IRC Log Viewer » #firebreath » 2011-02-19

IRC Nick Time (GMT-7) Message
FB_GitHubBot 01:02 FireBreath: firebreath-1.4 Richard Bateman * 3ffa6c3 (3 files in 3 dirs): Fixed issue #147 - possible crash on IE -
FireBreath: master Richard Bateman * 3ffa6c3 (3 files in 3 dirs): Fixed issue #147 - possible crash on IE
FireBreath: master Richard Bateman * 1827488 (3 files in 3 dirs): Merge branch 'firebreath-1.4'
FireBreath: master commits fdd44c8...1827488 -
FireBreath: firebreath-1.4 Richard Bateman * 1827488 (3 files in 3 dirs): Merge branch 'firebreath-1.4' -
taxilian 15:02 morning all
taxilian 16:02 why do people always give me good ideas when I have other things I should be doing rather than implementing those ideas?
FB_GitHubBot 17:02 FireBreath: master Richard Bateman * 800a960 (12 files in 6 dirs): Merge branch 'newmacwindows' into cmake-refactor ...
FireBreath: master Richard Bateman * 1cf9f75 (46 files in 31 dirs): Merge cmake-refactor and newmacwindows to master, FireBreath 1.5 ...
FireBreath: master commits 1827488...1cf9f75 -
taxilian 17:02 for any who may actually be here and paying attention, the master branch has now officially diverged from firebreath-1.4
so if you need to be stable I recommend you switch to the firebreath-1.4 branch
though this is likely stable, there are already some breaking changes
taxilian 20:02 but which wiz?
imofftoseethewiz 20:02 imofftoseethewiz == Pat Lasswell
taxilian 20:02 I guessed that, but I was just wondering which wiz you were off to see
I'm Richard, btw
imofftoseethewiz 20:02 I tried a different approach, much like the one you described in on firebreath-dev, and it crashes like my more complicated version.
Nice to meet(ish) you.
taxilian 20:02 so I've been thinking about your suggestion, and I really like the reference class idea
unfortunately getting all of it through the ActiveX abstraction is painful :-P
but I think I've almost got it figured out
imofftoseethewiz 20:02 Does it work? I'm running on Ubuntu/Chrome and Firefox.
I get crashes on what are ostensibly good pointers.
taxilian 20:02 I created a FB::ShareableReference template class that does nothing but be capable of living in a shared_ptr and hold a ptr to whatever type you give it
well, not yet, but I think it will
imofftoseethewiz 20:02 I.e. the target object has not hit its destructor, the cached pointer is identical to the one put in...
taxilian 20:02 I'm still working through the IE side; once I get that working I think the NPAPI will be easier
give me... hmm. probably 25 or 30 minutes and I can probably push something up somewhere for you to look at
second pair of eyes can be helpful for something like this
imofftoseethewiz 20:02 Ok. Glad to help.
taxilian 20:02 I'm impressed which how well you found your way through the abstractions, btw; very few people understand that part of the code
I really wish more did, because I need a lot of help =]
and currently nobody else has time to contribute code
imofftoseethewiz 20:02 It's an important project, given the way things are moving to browsers.
taxilian 20:02 I've honestly been surprised at how much things have been picking up lately
imofftoseethewiz 20:02 I'm working on providing some basic services to Javascript: file system access, tar, git, svn, awk, sed, that sort of thing.
taxilian 20:02 the interesting thing is that website traffic is higher than ever but the mailing list and chat rooms have been relatively quiet
imofftoseethewiz 20:02 It surprising that there is not as yet a parser-generator stable backed by an NPAPI library, given all of the horrendous home-written JS-based parsers out there.
taxilian 20:02 I guess that means most people can figure out the docs
imofftoseethewiz 20:02 The docs are fairly good for such a young project.
taxilian 20:02 well, a plugin such as you describe will have to be *really* careful with security
imofftoseethewiz 20:02 Yes.
taxilian 20:02 well thanks =] I've worked hard to make it so, and of course many people have contributed
imofftoseethewiz 20:02 I noticed that boost has a logging library, and that fb doesn't have logging.
