IRC Log Viewer » #firebreath » 2012-10-02

IRC Nick Time (GMT-7) Message
JuanDaugherty 04:10 PluginCore release config on VS on the pull I took on the 26th was missing the gecko-sdk lib in the include path
reichi 04:10 hmm
i pulled and merged yesterday
and everything compiled fine for me
JuanDaugherty 04:10 can't remember if debug had, but just reporting
reichi 04:10 1.7 branch
JuanDaugherty 04:10 y
reichi 04:10 but I'm on linux
with my own platform classes
JuanDaugherty 04:10 well VS means Visual Studio, it's strictly limited to that
and actually was the 28th
Guest64715 04:10 Hi, i am using Dialog manager abstraction, it opens the dialog folder choose, and when i choose it calls my callback that is currently empty. How do i use the callback to fire an event so the javascript can get the folder result string?
JuanDaugherty 04:10 I do not know chen/Guest64715, I don't use JSAPI, take it this is a part of that
reichi 04:10 Guest64715: standard events
fb supports registering events
regsiter one with std::string as argument
JuanDaugherty 04:10 s/lib/headers/
reichi 04:10 example in my *API.h
or in yours
FB_JSAPI_EVENT(FileSelected, 1, (std::string));
and in your *API.cpp
fire_FileSelected(mystring);
on the javascript side you can than use standard event handlers
which means you could either use
myObject.onFileSelected = callbackfnc
or using myObject.addEventListener(....
you'll need a browser-switch for the second one to work on all versions of IE, though
(or use a standard framework for it)
in jquery you would use $.('#myNpapiPlugin').on('FileSelected', callbackfnc); e.g.
chen 05:10 Thanks reichi ! , i did what you suggested, still i have this question : I included the code from https://gist.github.com/1368648 to my project this brings up the folder choose dialog, once the user
Guest68762 05:10 selects the folder the method called "DialogManagerWin::_showFolderDialog" invokes a callback, how do i fire the event fire_FileSelected from that callback?
reichi 05:10 well
i have no idea about the whole dialog stuff
sorry
Guest68762 05:10 You have been helpful non the less
ManojGogoi 06:10 Hi
Guest68762 06:10 Hi, How can i call one of my pluginAPI functions from another class? should i pass a pointer of pluginAPI to the class constructor?
JuanDaugherty 08:10 the chinese was from all the defines passed to the resource compiler. Prolly should be none in release, _DEBUG only in debug
taxilian 09:10 JuanDaugherty: nothing related to the gecko-sdk lib changed; it is still working fine for me include path and all. I don't know what you were experiencing :-?
JuanDaugherty 09:10 taxillian, same as other user. Just ran into again because cmake ran and restored the X11 files. As before, deleting them cleared it. Of course the resource thing is unrelated to this.
taxilian 09:10 did you update?
because I fixed the x11 thing days ago
JuanDaugherty 09:10 i'm frozen on what I forked last thursday
taxilian 09:10 the problems come from a something-or-otherX11.cpp in the PluginCore project
JuanDaugherty 09:10 which was master
taxilian 09:10 there is nothing actually in that file
so just delete that .cpp file
and it'll be fixed
JuanDaugherty 10:10 as of 15:00 or so Eastern time on the 27th
right just deleting the complaining X11 related files clears it for me
but if cmake runs it restores them
taxilian 10:10 no, delete the .cpp file
not from the project
the file itself
cmake doesn't restore missing files
it just finds the ones that are there and adds them to the project
JuanDaugherty 10:10 ah you mean in the fb tree
taxilian 10:10 yes
that's what the fix for that issue was
I pulled in a pull request and discovered that it hadn't been tested on windows; just needed to delete the .cpp file
JuanDaugherty 10:10 have you used upstream linking or have others upstream linked to you on github?
taxilian 10:10 I'm not sure what you mean by upstream linking
you mean submodules?
JuanDaugherty 10:10 (talking about fb of course, not other repos you might have but procedures same I presume)
so as I understand it
if I linked my github fork of fb to master, it would get updates
taxilian 10:10 … automatically?
I mean, it's not hard to merge the updates from the main fb fork, but I don't think you can do it automatically
and I don't think you'd want to
JuanDaugherty 10:10 dunno, that's why I was asking for vague qualitative feedback had you used it
taxilian 10:10 AFAIK such a thing does not exist
JuanDaugherty 10:10 yeah, probably not
taxilian 10:10 it's not the "git way"
but you know how to use remotes?
JuanDaugherty 10:10 oh no it definitely does exist
taxilian 10:10 link?
JuanDaugherty 10:10 http://gitready.com/intermediate/2009/02/12/easily-fetching-upstream-changes.html for example
it's actually an integral part of git FWIU, not just github
taxilian 10:10 oh
I've never heard that called upstream tracking
but yeah, that's just using remotes
I use that all the time
JuanDaugherty 10:10 the other upstream huh?
