forked from servo/servo
-
Notifications
You must be signed in to change notification settings - Fork 0
Meeting 2013 07 22
Josh Matthews edited this page Jul 22, 2013
·
1 revision
- Reftesting
- Float leftovers
- JS Memory
kmc, jack, jdm, eatkinson, azita, tikue, simonsapin, pcwalton, brson
- jack: reftests landed and we should require tests for things that can be tested
- (some discussion)
- jdm: are we agreed that we should require reftests when possible?
- all: yes
- jack: what is contenttest?
- jdm: it's like gecko's mochitest. it has a harness and some asserts to test various scripting things.
- pcwalton: we dont' use mochikit do we?
- jdm: no
- jack: does it currently work, or is it out of date? does it run as part of make?
- jdm: it is almost up to date, and it can be automated.
- jack: what's left to do on floats?
- pcwalton: we don't have clear.
- eatkinson: mixing floats and text doens't work yet.
- simonsapin: do we have margin collapsing?
- eatkinson: no
- pcwalton: is that in acid1?
- jack: perhaps. there are some margins.
- simonsapin: margin collapse is only in block direction, so top and bottom.
- jack: based on screenshot, it looks like we're calculating (margin widths?) wrong. ...
- pcwalton: let's get bugs open on all of the problems we can identify. this is valuable.
- (some discussion of problems apparent in acid1 screenshot)
- jdm: i'm starting to look into the various segfaults. i'm taking the easiest path, which requires ~8 lines of changes to spidermonkey. it uses reserved slots and then reserves more slots when possible.
- jack: is it in a state where others can help?
- jdm: not quite yet, but i can publish it to a branch if others want to look at it.