>> Admittedly rPath's Wiki is not the most user-friendliest one. So I >> could invest a few hours to write a summary myself. >> > > > Sure, why not. It is always good to know what's going on around us. > I would definitely like to read something like this. > > But then from the practical point of view I think it is a way too late for > Sun to suddenly switch to something else like conary. So disputing IPS vs > conary to convince Sun to change their approach is just a waste of time. > > But then nothing stops you from creating a new project at open Solaris (or > not if you wish) and try to build a distribution based on conary. Since you > seem to infer that it is mature and feature-rich product you should be able > to get there relatively quickly, especially if you would write some tool co > conver ips manifests to conary manifests (or whatever they are using). And > then who knows, everything might happen :) > > > -- > Robert Milkowski > http://milek.blogspot.com
sounds fair :) ok, I'm no longer trying to convince Sun to drop IPS. That was 1.5 years ago. But I could indeed finish the long promised conary based distro. (after updating my old stuff to match a current Distro Constructor checkout). Plus: My thought was, that parts of conary could just serve as a reference for IPS. I write up that summary. Probably I learn a lot more about IPS's actual implementation then, by directly comparing the two's Python code and the 2's C-backend and pluggable db backend (sqlite, conary also supports other db's). Martin _______________________________________________ pkg-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