Do you know if anyone is working on that right now?
taxilian 20:02 lol. I'm trying to work on this and my 19 month old son keeps climbing over me on his migratory trek back and forth across the couch
actually, FB *does* have a logging lib
it's just not enabled by default
and I haven't had time to document it yet
imofftoseethewiz 20:02 Ah, so I guess the docs aren't t...
taxilian 20:02 and none of those I've explained it to have chosen to do so
imofftoseethewiz 20:02 Where do I find it?
taxilian 20:02 also, boost::log is the worst piece of junk I've found so far
I spent 8 hours trying to make it work
wasted time
imofftoseethewiz 20:02 Gotcha.
taxilian 20:02 and settled on log4cplus, though others could be added easily
for kicks, see if you can find anything useful by searching for "log4cplus"
if you can't find enough info there I'll give you the rundown
qwebirc590539 20:02 I missed that... accidentally blew up my browser.
oops.. I should be imofftoseethewiz
taxilian 20:02 hehe. happens a lot when you're developing plugins... =]
(8:19:59 PM) taxilian: for kicks, see if you can find anything useful by searching for "log4cplus"
(8:20:12 PM) taxilian: if you can't find enough info there I'll give you the rundown
qwebirc590539 20:02 I've noticed.
Ok. I'll have a look.
taxilian 20:02 if I had a dime for every time I've crashed a browser, I'd be a very wealthy man...
the log search tool is still pretty young (also open source), but it seems to work okay; I've never needed to use it to try to actually find something, though :-P
taxilian 20:02 this is frustrating... I'm fighting a stupid obscure compiler error
qwebirc590539 21:02 vexing...
taxilian 21:02 error C2166: l-value specifies const object
except that the lvalue *shouldn't be* a const object
I'm not in a const function, I haven't specified const
I have no idea where the const is coming from
I guess the map might force the first piece of the pair as a const...
but even still, I don't know why it would matter
*sigh* I guess I might have to use something other than a map and search through it :-/
taxilian 21:02 did you find anything useful on logging?
qwebirc590539 21:02 Yep. Thanks.
Got a workaround for c2166?
taxilian 21:02 c2166?
ahh. turned out that I was missing #include <map>
though I have no idea how that translated into that error
see if I can make the whole thing build now
this is taking way longer than it should
qwebirc590539 21:02 C++ has some of the more delightfully obtuse error messages.
taxilian 21:02 yeah
particularly when templates are involved
hey, it builds!
qwebirc590539 21:02 congrats.
taxilian 21:02 well, it crashes, but it returns true when I compare two objects :-P
and crashing isn't that big of a deal, right? ;-)
oh... of course
I didn't do an addRef =]
any questions on the logging stuff?
>> plugin().simpleMath
<JSAPI-Auto Javascript Object>
>> plugin().simpleMath == plugin().simpleMath
>> a = plugin().simpleMath
>> b = plugin().simpleMath
>> a == b
that's in IE
let me know if that's at all what you were thinking
I think it works, though
qwebirc590539 21:02 The logging stuff looks totally straightforward.
Ok.. I'll pull that and give it a look.
taxilian 21:02 cool. you can add your own logging engine if you want, bwt
taxilian 22:02 let's see if I can finish this NPAPI stuff in the next 5 minutes before my wife makes me get off the computer... =]
I'm glad you got me thining about this... we didn't have the weak_ptr / shared_ptr stuff from boost when I last visited this issue
...and I think I have it done, just need to see if it builds and works
qwebirc590539 22:02 I'm getting an error....
CMake Error at projects/SecureConnectionServices/CMakeLists.txt:11 (include): include could not find load file: /common.cmake
that's from
taxilian 22:02 yep
that's because you haven't read the link I sent to the mailing list =]
you're on the 1.5 branch
which means you have the cmake refactor
there are a couple of changes you'll need to make
sorry :-/
qwebirc590539 22:02 no worries. I just got on the mailing list, so I'll check the recent msgs.
taxilian 22:02 ok; if my browser ever pops up I could get you the link
but it's being stubborn :-P
it's under "version history" as a subpage
there it is
qwebirc590539 22:02 thx
taxilian 22:02 that'll give you something to do while I commit the rest of this
firefox is working
not bad for free tech support, eh? ;-)
FB_GitHubBot 22:02 FireBreath: master Richard Bateman * 235dea2 (8 files in 4 dirs): Added caching of browser objects to IE
FireBreath: master Richard Bateman * cb0de5f (5 files in 1 dirs): Added caching of browser objects to NPAPI
FireBreath: master commits 1cf9f75...cb0de5f -
taxilian 22:02 ok, it's in both my fork and in the master branch on the main repo now
please let me know if you find any problems with it
qwebirc590539 22:02 ok. thanks.
taxilian 22:02 I need to go now; I haven't made it to bed before 2am in a week, and my wife decided I need more sleep than that =]
don't be a stranger; could really use your help on things if you're interested. you seem to have a decent grasp on how things are put together
dan2: FYI, in case you're not on the mailing list, the cmake refactor stuff I did the other night is in the master branch (which will be 1.5)
would still like to keep going through our laundry list as we find time, though
dan2 22:02 k
taxilian 22:02 g'night all
dan2 22:02 night