel ankit@redhat.com wrote:

Thanks for taking an initiative to define the criteria for language support. However, I am not sure whether the term (Language Support) used over here is correct or not. Because any package or feature of the application included in Fedora have any specific criteria for *support*. The criteria defined on the above URL could be called as a *Language Inclusion Criteria* in terms of I18n and L10n both instead of *Language Support Criteria*. Anyways, FLP team would be happy to define the Language Inclusion Criteria in terms of L10n I think.

I could be wrong here, please correct me if I am. Because as far as I know, for Fedora, "Support" means voluntary help from community contributors through mailing lists, forums or IRC. That's it, I think! No specific criteria for it.

IMHO, 'inclusion' does not sound an appropriate word here. For example, it would mean that the listed criteria be met even before the inclusion process starts. Whereas the purpose here is to have a reference to check after the inclusion process is completed.

The targeted problem here is that, how do we crosscheck if a particular language is actually a part of fedora. The problem with l10n/i18n work is that it is scattered through too many packages and sub-components of the system. This makes it difficult to identify if the language is usable or not on a given system. This also adds ambiguity for developers to make the language go through a _feature's_ process. Since a 'new language' in fedora is a significant achievement for user experience and amount of work for l10n and i18n teams, it makes sense to identify it through the features documentation and even possibly the 'release notes'. Thus, if we propose a new language for a particular release as a feature we have to ensure that a third party (anyone beyond i18n/l10n team) has a set of checkpoints to verify the completeness of the feature. This has been the main reason to document the minimum criteria for support.

Now, I agree that the term 'support' may have varied interpretations depending upon the context. For fedora, I think it should mostly be from community perspective. The same term has been used by gnome as well. The objective is not-to-discourage particular activity on a particular language and yet to identify few others that have good status of overall work and usability. This helps developers to set a milestone and users to set expectations. The same could be true for fedora. (We have already used it in past but more vaguely.)

This is not proposed as a criteria for 'commercial' support, but I think the same can be derived on similar lines if needed.

Thanks,