IRC Log Viewer » #firebreath » 2012-09-26

IRC Nick Time (GMT-7) Message
matutetandil 10:09 Hi all!!!!
how are u??
I have a question about mac
I compile my plugin and itś generats a .plugin file... that works fine... but I need a .bundle file to use with chrome extensions...
do you know how to do that?
taxilian 10:09 the .plugin is a bundle file
that's what plugins are
I don't know how to create extensions, but the plugin part of an extension is going to be a .plugin/
so if there is a .bundle involved, it's completely an extension thing, not a plugin thing
and I wouldn't know anything about it
matutetandil 10:09 ok, so you said that if I change the extension to .bundle it should work??
taxilian 10:09 no, I didn't say that
I said that plugins are bundles, and that the extension used is .plugin/
unless the chrome guys are completely out of their minds and arbitrarily forcing you to change the extension of the plugin bundle (which is possible) the .bundle thing you're talking about is probably either unnecessary or has nothing to do with plugins
but I am just guessing
because I don't do extensions
you'd be far better off asking in a place dedicated to chrome extensions
ocholetrasaleato 12:09 hi!
just some basic question
if i want to add my custom code files
i have to do it on the Project folder generated by fbgen
2010
and then run prepscript for VS C++ 2012
am i right?
taxilian:
taxilian 13:09 yes
lol
matutetandil 13:09 Hi again... I still having troubles with mac... :S:S
taxilian 13:09 ok
matutetandil 13:09 I load the plugin making a simbolic link to it... in safary works... but in chrome not... :S:S
taxilian 13:09 did you build it 64 bit only?
matutetandil 13:09 I don't know, I build with cmake with no extra options...
taxilian 13:09 make or xcode?
matutetandil 13:09 make
taxilian 13:09 make only does one configuration
so it probably built the 64 bit version
you need 32 bit if you want to support chrome
build it with xcode
I can't think of any reason it would be a better idea to build with make
matutetandil 13:09 ok... i'll try... but in the past xcode give me some linking errors, that make don't... so I used make
but I'll try, and the I tell you
taxilian 13:09 if that were the case it would simply mean that you didn't have your project set up right
or that you're linking libraries that aren't built for 32 bit, I suppose
matutetandil 13:09 perhaps... I'll look arrownd
taxilian 13:09 JuanDaugherty: good news is it's now building again. we'll see how much I broke… =]
JuanDaugherty 13:09 you mean from just pch or the whole magilla with unsolicited and possible rearchitected streams?
*possibly
taxilian 13:09 both, actually
though still no AX support for unsolicited streams
JuanDaugherty 13:09 i can probably apply whatever it is mutatis mutandis to that leg
taxilian 13:09 maybe; it'll be interesting to see. a stream is a stream on NPAPI; on AX they may be totally different
JuanDaugherty 13:09 my app has better than 90% common code between the two
that's which I thought IE used NPAPI
s/which/why/
a stream is a stream is a stream
to paraphrase g. stein
taxilian 13:09 that's FireBreath's goal; so far it's not a guarantee, though
JuanDaugherty 13:09 so the bot will note the push/commit?
taxilian 13:09 it will
of course, my computer just went haywire
grrr
display driver keeps shorting out (not sure how to better describe it)
let's hope it's the driver and not the display hardware itself
JuanDaugherty 13:09 well if you want to make it available to me for checkout before pushing into the pubic repo can do
taxilian 13:09 so I have it building w/out PCH, but now PCH is broken :-P
JuanDaugherty 13:09 otherwise I can wait
taxilian 13:09 if I can fix that I can push it
though I haven't tested it *at all* yet, so likely there are some significant bugs/oversights
JuanDaugherty 13:09 well you needed to straighten that out anyway, right?
yeah you shouldn't push untested code
taxilian 13:09 well, some of it is already pushed, just not to the main repo
I have my own fork that I develop in
JuanDaugherty 13:09 ah
taxilian 13:09 so I can test on multiple platforms before pushing live
JuanDaugherty 13:09 how far diverged are thay ATM?
just this thing or a whole bunch?
taxilian 13:09 just this thing and other things that I've done while working on this thing
(I get sidetracked sometimes)
it's not hidden or anything. https://github.com/taxilian/FireBreath
JuanDaugherty 13:09 got it. I'll prolly spend the rest of today working on the GUI stuff which does appear sofar to not require any changes
i guess you have to expose the underlying platforms GUI in your pkg unlike the other stuff
the plugin infrastructure as such stuff
taxilian 13:09 yeah; I'd like to build a bit more of an abstraction around that, but there is really a limited amount I can do without building a full GUI framework, and that is one thing I really don't want to do
but I'd like to add some helpers, e.g. one for opengl contexts, one for drawing a pixel buffer/bitmap
things that are common and can be abstracted
JuanDaugherty 13:09 I'm only concerned with dos in this job
taxilian 13:09 you mean windows?
