Keith M Wesolowski writes: > The best opportunity for compromise was offered by the Companion > consolidation, which could have been used as a vehicle for unfettered > aggregation of arbitrary (and arbitrarily toxic and badly-integrated) > software, while the other consolidations continued using strict > quality-focused processes. But that opportunity has been lost and the > push to make everything look more like what the Companion used to be > suggests that many believe this disagreement has been resolved in > favour of the populist position that more putbacks = better. It would > be useful to know whether that's the position of the ARC and the > C-teams.
The ARC is inanimate (not sure about the C-team ;-}) and doesn't have opinions, but as one ARC member, I agree with everything you've said. Architectural review itself supposes that we're still building a system and that we need constraints in order to get the result to look as much as possible like a single work rather than an accumulation of artifacts. If we're ham-stringing the review because we've decided that we don't really want a single system anymore, then that's just plain silly. Instead of that, we ought to redo the rules so they match what 'we' want. I'm wildly in favor of making as much other software available on OpenSolaris as possible, and making it as easy to get as possible. There are many components of doing that job _well_ that cost a great deal: committing fixes upstream, participating in other communities, reworking parts that don't integrate well with our native mechanisms, and setting up well-mirrored sites with clear integration requirements, to mention a few. (I don't see how we're actually paying attention to those costs, which is a potentially concerning situation for the future.) I also (separately) like having a repository of pre-compiled stuff even if the only work done on it is putting it in package form. I'm a big fan and user of blastwave, even when they won't take my contributions (:-/). Having yet another repository that perhaps doesn't have a two-drink-minimum or has different component reuse guidelines might also be nice. However, as an ARC member, and thus as a fan of integrating software, I don't think "more" is always "better." The world already has a bunch of Linux distributions, and I don't see how we'd help by contributing another. Assuming that Companion == John Plocher's other Universe == extra stuff we don't really _need_ in SFW and don't plan to align with OpenSolaris architecture, I'd vote to bring back the no-ARC Companion CD rather than lard up SFW with partly-reviewed projects. -- James Carlson, Solaris Networking <james.d.carlson at sun.com> Sun Microsystems / 35 Network Drive 71.232W Vox +1 781 442 2084 MS UBUR02-212 / Burlington MA 01803-2757 42.496N Fax +1 781 442 1677
