IRC Log Viewer » #firebreath » 2011-12-29

IRC Nick Time (GMT-7) Message
reichi 04:12 hiho
Sunil_ 05:12 hello
reichi 05:12 Sunil_: if you have questions, ask them here
stop querying people
it is HIGHLY annoying
you don't query people on irc if you never ever sopke a word to them before
ask you question here
and wait for someone to answer
reichi 06:12 i have the requirement to return a newly created Document element containing xml, can i do that with FireBreath?
well, let me alter my question a little bit
dougma 06:12 i wouldn't think so.
reichi 06:12 can i do that with FB without going the hard way of executing javscript code
dougma 06:12 you can call dom methods
reichi 06:12 dougma: i think it should work using object->invoke
Invoke, sorry
dougma 06:12 but creating a "document" might be hard
reichi 06:12 you can do that
i think
dougma 06:12 createElement
and friends i think, off the top of my head
reichi 06:12 document.implementation.createDocument
dougma 06:12 oh!
reichi 06:12 :)
dougma 06:12 well, go for it. :)
that's well supported is it?
reichi 06:12 i would have preferred doing it the native way ;)
oh well
my plugin will only run within "my" webkit
dougma 06:12 i see, so you'll provide that
reichi 06:12 so no focus on compat
that's a standard functionality
dougma 06:12 oh (again)
firebreath helps you, you don't need to execute javascript
reichi 06:12 well yes
i can still code native stuff
and get the newly created document object from the dom
i think FB::DOM:Document.create will do the rest for me
dougma 06:12 not quite
reichi 06:12 hmm i see, firebreath is (at least atm) not meant to manipulate the dom
dougma 06:12 it can
but FB::DOM:Document::create won't create a document
reichi 06:12 i have to passe my createDocument JSObjectRef to it
at least that's what i thought
dougma 06:12 or you can getHost()->getDomDocument()->invoke("implementation")->invoke("createDocument)
in psuedo-code. :)
if you follow...
reichi 06:12 yay
seems to make sense ;)
dougma 06:12 off now. good luck.
reichi 06:12 thx
reichi 08:12 hmm
taxilian 08:12 FireBreathBot: tell sunil_ when you come in just ask your question; you've been dropping in when I'm asleep, but we can leave messages for you like this
FireBreathBot 08:12 taxilian: I'll pass that on when sunil_ is around.
reichi 08:12 how do i properly return null when my return value is a FB::variant
taxilian: bots rock ;)
taxilian: i explained that to sunil about 3 times :/
he kept querying me :/
taxilian 08:12 yeah; having the page up means lots of people can get on that don't really understand IRC; that's both good and bad
reichi 08:12 it's not a problem for me
taxilian 08:12 so like dougma was telling you, Document::create is not for creating a new document object
reichi 08:12 yes
taxilian 08:12 returning null is easy; just return FB::FBNull
reichi 08:12 already got that ;9
taxilian: that's too easy for me, isn't there a harder way?
taxilian 08:12 hmm. Well, I think if you look in APITypes FBNull is a typedef for something more convoluted
so you could look at that and find the more convoluted example if you want ;-)
reichi 08:12 dunno why i missed that :/
taxilian 08:12 probably because it's not terribly well documented
!wiki FBNull
FireBreathBot 08:12 0 results found. Note: Results limited to 8
taxilian 08:12 Everyone: I will be working on trying to catch up on jira issues today. If anyone is available to help, it would be greatly appreciated
linearray 08:12 I fell sick, but I will try to help
might drop out at some point though
taxilian 08:12 linearray: if it helps, I'm sick as well
as is my whole family :-/ so I have 2 sick kids to help my sick wife with while I work
so I know how that goes :-P
I haven't really gotten started on the day yet, still waiting for my oldest child to wake up so I can take a shower and try to rinse my sinuses out and make noise and such, but I'm about to start looking through the jira issues
on a totally unrelated note, questions like this scare me:
reichi 08:12 well
i found a nice thing here
i am actually using chrome(win)/chromium(linux) for daily browsing
anytime i replace my plugin on my linux machine
chrome dies hard
because it tries to acces the (replaced) plugin .so
and i somehow doubt it's meant to be that way
linearray 08:12 so, there is 23 tickets
I tested firebreath-1, it works as advertised
FireBreathBot 08:12 FIREBREATH-1: Summary: Cross-platform timer abstraction.
