First, thanks everyone for support and ideas!
On Tue, Apr 23, 2013 at 12:02 AM, Fred Kiefer <address@hidden> wrote:
Hi Ivan
I had long ago decided that I wont mentor for GSoC this year, but your
mail led me to rethink that position. Your proposal is very convincing and
would be highly beneficial for GNUstep. ...
Sorry, but I wont be able to step in here. Hopefully somebody else in the
GNUstep community will.
Thanks!
Best wishes and hope to see you in Cambridge,
Ditto!
On Tue, Apr 23, 2013 at 9:31 AM, David Chisnall <address@hidden> wrote:
On 22 Apr 2013, at 21:43, Ivan Vučica <address@hidden> wrote:
Ping! Student registrations have started. Any prospective mentors?
I've registered as a mentor. I'd also be most interested in the back end
being refactored to support using Opal.
Based on feedback everyone gave, yes, the importance of this (and the
importance of integration with Core Animation) seems to be the consensus.
Currently, basic idea that I have is to copy the entire Cairo backend (or
maybe use derived classes, although long-term that seems a bad idea) and
ensure that every Cairo surface is instead backed with an Opal context. I'd
temporarily expose the Cairo surface from the Opal context, and ensure that
calls are directed to the Cairo surface backing the Opal context instead of
to the Cairo surface inside the backend.
Then, I'd be removing the Cairo surface from the backend and replacing
Cairo calls with Opal calls.
Finally, I would be continuously looking at options for integrating Core
Animation with AppKit (based on feedback).
I'll probably need more supervision compared to implementation of UIKit,
but this is definitely a more useful long-term project for GNUstep.
I'll put together a proposal, register it on GSoC's Melange, and also send
it to the mailing list.