:)
taxilian 10:10 there is no upstream in git
there are just other repos
you add a remote repo for the firebreath main repo and you can pull from it easily whenever you want
you can create an alias like he did if you want; I don't, but you can
JuanDaugherty 10:10 but yeah, there's "origin" in core git apparently, that's it
taxilian 10:10 origin is just the default name of a remote
it's nothing special; you could change it
it's wherever you cloned from
reichi 10:10 usually git automatically links your local projects to the remote you cloned them from
taxilian 10:10 by default
reichi 10:10 you could either do
git pull
which will automatically try to merge if it has to
or you could rebase
but either way you have to commit your local changes before you call merge remotes
s/call/can
JuanDaugherty 10:10 taxilian, what version of chrome is it you're able to attach to in Windows?
i'm having no luck with 22, though it does appear to recognize and possibly to have loaded the plugin
taxilian 10:10 22.0.1229.79
JuanDaugherty 10:10 same except mine as a "m" after the 79
*has
taxilian 10:10 oh, mine does to
too
JuanDaugherty 11:10 what do browsers generally do when two plugins claim the same MIME type?
taxilian 11:10 the browser itself will probably pick whichever is first in its arbitrary, undocumented method of finding new plugins
the browser manufacturers, however, will blacklist whichever plugin doesn't actually own the mimetype (if applicable)
(the main case of this is people who try to hijack the flash mimetype)
JuanDaugherty 11:10 not sing wide char casts and functions was what ultimately cleared the chinese thing
i guess HTML5 short circuits that
(hijacking Flash)
taxilian 11:10 heh. if only
JuanDaugherty 11:10 i mean that way
taxilian 11:10 unfortunately there are still many things that html5 doesn't do that flash does
JuanDaugherty 11:10 by trynna be Flash
taxilian 11:10 fortunately the main one is video, so fewer and fewer websites are building whole websites using flash
actually I think it's more flash was trying to be what html5 mostly is =]
JuanDaugherty 11:10 yeah, I try to explain to people about html5 but it's pointless
taxilian 11:10 because flash was never actually good for that stuff
it's great for video; it's great for certain types of animation
JuanDaugherty 11:10 so I don't now, just refer them
taxilian 11:10 but it became used for things it never should have been used for
JuanDaugherty 11:10 take you mean like a general client server mechanism
*take it
taxilian 11:10 really I just mean "general web page functionality". flash is used (even still) to create applications on the web
JuanDaugherty 11:10 HTML5 has definitely taken a bite out of the Flash market though, last I looked, dev wise
taxilian 11:10 there are a few that I have had to use that just annoy me in particular, so it's a pet peeve =]
yeah
thank goodness
of course, most of what people call html5 isn't actually
it's just using javascript properly
JuanDaugherty 11:10 right there's people that do whole sites in flash
taxilian 11:10 html5 has added capabilities to javascript and the web that solve some of the big issues
exactly; and that is an abomination =]
JuanDaugherty 11:10 fairly common antipattern
reichi 11:10 so the only problem left ist
javscript itself :p
+a
JuanDaugherty 11:10 "i know lang X, so everything has to be in lang X"
*lang/pkg
taxilian 11:10 heh. you know, Javascript isn't nearly as bad as I used to think; the main problem is that so few people know how to architect code well in it
it's actually an impressively powerful language
when used properly
reichi 11:10 well
taxilian 11:10 our entire stack is in javascript, and I've been a little shocked at how much I like it
reichi 11:10 all the browser quirks really make a proper use hard
taxilian 11:10 node.js, mongodb, ajax web front end, etc
JuanDaugherty 11:10 talkin about node and such I take it
taxilian 11:10 yeah, but libraries are getting better to abstract those issues
JuanDaugherty 11:10 and of course a huge part of mozdev is now
taxilian 11:10 still frustrating =]
anyway, I'm going to go get lunch; be back in an hour or so
JuanDaugherty 11:10 some of it is really ugly though, like jquery
taxilian 11:10 jquery source is ugly, maybe, but the library itself is wonderful
JuanDaugherty 11:10 I guess computing envs are like people, the hotties are few and far between
yeah it's very useful
taxilian 11:10 when used properly, of course; javascript always gives you enough rope to shoot yourself in the foot with =]
reichi 11:10 "my" project uses prototype
JuanDaugherty 11:10 but at the lang level actually using it, it (jquery) is fugly
reichi 11:10 when we started with it prototye was like state of the art ;)
taxilian 11:10 I guess it's all in preference =] I love jquery syntax
JuanDaugherty 11:10 yeah prototype is nice
j-b 11:10 Hello :)
taxilian 11:10 TBH, I've never really looked at prototype, so I can't compare
hi j-b
reichi 11:10 hi j-b
taxilian 11:10 I'm about to leave; any quick questions before I go? (I'll be back after lunch)
JuanDaugherty 11:10 yello j-b
go eat!
:)
j-b 11:10 I am still around :)
JuanDaugherty 11:10 I think node.js and such is a mistake, but it's not a positive evil
just a perseverated misjudgement, the one mentioned, i.e. "I know/like lang X, so everything needs to be in X"
and then there's the opposite idiocy, "js is not 'real' lang"
*a