June 14 Gecko 1.9/Gran Paradiso Planning Meeting

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view

June 14 Gecko 1.9/Gran Paradiso Planning Meeting

Vladimir Vukicevic-3

The next Gecko 1.9/Gran Paradiso meeting will be this upcoming Wednesday,
at 11am.  This meeting was originally going to be every other week, but there
is enough to do that it will be held weekly.  The agenda is at http://wiki.mozilla.org/Firefox3/StatusMeetings/2006-06-14 
-- please add any items that need discussion there.  Also, before the meeting,
please update the Gecko 1.9 Feature List at http://wiki.mozilla.org/Firefox3/Gecko_Feature_List|Gecko
with any items that are not yet on that list.

Meeting Details
* Wednesday, June 14, 11:00 am PDT (18:00 UTC)
* Mozilla HQ, 1st floor conference table
* +1 866 216 2181 (or +1 303 228 0514)
* join irc.mozilla.org #granparadiso for back channel

    - Vlad

dev-planning mailing list
[hidden email]
Reply | Threaded
Open this post in threaded view

Re: June 14 Gecko 1.9/Gran Paradiso Planning Meeting

correct adress:


dev-planning mailing list
[hidden email]
Reply | Threaded
Open this post in threaded view

Re: June 14 Gecko 1.9/Gran Paradiso Planning Meeting

Darin Fisher-2
In reply to this post by Vladimir Vukicevic-3
Here are my notes from the meeting:

Gecko 1.9 meeting - 6/14/2006

attendance: bryner, bsmedberg, beltzner, sicking, dveditz, ispiked,
dbaron, schrep, mrbkap, sherman, pavlov, rhelmer, vlad, mconnor, darin

Gecko 1.9 Alpha (not a Firefox or Gran Paradiso release):
  - places: in or out - AI: talk to brett
  - this release does not include dbaron's reflow branch
  - firefox on xulrunner?  depends on build engineering, installation work

Firefox 3 schedule?
 - does a firefox 3 alpha1 occur shortly after firefox 2 final?
 - Firefox 3 beta 1 in january
 - mconnor: places needs to be in an alpha shippable state by october
 - schrep: concerned about state of gecko trunk (stability, content
regressions, etc.)

Prioritize features:
= cairo gfx (p1)
= cocoa widgets (p1)
= fix units in gecko "means: real content zooming" (p2)
 - dbaron: this is something that can be persisted, while text zoom cannot
 - mconnor: probably a p2
 - pavlov: additional work required on top of Eli's patch
 - dbaron: extra work is probably reasonable
= view removal (p1)
 - necessary to fix regressions
 - unclear what roc's timeframe is for this work
 - absolute positioning across XUL and HTML... dbaron: small patch,
but it hurts performance by 10%... need to figure it out.
 - hoist plugins to toplevel children is included in this task
= text frame rewrite (p2)
 - pavlov: big win for performance, but it's a big job
 - schrep: is this the lowest hanging fruit?
 - vlad: yes, probably, but it'll be a lot of work
 - pavlov: make stuff work better (word breaking, justification, etc.)
= frame display lists (p1)
 - vlad: depends on view removal
 - dbaron: goal is to fix a bunch of bugs
= reflow branch (p1)
 - dbaron: goals: simplify code and fix a bunch of incremental reflow
bugs and bugs with intrinsic sizing
 - dbaron: on the whole, somewhat more compatible with IE... in some
cases, less.
 - dbaron: will make it easier to implement features that web authors
want (e.g., inline-block and integrating the xul box model with html)
 - dbaron: good bit of risk (months of regressions)
 - qa ramping up automated layout tests
 - with the reflow branch, we pass acid2
= grayden's new cycle collector (p2)
= fighting leaks (p1)
 - maybe grayden's cycle collector will help here
 - fix leaks on our own, try to do something about the extension
problem (evangelism?)
= residual style handling (p3)
 - mrbkap: not had a chance to work on it yet.  would be nice to have.
 - mrbkap: makes it so that if you have malformed content, style
matching depends on when chunks of content arrive.
= CSS XUL box model (p3, p4)
 - what safari implements
= Caret painting rewrite (p1)
 - necessary for landing the cairo work
 - already landed, just need to fix regressions
= XUL2D (p3)
 - vlad: needed to support content level zooming and maybe other transforms
 - no vector gfx
 - pavlov: instead of calling this XUL2D, split it up into zooming
APIs and simple content "whatever"
= XBL2 (p3)
 - sicking: new feature, not going to have time to leverage it at our
application level
= JS2 (p2)
 - mrbkap: pretty close to being done with JS1.7
 - mrbkap: not sure what brendan's plan is for JS2
 - schrep: probably going to be JS1.9 in the Gecko 1.9 timeframe
= E4X4DOM (p3)
 - mconnor: brendan soured on the idea of this
= Language-agnostic DOM (p2)
 - schrep: sunk cost do not make things a p1
 - currently fighting some regressions
= SVG images (p3)
= XMLHttpRequest cross-domain (p1)
 - sicking: no happy with the POST security model
= Firefox on xulrunner (p1)
= XPCOM restart in-process (p3)
= offline operation (p3)
 - bsmedberg: cache pinning
 - vlad: other features as well
 - sherman: tie in to local sqlite database
 - vlad: targeting web-apps so that they can work offline
= killing nested event queues (p1)
 - have regressions still

What are the high level goals of Gecko 1.9?
 - how does that compare to what we are actually doing?
 - mostly an open question that we should answer
dev-planning mailing list
[hidden email]