IRC Log Viewer » #firebreath » 2013-09-27

IRC Nick Time (GMT-7) Message
SinnerSmile 09:09 Hello
anybody here?
taxilian 09:09 nope
we're all gone
SinnerSmile 09:09 that's sad, are you will ever return?
And when you'll be back form saving the galaxy, i wrote a bit here last time, can you give me any suggestions?
reichi 09:09 the galaxy is big, you know
and fighting 12-fooded 9 armed, 4-brained pink aliens is quite a task!
taxilian 09:09 SinnerSmile: I don't remember the details of what you said last time
give me a recap?
reichi 09:09 he's trying to run cef inside chrome iirc
or wants to
taxilian 09:09 oh; you're using a modal dialog?
reichi 09:09 which reminds me...
i wanted to comment on the whole dropping npapi on chrome thing...
SinnerSmile 09:09 Well, the firs thing was about that when i'm trying to run simple window from firebreath browser process itself stop working (i dont mean "not responding"), but just cant make any action with it.
taxilian 09:09 SinnerSmile: how are you showing the window?
SinnerSmile 09:09 well just throug simple function in my ...API.cpp that creating window cia CreateWindowEx and shows it cia ShowWindow
via*
window class registered in StaticInitialize in TesPlugin.cpp
WinProc function declared and put near StaticInitialize function in the same .cpp
and the second one was about, yes, that i cant run CEF under that created window, cos chrome gives me assertion failed
taxilian 09:09 SinnerSmile the reason that it would lock up is because you are some how blocking the main thread
you can't block the main thread
SinnerSmile 09:09 at expression BrowserHost::getInstanceCount() == 0 http://s017.radikal.ru/i435/1309/33/a71333849400.png
ok, how can i start it in another thread?
taxilian 09:09 the image you sent indicates that you aren't releasing all instances of browserhost
so you have a circular reference or maybe a memory leak
SinnerSmile this is a windows question, not a plugin question, so you should be able to find resources to help; I don't have time to help you troubleshoot it right now
SinnerSmile 09:09 the problem is, that I don't understant what "releasing all instances" mean and how can I release it or just what can I do
yes with making another thread i think I will handle it
But maybe any suggestions about this instances? and the firefox doesnt gives me such assertion failed, window pops up, but CEF don't loads itself
And overall, as it seems, that Winter is Coming to the world of NPAPI, maybe it's sensless now to do such things via plugin and better to look at some simple IPC implementation, cos truly I need just small string data to be send to application from plugin, then app will handle the rest
What do you think, is it achievable with reasonable difficulty?
taxilian 09:09 that would be the question. We are looking at options.
eshedoni 09:09 do you guys at Firebreath plan to look for a solution to continue supporting Chrome in their post-NPAPI era?\
taxilian 09:09 go subscribe to the google group
we're discussing them there
rcompton78 10:09 I just found out about npapi getting dropped and threw up a bit in my mouth...
taxilian 10:09 yeah, it should be fun
reichi 10:09 fun is such a flexible word :/
rcompton78 12:09 Anybody know of some more advanced Native Messaging API examples out there. Having problems finding more than launching a calculator. Can't find good documentation
taxilian 12:09 nope, but please let me know what you learn; in particular, I need to know if we can send more than just a string?
rcompton78 12:09 haha I was just about to ask you that. I will keep you posted
taxilian 12:09 I havne't had time to deal with it this week
hopefully I can take a day or two next week
rcompton78 12:09 It is not officially my priority at work so I will keep you posted on my findings. Every other project got thrown out pretty fast
taxilian 12:09 hehe
rcompton78 12:09 I was reading the google groups...you still think to access hardware LoadLibrary on some device driver SDKs would be the native messaging? Any other options you know of that I can look at?? NaCl as far as I can tell won't work since we get sandboxed with almost no outside communication
taxilian 13:09 in chrome, native messaging seems the best bet for that other than websockets
websockets is the most generic option, but I'm not sure if it'll work from SSL
need to test
rcompton78 13:09 so you main app would open a socket and js would talk via that hmmm
hmm indeed
you tricky bugger
taxilian 13:09 SSL is the concern
rcompton78 13:09 ya I have done some websocket stuff but not enough to know that..I will look into that as well
taxilian 13:09 awesome
kylehuff 13:09 I am currently testing the dlopen feature of NaCl to see if I can cause NaCl to interact with a library that is not completely constrained by the sandbox... I doubt it will work though.
taxilian 13:09 seems unlikely
but let me know =]
rcompton78 13:09 ya let me know that would be great but I think all NaCl apps have to go through the google store which is crap as well
the native messaging needs an extension which also (I think) has to go throug the store
so right now the websocket seems the best 'hack' in my mind.
taxilian 13:09 you can install chrome crx extensions from an installer
rcompton78 13:09 Also the most cross browser
taxilian 13:09 though I know very little about how that works
rcompton78 13:09 hmm I'll look into that as well. We used to have a crx and it got flagged last year when chrome started the extensions 2.0 thing
so we skipped plugin wrapped in an extension and just used an installer
that way we had full control and had no red tape on releases
codefame 22:09 hi everyone -- I'm troubleshooting the sendmulti function in jUART (https://github.com/billhsu/jUART). the function isn't registered with the .dll that's already built in the repository. I dloaded & rebuilt without changing anything, registered the new .dll with regsvr32 and replaced the old .dll with it
now the function is registered and can be called, but I'm getting an error related to converting from std::string to std::basic_string, and I'm not really sure where to go from there
I'm far from a dev, but I'd really like to figure this out