On Thu, Dec 22, 2011 at 7:55 PM, Christophe Fergeau <cfergeau@xxxxxxxxxx> wrote: > On Thu, Dec 22, 2011 at 07:33:22PM +0200, Zeeshan Ali (Khattak) wrote: >> The patch improves the situation as it makes the whole API very >> consistent w.r.t what exactly is the namespace here. > > Imo the namespace really is GVir::Config, not a GVirConfig namespace > totally separate from the GVir namespace, so it does not make the whole API > "very consistent", it just changes things. Even if the namespace is 'GVir::Config', my assertion that 'Config' is part of the namespace and not just a symbol prefix is pretty much correct so I don't understand what you are trying to discuss here. >> I also >> agree that nested namespaces will be better. If we decide/manage to go >> towards nested namespaces, this patch actually helps in that regard as >> well since existing API is not consistent/correct for that purpose >> either. > > It helps *but breaks every library user*. Which user? Currently the API is hardly used by 2 apps. Keeping in mind that library is still at its infancy and missing a lot of essential API, this shouldn't be a concern at all since breaking things now is preferable to breaking it later when apps really depend on it and we promise some stability. > Which is why you have to > carefully weight the pros and cons. It makes things slightly nicer, > slightly more consistent but *it breaks every user*. This is what makes it > special and worth more considerations than a quick ack while everyone is on > holidays. I disagree and think you sometimes worry way too much. :) > Really, let's just wait until the holidays are over, as far as I'm > concerned I wouldn't like having such a patch go in before I get a chance > to see it even if I agree with it. Thats what I am going to do if there is no ACK for it anyway. -- Regards, Zeeshan Ali (Khattak) FSF member#5124 -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list