FIREBREATH-1: Assigned To: [email protected]
FIREBREATH-1: Priority: Minor, Status: Open,
taxilian 08:12 linearray: can you create a current branch and merge in the timer abstraction so I can take a look at it more easily? also, what platforms have you tested it on?
FireBreathBot 08:12 JIRA issue issue commented by richard "as an FYI, your theory is only partially true; all versions of IE use the ActiveXStreamRequest ob..."
linearray 08:12 I checked out mitals repo
taxilian 08:12 can you rebase it on the current master?
and merge anything that needs it?
then push it to a branch on your own github fork
linearray 08:12 let's see
and I tested on VS2010 probably... can't remember
but it's using boost::asio, so unless asio is broken, it should work everywhere
taxilian 08:12 right; still good to test. if you can get it into a branch based off of the latest trunk so it's ready to just pull into trunk we should be able to devise a unit test of some sort for it in FBTestPlugin
and then get it into the trunk and finally closed
I feel kinda bad that mital did all that work and I *still* haven't gone through it completely
but I just haven't had much time and it hasn't been as high of a priority lately; earlier more people were asking for it
reichi: I've seen issues like that before, but it surprises me that you're seeing it on linux; it shouldn't matter if you delete the plugin it's using because on a linux filesystem it'll keep the data around 'til it closes the .so before actually deleting it
reichi 08:12 well
i'm replacing them
and on the next action in chrome
it dies
with some ld-error in my syslog
taxilian 08:12 reichi: don't replace them; put a new file with a different filename and delete the old one
reichi 08:12 well
i think it shouldn't crash anyways
but bu thmm
if it helps :)
taxilian 08:12 reichi: I tend to agree with you, but the method you're using is not the "approved" way to update a plugin, at least according to the mozilla guys, and I'm betting they do it the same way
reichi 08:12 hmm
well, i'll got that way than
actually i am not even testing it in chrome
that's why it sucks even more, that it crashes all the time ;)
taxilian: thx for the hint :)
taxilian 08:12 I love my SSD… visual studio in a VM goes *so* much faster
now if I can just figure out why it isn't working
reichi 09:12 SSDs rock
got one myself a few weeks ago
taxilian 09:12 I admit I was skeptical until I finally forked out the cache to get one (Because my laptop drive kept having issues from the laptop getting bumped around so much)
and… wow. *so* much faster
linearray: do you mind if I assign a couple of issues to you to take a look at when you're done fiddling with that one?
FireBreathBot 09:12 JIRA issue issue assigned by richard "this should be pretty straightforward; it's not a major change"
JIRA issue issue resolved by richard "I don't have a way to test or fix this. If you want to submit a patch or help try to make this wo..."
taxilian 09:12 kylehuff: are you around and interested in helping today?
FireBreathBot 09:12 JIRA issue issue updated by richard
linearray 09:12 go ahead, but I will revenge by assigning FIREBREATH-144 to you
FireBreathBot 09:12 FIREBREATH-144: Summary: Allow logging with macros without 'src' argument
FIREBREATH-144: Assigned To: amxx
FIREBREATH-144: Priority: Major, Status: Open,
linearray 09:12 I can't test the remaining platforms
taxilian 09:12 that's fair =]
linearray 09:12 and I think it's not even necessary... it should work according to documentation
taxilian 09:12 actually that's on my list of things to pull in today
particularly because I feel bad that it's taken me so long to do so =]
FireBreathBot 09:12 JIRA issue issue assigned by amxx
JIRA issue issue resolved by richard "Nobody else is experiencing this, so I'm going to close the issue; if you want to help me do some..."
