|IRC Nick||Time (GMT-7)||Message|
|FireBreathBot||02:04||JIRA issue http://jira.firebreath.org/browse/FIREBREATH-2 issue commented by noantz "Hi,|
|FireBreathBot||05:04||JIRA issue http://jira.firebreath.org/browse/FIREBREATH-33 issue commented by qifuren1985 "Thanks! It works fine, but can you add a new function called GetName in jsfunction.h, because if ..."|
|FireBreathBot||09:04||JIRA issue http://jira.firebreath.org/browse/FIREBREATH-35 issue resolved by richard "Unfortunately, there is, as far as I can tell, absolutely nothing that can be done about this. Y..."
JIRA issue http://jira.firebreath.org/browse/FIREBREATH-2 issue commented by richard "libcurl certainly makes sense; I would expect WinHTTP to be able to retrieve the proxy informatio..."
JIRA issue http://jira.firebreath.org/browse/FIREBREATH-33 issue commented by richard "if you call .toString() it will give you "methodname()". Since the purpose of having a JSFunction..."
|taxilian||10:04||I rewrote MethodConverter to use boost_pp
it now supports up to 50 arguments, instead of just up to 15
|FireBreathBot||11:04||JIRA issue http://jira.firebreath.org/browse/FIREBREATH-2 issue commented by nitro "It's not possible to retrieve the proxy settings (at least the password) from WinHTTP due to secu..."|
|FireBreathBot||11:04||JIRA issue http://jira.firebreath.org/browse/FIREBREATH-36 issue created by richard
Commit b05ec63 on master by Eric Herrmann: "FIREBREATH-36 Fix memory leaks" http://goo.gl/w8PgR
Commit b77930d on master by Richard Bateman: "FIREBREATH-36 Fixed npobject leak on plugin shutdown" http://goo.gl/xpBrp
|FireBreathBot||12:04||FIREBREATH-30: Summary: crash in Dynamic create plugin
FIREBREATH-30: Assigned To: richard
FIREBREATH-30: Priority: Major, Status: Resolved, http://jira.firebreath.org/browse/FIREBREATH-30
|FireBreathBot||12:04||e09fc33 by Richard Bateman: FIREBREATH-30 Added test to verify issue fixed http://goo.gl/vn6O7
JIRA issue http://jira.firebreath.org/browse/FIREBREATH-30 issue commented by richard "As a side-note, though this is fixed in IE it's still not a good idea to do it this way; I just f..."
|FireBreathBot||13:04||Commit 468e2f3 on master by Richard Bateman: "Disable dangerous injection, crashes ff4 w/ ipc" http://goo.gl/NcxVr|
|taxilian||13:04||doth this now work? my stupid IRC client is acting really strange the last few days|
|taxilian_web||13:04||well, that looked a lot better, anyway|
|taxilian||13:04||hey, I am getting messages again! I can talk to myself!|
|FireBreathBot||13:04||JIRA issue http://jira.firebreath.org/browse/FIREBREATH-36 issue resolved by richard "This is believed to be fixed"|
|FireBreathBot||14:04||Commit 3921284 on firebreath-1.5 by Richard Bateman: "FIREBREATH-18 added examples to fbgen template" http://goo.gl/ZWVhp|
|nirvdrum||14:04||taxilian_away: The release notes should link to something about that breaking change or expound upon it.
Ahh. It is linked, just not where I expected.
|taxilian||14:04||nirvdrum: feel free to update it =]
so that it makes more sense
|nirvdrum||14:04||Yeah, will do.
I couldn't author the content, but now that I see it's there, I'll update the links.
|taxilian_web||16:04||stuartmorgan: Just as an FYI, Safari expects the CALayer to be retained before you return it to the browser; it crashes if you don't. So we have to check to see if it's Safari or Chrome/FF, since Safari will crash if we don't retain and Chrome/FF will leak if we do|
|stuartmorgan||16:04||taxilian_web: you mean it releases what it's given?
Even in a current version?
|stuartmorgan||16:04||That's horribly broken|
|taxilian_web||16:04||at least, that is my understanding|
|stuartmorgan||16:04||Have you filed a bug against WebKit?
