[Bf-committers] Qdune code
joeedh at gmail.com
Fri Nov 9 16:44:59 CET 2007
Alexander Ewering wrote:
> On Thu, 8 Nov 2007, Joe Eagar wrote:
>> Hay! That's not true. I for one have spent maybe horrible hours trying
>> to get DSM ready since the end of SoC. And I'll be spending many
>> horrible hours more.
> That's of course very laudable, and I fear we would get off-topic with this
> discussion :)
> On the whole, I think the project deadline criteria in GSoC were not defined
> well enough. There should have been a "No fully working inclusion in trunk,
> no money" policy. Because as it is, dozens of thousands of dollars have been
> spent this and last year on coding hours (no matter by whom) which were lost
> in nirvana. And that's not always the pupils' fault.
> Alexander Ewering
Well in my case it was decided to change the project goals halfway
through. I implemented exactly what I originally meant to a month into
GSoC. The time since has been spent mostly tracking down obscure bugs
in the calculus-leaden full DSM implementation it was decided to do instead.
Also I have to disagree on such a policy; it'd encourage students to
abandon projects if they feel theres no way to meet the deadline, even
if they spend a great deal of time on 'em. This year several of the
GSoC projects did get pretty much working on-time (mine before goal
changes, the audio one, GLSL and the bevel one I believe). None were
acutally merged into trunk, which is weird. Certainly GLSL, bevel and
audio seem to be relatively finished.
More information about the Bf-committers