linearray 09:12 this might work
taxilian 09:12 awesome; I even still have your repo in a remote
okay; I need to go take a shower and try to feel human again (like I said, I'm sick), then I'm going to go down to my main workstation since visual studio isn't working on my VM on this computer and get some real work done
in the mean time I have the timer branch building on this system (mac) to test
linearray 09:12 see you
reichi 09:12 mmh
can i create a DOMParser object using firebreath?
or mor general, can i instantiate something firebreath itself doesn't know/handle
linearray 09:12 git pull request
FireBreathBot 09:12 2 open pull requests:
linearray: Changed unsigned int to uint32_t.
linearray: [FIREBREATH-144] Added logging macros that take a single arg...
linearray 09:12 bah
git pull request
FireBreathBot 09:12 2 open pull requests:
linearray: [FIREBREATH-142] Changed unsigned int to uint32_t.
linearray: [FIREBREATH-144] Added logging macros that take a single arg...
linearray 09:12 \o/
linearray 09:12 btw. two tests in FBTestPlugin fail for me in chrome 16
Use OBJECT custom JSAPI coming in from script (should FAIL!) failed!
ok, see, that's the fever talking
taxilian 09:12 reichi: not easily; there is no way to call "new" from a NPAPI plugin
linearray 09:12 didn't even see that text in parens
taxilian 09:12 however, you could inject a js function into the DOM and call that
linearray: actually those unit tests should pass on each platform; they pass because if trying to do it fails, it passes...
but aparently you can do it on some platforms, not on others
pretty cool, huh?
linearray 09:12 very sophisticated
reichi 09:12 taxilian: mmmh
taxilian: if i use setInnerHTML
FireBreathBot 10:12 JIRA issue work started on issue by amxx
linearray 10:12 the button was irresistable
FireBreathBot 10:12 JIRA issue issue assigned by amxx
FireBreathBot 10:12 Commit 67b61fc on master by Max Amanshauser: "Added logging macros that take a single argument."
Commit 1367d72 on master by Max Amanshauser: "Merge branch 'master' into logger
JIRA issue issue resolved by richard "Pulled in"
linearray 10:12 what on earth are we going to do about ?
FireBreathBot 10:12 FIREBREATH-122: Summary: Support Mousewheel Event
FIREBREATH-122: Assigned To: richard
FIREBREATH-122: Priority: Major, Status: Open,
linearray 10:12 considering that MouseScrollEvent is already implemented
taxilian 10:12 I think it may only be implemented on mac, or just on mac cocoa
or something
not sure
FireBreathBot 10:12 JIRA issue issue resolved by richard "As far as I can tell this is already implemented; if you think this is not the case please be spe..."
Commit 1222842 on master by Mital Vora: "Adding Windows Timer functionality."
Commit 0f5b8ad on master by Mital Vora: "Moving getTimer() implementation to WinTimer.cpp."
Commit ec5812a on master by Mital Vora: "Replacing windows only timer implementation with boost::asio..."
Commit d685c49 on master by Mital Vora: "Adding TimerService Mutex. and fixing the tests."
JIRA issue issue resolved by richard "After far, far too long I have finally merged this into trunk and will be testing it.
taxilian 10:12 okay, I'm headed down to my main workstation; be about 15 minutes, need to eat breakfast first
linearray 10:12 and jira dies
and lives again
taxilian 11:12 huh.
linearray 11:12 do you have an XP VM with VS2005?
taxilian 11:12 I think I do; I need to check
that one is on my list to fix today
kylehuff 12:12 taxilian: absolutely. I am in and out, but if I can do something, just let me know.
