On Wed, Oct 01, 2008 at 05:40:12PM -0500, Shawn Walker wrote: > Nicolas Williams wrote: > >On Wed, Oct 01, 2008 at 04:47:55PM -0500, Shawn Walker wrote: > >>Having an OS support more than one-language on a single medium is fairly > >>common, and the cost of maintaining per-language media is likely to be > >>prohibitive. That's probably why most vendors have adopted a "Language > >>Pack" approach from what I've seen. > > > >Well, the cost of keeping install image size down is not zero either. > > > >Can the cost of each of the two possible activities be quantified, or > >estimated? If it ever comes to where nothing else can be sacrificed but > >the multitude of locales, then we'll know for sure which is cheaper :/ > > Either direction has to be proven as optimal, but my guess based on
Has that been done? > observation is that it has proven for others to be more cost-effective > to use language packs instead of per-language media approaches. I've not observed what others have done and, more importantly, why they've done things this way. So I might have to take your guess for it. But note that the fact that "every Linux distro" (or whatever) has multiple locales on their install CD doesn't mean that their reasons for choosing to do so apply to us. > >>As such, I think a per-language media approach is not an optimal one. > > > >Has anyone tried? Perhaps it's easier to automate than you make it > >sound. > > Even if it could be automated, there is still manual testing involved, > management of multiple images, additional resources and bandwidth > required, etc. Manual testing wouldn't necessarily be much harder. You'd have to burn more CDs, but you still have to boot N times for N install locales. Or you could use VirtualBox, in which case there's no need to burn CDs. Nico -- _______________________________________________ indiana-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/indiana-discuss
