Hi guys! On Fri, Feb 25, 2011 at 5:05 PM, Colin Guthrie <gmane at colin.guthr.ie> wrote: > 'Twas brillig, and Tanu Kaskinen at 25/02/11 15:44 did gyre and gimble: > Are you happy with the patch other than this additional comment in the > commit message? Thanks Tanu for giving me my share of credits :) my 2c, although it's been almost two years since I started this patch, I remember I was convince that it was the right approach, in my mind, it simplified the volume computation in "complex graph" cases (and also some weird crazy volume computation loops when you have 3 sinks in Y shape that should really just be the same volume in fact - that's why it was call flat-sink originally). But I never managed (or had time really) to properly put those reasons into words, and convince Lennart, with whom I discussed it, but there is no written trace of that discussion, I don't think it was even submitted on the ML. So, although it looks weird, this patch (or it's version on N900) is fairly safe and solve some real use cases. There is no obvious way to improve it. If it can help Nokia and other people to get this in (instead of having to maintain it elsewhere), I would say go for it, yes! regards -- Marc-Andr? Lureau