taxilian 12:12 kylehuff: think you could glance at
FireBreathBot 12:12 FIREBREATH-150: Summary: GTK dependency cannot be removed by defining FB_GUI_DISABLED
FIREBREATH-150: Assigned To: richard
FIREBREATH-150: Priority: Minor, Status: Open,
Commit e378d38 on master by Richard Bateman: "Merge pull request #38 from linearray/uint32
kylehuff 12:12 I ran into that one just before I left and never had time to investigate; It doesn't link against GTK, it just requires it even though FB_GUI_DISABLED is set to 1.
taxilian 12:12 sounds fixable
kylehuff 12:12 the other part regarding gdkkeysyms.h I hadn't noticed
taxilian 12:12 oooh…. linearray, you still aroudn?
linearray 12:12 yes, trying to figure out FIREBREATH-149
FireBreathBot 12:12 FIREBREATH-149: Summary: Crash after SimpleStreamHelper::AsyncGet completes
FIREBREATH-149: Assigned To: richard
FIREBREATH-149: Priority: Minor, Status: Open,
taxilian 12:12 cool
so, your chnage to support not passing the first param to the logging functions
seems to have broken, like, everything in FBTestPlugin on windows
linearray 12:12 yes
this is bad
taxilian 12:12 unless I merged it wrong
so that is what I'm trying to figure out
linearray 12:12 VS2010?
taxilian 12:12 yes
the specific problem is: FBLOG_WARN("SystemProxyDetector", "Could not detect system proxy: " << e.what());
with the previous way, the FBLOG macros created a std::stringstream object
which made it possible to log any type, not just a string; also, you could use << to build your log message
however, this is not working after I pulled in your change
linearray 12:12 I'll see what I can find out
taxilian 12:12 I'll be looking at it as well, so keep me posted
it doesn't look like it should break...
linearray 12:12 I did not change that original way things work
just added a second macro
taxilian 12:12 yeah, I'm starting to see that and how it works
never tried something like this, though
wonder if maybe something in boost preprocessor would do this better
linearray 12:12 most likely
but boost::preprocessor seemed even more painful to get my head around
taxilian 12:12 hehe. well, I've wrapped my head around it a couple of times; let me see what I can manage
I had to figure out a lot to get the event firing function macro to work
which I still think is really cool =]
linearray 12:12 just tried again on mac. here it works.
taxilian 12:12 with the latest trunk?
linearray 12:12 yes
taxilian 12:12 good to know
so I didn't merge wrong
linearray 12:12 I take it it already fails at compiling
taxilian 12:12 yes; it's not compiling on windows
kylehuff 12:12 so, regarding the gdkkeysym.h definitions -- should I just reproduce the needed definitions in the KeyCodesX11.cpp file? I'm not sure what the best course of action would be in this case.
linearray 13:12 ok, getting there
taxilian 13:12 kylehuff: hmm; is it possible to just compile out the code that wouldn't work?
put an #ifdef around them?
if they disable gui, they don't get events either
so we shouldn't need them
linearray: on which problem?
kylehuff 13:12 ahh, okay. I wasn't aware of that. I'll do that
linearray 13:12 I think the problem is not the trick with the macro chooser
I changed it back to: # define FBLOG_TRACE(...) FBLOG_LOG_BODY2(trace, __VA_ARGS__)
and FBLOG_TRACE("FunctorCall", "Destroying FunctorCall object (void)"); will still emit a warning:
warning C4003: not enough actual parameters for macro 'FBLOG_LOG_BODY2'
taxilian 13:12 hmm
so I can actually use either FBLOG_LOG_BODY1 or BODY2 and it will compile
but if I use FBLOG_TRACE it dies
linearray 13:12 Visual Studio 2010 has this bug which treats a __VA_ARGS__ argument as being one single parameter.
says the internet
taxilian 13:12 hmm; that seems a bit weird
linearray 13:12 the solution is... interesting
taxilian 13:12 so let me ask you a question; is this feature really worth the trouble it's taking?