That's a clear violation of the spec
|taxilian_web||16:04||on the other hand, they supported it that way before the spec was written =]|
|stuartmorgan||16:04||Then it should leak plugins that don't match the spec
Not crash those that do
|taxilian_web||16:04||one would think :-|
|stuartmorgan||16:04||Please file that bug; as long as Safari is wrong it's going to encourage plugins to be broken|
|taxilian_web||16:04||I'll see if I can get Eric to do so; I don't have all the info|
|stuartmorgan||16:04||Bleh, they totally do
They push it into a scoped pointer
I can actually file this bug from source
thanks for the heads up
|taxilian_web||16:04||ahh, excelent =]
less work for me... :-P
and we'll have to add a version-dependent retain
the really fun bit is that when/if they fix it, anything that was doing it "correctly" will suddenly start leaking
|stuartmorgan||16:04||"correctly" doesn't apply here even in quotes
Apple created a really stupid mess here
yes, yes they did
|stuartmorgan||16:04||https://bugs.webkit.org/show_bug.cgi?id=58282 if you want to follow it
(to be clear, I don't mean "argued" in a negative sense in the report; clearly making your framework not crash under Safari is the right thing to do)
|taxilian_web||16:04||I will do so
as soon as we know when it'll be in I'll need to update FireBreath
|taxilian_web||17:04||anyone sick of waiting for FireBreath projects to compile that wants to help me test something?|
|FireBreathBot||17:04||JIRA issue http://jira.firebreath.org/browse/FIREBREATH-37 issue created by richard
Commit eb69607 on firebreath-1.5 by Richard Bateman: "FIREBREATH-37 Support prebuilt static libs" http://goo.gl/qMJuK
Commit db47f46 on firebreath-1.5 by Richard Bateman: "FIREBREATH-37 with FB_BINONLY dont require project dir" http://goo.gl/BgiiW
|dan2||18:04||I must eat too much pizza when the pizza place I call knows who I am on the phone and doesn't even bother asking for my phone or address anymore|
probably a good indicator, yes
|dan2||18:04||they deliver on bicycles
rain or shine
|dan2||18:04||now they have like the smallest delivery area ever of a pizza place
but it's damn good pizza
certainly NYC style
|taxilian_web||18:04||you're making me hungry =]
you saw that 1.5 is in RC?
|dan2||19:04||did not see that yet|
|taxilian_web||20:04||and naturally I've already made several changes since I released the RC :-P
but not to core functionality; just tweaks
|dan2||20:04||I take it the new event definitions made it into 1.5|
|taxilian_web||20:04||dan2: they did
maybe one of these days I'll even document how to use them =]
pretty straightforward to me
|taxilian_web||20:04||I just updated Facebook's plugin to use them, though; it's much nicer|
|dan2||20:04||more straightforward than the previous|
|taxilian_web||20:04||oh, it is... but if you don't know about it, it still doesn't get used =]|
|dan2||20:04||what are tracepoints in vs?|
|taxilian_web||20:04||not a clue|
|dan2||20:04||I can tell you that link time code generation in vs2010 cuts my binary sizes by over 2 megabytes
|taxilian_web||20:04||? what do you mean by link time code generation?
huh. never heard of this before
|dan2||20:04||there's a switch called LTCG
basically you compile and link everything with that switch
it compiles to intermediate code
and does super aggressive interprocedural optimizations as well as inlining at the link stage where it does the real compilation
|taxilian_web||20:04||huh. probably takes a lot longer?|
|dan2||20:04||last stage yes
but it runs a lot faster
and the binary is definitely smaller in my case
I'll have to look into that
|dan2||20:04||not that I can tell yet other than longer compile times|
|taxilian_web||20:04||huh. so I set it up so that I can build the 4 Core projects once and reuse them (using export, haven't done the install(export..) bit yet) and it cuts the build time of my plugin significantly
if I were doing binary releases anyway I don't care if the build time is a bit longer
I'll have to play with this more as I go
|dan2||20:04||it might be time for a visual studio build wipe|
|FireBreathBot||21:04||JIRA issue http://jira.firebreath.org/browse/FIREBREATH-33 issue commented by qifuren1985 "If I call .toString(), I must parse the String, it may run slower, but I can accept.
JIRA issue http://jira.firebreath.org/browse/FIREBREATH-33 issue commented by richard "empty should be void or undefined, null should be null; so yes, that may be a bug. That should be..."