JuanDaugherty 13:09 y
should the fb windows plugin work with opera there?
taxilian 13:09 might. I haven't checked recently
they keep breaking random things
not worth the trouble to support =]
JuanDaugherty 13:09 if you click on a file for which the plugin is the mime type handler, like in explorer (regular not internet) should that start the AX mode of the plugin?
taxilian 13:09 to be honest, I've never made a plugin that was a default handler for a mimetype
I always instantiate using object tags
so I don't actually know
JuanDaugherty 13:09 well this will be the finding out?
taxilian 13:09 I would think so; it's registered for that mimetype
JuanDaugherty 13:09 s/?//
taxilian 13:09 but I have no idea how that would work
let me know =]
JuanDaugherty 13:09 your AX stuff isn't a control?
taxilian 13:09 it is
JuanDaugherty 13:09 anyway, yeah, I'll let you know
the control just needs a container, it's quite possible dos 7 provides a default one
operationally the stuff we've been discussing is key
i.e. unsolicited
taxilian 13:09 dos 7? are you talking about win 7?
or ie?
seriously, dude, dos makes no sense at all in context of windows nt =]
so I'm never sure what you're talking about
grrr… someone in here was having a problem with compiling C files w/ PCH enabled and found a fix
and they told me what it was
and now I can't remember it or find it
JuanDaugherty 14:09 yeah I say dos for windows. I'll adapt, as you know.
JuanDaugherty 14:09 did you say you don't actually build 64bit? I have and have had a NixOS 64 bit machine but really only now making the transition with this job and a current move of one of my public servers
taxilian 14:09 I do on mac and linux
just not on windows
JuanDaugherty 14:09 the situation on Mac is occluded. but not developing there so not worrying about it
on linux it seems to be a non issue unless I presume you try move binaries from a 32 bit host
which I just realized I tried to do with an expat so being used by a perl pkg this morning
taxilian 15:09 moving binaries on linux is a problem any way you do it
this is infuriating; the idl generated code builds as a .cpp file when pch is on, only as .c if it's off
and it makes no sense at all
it shouldn't matter either way
JuanDaugherty 15:09 no it shouldn't. I only ever associate pch with stdafx.h on windows but other platforms do do it, think boost may do x
and stdafx seems really different from what you're doing
taxilian 15:09 it isn't a boost thing; it's just the idl
when I set it to .cpp and build w/out pch it says I'm missing all of the symbols from the midl generated file
which makes no sense at all
JuanDaugherty 15:09 since it's working in vs 8 i'm not giving it attenttion
taxilian 15:09 but when pch is enabled it works
and I don't have hte slightest idea why that would be
JuanDaugherty 15:09 must be a broken/shifted dependency
your correct compilation state became dependent on pch
taxilian 15:09 no; when I rename it to .c (which is what it was before pch) it works
but it should work as a .c file
it doesn't work as a .c file w/ pch enabled
JuanDaugherty 15:09 ah, yeah the c++ processing is a factor
not surprising
taxilian 15:09 I can clearly see where the "missing" symbols are defined, however
and it seems to be building the file
and it's extern C'd
so I don't see where the problem could be coming from
very frustrating, really
I may actually have to set the filename extension based on the PCH state
that is infuriating
JuanDaugherty 15:09 so it's boost precompiled headers you're using not msvc?
or both?
taxilian 15:09 boost? I don't think boost has anything to do with pch
other than that sometimes the pch may include boost
this is the IDL file that I'm fighting
has nothing to do with boost at lal
all
JuanDaugherty 15:09 http://www.boost.org/boost-build2/doc/html/bbv2/reference/precompiled_headers.html
meant that
taxilian 15:09 not using boost build
so it doesn't apply
interesting to see that they support it in their build system, though
JuanDaugherty 15:09 ah
taxilian 15:09 huh
well, that's slightly frightening
it's working…
first run
as far as I can tell, everything that worked before still works
wow. on all browsers. now I'm really scared
I guess the test is to see if it still builds on mac...
JuanDuagherty: it should be ready to start looking at now: https://github.com/taxilian/FireBreath
I'm not pushing it live until I've done some more testing, but it seems to be working
FireBreathBot 15:09 Commit d92c402 on firebreath-1.7 by Byron Clark: "Move these messages to LogLevel_Trace.
Commit 3636a14 on firebreath-1.7 by Richard Bateman: "Merge pull request #78 from byronclark/getvalue-less-verbose..."
Commit d2d86af on firebreath-1.7 by Jason Frey: "Added support prepping on Visual Studio 2012 (Visual Studio ..."
Commit 213375c on firebreath-1.7 by Richard Bateman: "Merge pull request #87 from ManageIQ/add_vs2012_support
JuanDaugherty 15:09 OK, will do after I get the windows stuff which was goin smoove but then adding a regular resource file (.rc) to the end plugin project caused it to implode wiping out the whole tree in vs
i.e. the whole tree of the end plugin project, not the whole fb sln
and may let some other jobs get some cycles
taxilian 15:09 working on updating the API docs now :-/
FireBreathBot 16:09 Commit 355611d on master by Will Roberts: "FIREBREATH-171: initial implementation of handleUnsolicitedS..."
