On further thought, UTF-8 could be recognized as a protocol. It does identify a standard for various forms of data transfer (file, input, network). Perhaps something along the lines of "protocol::encoding-utf-8". However, according to the tagging guidelines everything should theoretically support Unicode.
> works-with::text:unicode > > Please do not tag programs with simple unicode support, doing so would > make this tag useless, but only if the editor has special support for > it, like code tables etc. > > Ultimately, all applications should have unicode support. Perhaps then, only packages that either add Unicode support to an application or are otherwise built specifically to address Unicode should be tagged with a "works-with::unicode"? Note that the "text" part of that tag is excessively abstract. A tool that addresses support for displaying UTF-8 encoded names from MP3 tags isn't going to occur to somebody as a "works-with::text". a solid, present example of such is the utf8-migration-tool package which is a tool for migrating a system to the UTF-8 locale. Though we generally think of Unicode as a "text encoding", Unicode effects much more than what users would commonly refer to as "text". This topic also has me thinking about another similar fundamental change, IPv6. Sure, everything is going to eventually support IPv6, but support is currently very limited. Do we currently identify all IPv6-supporting packages by labeling them with "protocol::ipv6" and eventually phase out the tags when ipv6-support becomes mainstream? Also, what about the non-Unicode supporting packages. How would I go about identifying that I don't want to browse packages known not to work with Unicode? How will I go about identifying that I don't want to browse packages known not to work with IPv6 when it achieves the same state? Regards, Tyln Sylverwind _______________________________________________ Debtags-devel mailing list [email protected] http://lists.alioth.debian.org/mailman/listinfo/debtags-devel