linearray 13:12 well, for now just revert the change
taxilian 13:12 okay. sorry :-/
linearray 13:12 but I really think at some point it should be addressed in some proper way, because passing parameters that are not used anywhere, makes no sense
taxilian 13:12 they are used
just not by the default logger =]
linearray 13:12 mhm
FireBreathBot 13:12 Commit a98f22c on master by Richard Bateman: "Revert "Added logging macros that take a single argument."
taxilian 13:12 :-/
linearray 13:12 no worries, maybe I will revisit it with boost PP
taxilian 14:12 hmm. definitely are some things I need to clean up in the TimerService stuff
dreid 15:12 Everything is not supposed to go from not working at all to working just fine when I add debug logging to figure out why it's not working at all.
taxilian 15:12 lol
and yet I've had that happen… many times...
FireBreathBot 15:12 Commit b8025eb on master by Richard Bateman: "FIREBREATH-142: Added missing include"
Commit 62f83ec on master by Richard Bateman: "FIREBREATH-1: Partial refactor of timer abstraction to preve..."
taxilian 15:12 linearray: so there was a memory leak in FIREBREATH-1, but I think I fixed it
FireBreathBot 15:12 FIREBREATH-1: Summary: Cross-platform timer abstraction.
FIREBREATH-1: Assigned To: [email protected]
FIREBREATH-1: Priority: Minor, Status: Resolved,
taxilian 15:12 it is definitely considered experimental until further notcie
FireBreathBot 15:12 JIRA issue issue resolved by richard "Thanks Max"
taxilian 15:12 kylehuff: any progress on that GTK dep?
linearray 16:12 is that apitypes.h needed for windows?
on mac os x it worked without it
taxilian 16:12 yeah
linearray 16:12 we really need a jenkins server :)
taxilian 16:12 we had one
but I reinstalled
and I haven't gotten the new one all the way up
linearray 16:12 i see
FireBreathBot 16:12 JIRA issue issue commented by richard "Is your intent that this will allow you to pass in an array of actual characters? Something like ..."
FireBreathBot 17:12 JIRA issue issue commented by [email protected] "hey thanks for taking care of this. I am really sorry about not able to contribute you on integra..."
FireBreathBot 17:12 JIRA issue issue commented by richard "No worries; I think I have gotten things figured out. There were a few more details to make sure ..."
Commit d02c54e on master by Richard: "FIREBREATH-138: Fixed build issue on xp/vs2005"
JIRA issue issue resolved by richard "Sorry this took so long; this is fixed in master, will make it into 1.6.1"
JIRA issue issue updated by richard
JIRA issue issue updated by richard
JIRA issue issue resolved by richard "We have a theory that this may be fixed, but I can't repro and haven't heard back, so I'm closing..."
Commit 7355b1b on master by Richard Bateman: "FIREBREATH-152: Fixed axstream stop method to shut down prop..."
JIRA issue issue resolved by richard "I was not able to repro this, but your suggested fix seems completely reasonable; I don't see any..."
Commit a611a88 on master by Richard Bateman: "FIREBREATH-149: Fixed potential crash in NpapiStreams"
JIRA issue issue resolved by richard "So it is definitely possible to get URLNotify after DestroyStream and vise versa; those two funct..."
taxilian 18:12 kylehuff: ping me when you get back
but if you don't have time that's fine; I know you don't have a lot of time 'til you go back, so family first =]
FireBreathBot 18:12 Commit 7137fc5 on master by Richard Bateman: "FIREBREATH-1: fixed build errors on linux"
Commit e5bfa98 on master by Richard Bateman: "Added mostly untested X11 version of SystemHelpers - Totally..."
JIRA issue issue commented by richard "I have pulled this into the repo; we need to write some tests to go with it."
FireBreathBot 18:12 JIRA issue issue commented by richard "Could you put this in a fork where I can really look at the whole thing easily?
JIRA issue issue commented by richard "I was really hoping to get to this today, but I'm going to have to push it back for a bit; this i..."
taxilian 18:12 bbl