On 10/5/07, Dave Miner <Dave.Miner at sun.com> wrote: > The problem with releasing patchadd/pdo is precisely that changing the > Nevada code doesn't get you all that close to something that will be > useful to you or any other Solaris 10 customer. Back-porting to Solaris > 10 is something that has to be done by Sun staff, it tends to be fairly > time-consuming, and we just plain don't have the bandwidth in the > engineering and sustaining teams to commit to taking contributions and > back-porting them. Since we can't make that commitment, I can't > classify releasing that code as anything other than "attractive nuisance". > > (FWIW, pdo.c is a swamp. I'm doing you a favor. Really.)
And this was exactly what I worried about when I heard all sorts of people from Sun saying that Solaris 10 was open source. I really hope that over time Sun can *really* open source Solaris[1] so that the bits that paying customers use can be fixed by paying customers. In this case back-porting would not be an issue. Presumably rules would need to be in place that it isn't considered for Solaris until the related fix is in OpenSolaris. 1. There will likely always be some encumbered stuff. But "it's so ugly it will turn you to stone" shouldn't be a sufficient reason to keep it closed. Arguably, this is even more of a reason to have it opened. I don't raise this issue expecting there will a fix for anything I complain about today. I raise it so that sometime over the next 2 - 5 years things maybe get better with input from Sun's customers that are also contributors to OpenSolaris. -- Mike Gerdts http://mgerdts.blogspot.com/