Commit ebdc3a1 on master by Will Roberts: "FIREBREATH-171: more information about unsolicited streams
Commit 90a6402 on master by Richard Bateman: "Disable PCH by default"
Commit cc071ca on master by Richard Bateman: "Fixed compilation sans PCH on windows"
Commit 308c1cd on master by Richard Bateman: "Update URI.h to support + char and add query params"
Commit 974b3eb on master by Richard Bateman: "Partial update of the BrowserStreams features"
Commit 9977bbb on master by Richard Bateman: "Improve constness of browserhost methods"
Commit acc85df on master by Richard Bateman: "More minor fixes to PCH support"
Commit 4e4020f on master by Richard Bateman: "Fixed compilation w/out PCH for IDL compiler; breaks PCH"
Commit 6b24801 on master by Richard Bateman: "change default models on mac to Cocoa/CoreGraphics"
Commit dbeae2d on master by Richard Bateman: "BrowserStreams refactor essentially done, but untested"
Commit 56b437c on master by Richard Bateman: "Fix so things build both w/ PCH enabled and disabled"
Commit 4792453 on master by Richard Bateman: "Improved documentation"
JuanDaugherty 16:09 so master is what people get when the follow the instructions on the download page ?
*they
taxilian 16:09 I think the download page actually points them to firebreath-1.6 right now
the firebreath-1.* branches are the "stable" ones
master is the development, but usually stable one
JuanDaugherty 16:09 oh good I have the damndest thing
even if I go back to git pull, fbgen, and prep I still have this resource corruption issue
taxilian 16:09 you're still rerunning fbgen all the time?
that's really just going to confuse the issue
just delete the build dir
what error message are you getting?
JuanDaugherty 16:09 but no just for this regression
first time in about a week I think
taxilian 16:09 oh good =]
JuanDaugherty 16:09 guess must be something squirreled in the windows user env
did have fb in fb-09-20 and and made fb-09-26 for this but to my astonishment the plug was not regenerated
*plugion
there are no errors but when the project comes up in vs
FireBreathBot 16:09 Commit a26e795 on master by Richard Bateman: "Add SynchronousRequest method and document SimpleStreamsHelp..."
JuanDaugherty 16:09 it just has resource file in the project project, not the normal source, header and generated filters
blew it away as an abhorrent monster or could give more detail
taxilian 16:09 hehe
JuanDaugherty 16:09 i always need to be able to recreate from some stable base of fb and that involves a fresh pull/clone, fbgen, build, merge and rebuild
it's orders of magnitude slower that the VS only 90s era legacy stuff's build
taxilian 16:09 I still don't understand the inclusion of fbgen in that list, but otherwise I agree =]
JuanDaugherty 16:09 but there's a price for everything
even when I removed all the resource files (there's about 10 or so including the rc) I still get this bizarre thing so will take a break and deal with it in a bit
was at the point of adding in all the custom resources, bmps, cursors and what not
the mind blowing part is that even going back to the newly gened plugin with nothing in it in a fresh, clean pull of fb does it
taxilian 16:09 yeah that is kinda weird
JuanDaugherty 16:09 sometimes we just have too many balls in the air
that's why I always combine the first law of Data Processing with having a stable base to reach high from
that being Back Thyself Up
taxilian 16:09 hehe
JuanDaugherty 17:09 so I always need to be able to apply my sources to your fbgen'ed stub and move forward from a given checkpoint
in this case the one just before I tried adding in the resources
JuanDaugherty 17:09 now that' I've restored it I see it was more badly messed that I firs thought. I think I slurred the old and fb's
*old and new
FireBreathBot 17:09 Commit a73761c on master by Geoffroy Couprie: "generate a DMG installer for Mac plugins"
Commit db17e1a on master by Geoffroy Couprie: "remove the default installer's background picture"
Commit 7bc3505 on master by Geoffroy Couprie: "Use FireBreath's logo as background for the DMG installer"
Commit 983ecc1 on master by Richard Bateman: "Add tweaks to DMG installer to be more flexible on program p..."
Commit 6cbe573 on master by Richard Bateman: "Add DMG installer paths for FBTestPlugin, disabled by defaul..."
Commit 6929421 on master by Geoffroy Couprie: "Add a CMake function to build a Chrome plugin"
Commit d715745 on master by Geoffroy Couprie: "use the plugin version from PluginConfig.cmake in manifest.j..."
Commit ab519e1 on master by Geoffroy Couprie: "fix add_chrome_package usage"
JuanDaugherty 18:09 that whole deal was just the normal IDE view change from adding resources I think
taxilian 18:09 huh
JuanDaugherty 18:09 can't know but think and will try to merge into generated this time
you get tired and stupid, what can I say
FireBreathBot 18:09 Commit 5416b42 on master by Geoffroy Couprie: "Add a CMake function to build a XPI package using XPISigner"
Commit 772825c on master by Geoffroy Couprie: "use a more explicit file name to store the passphrase"
Commit 2a3c37f on master by Geoffroy Couprie: "use the plugin version from PluginConfig.cmake in install.rd..."
Commit 5ce5806 on master by Geoffroy Couprie: "Remove debug message"
taxilian 18:09 =]
FireBreathBot 18:09 Commit b3410c1 on master by medusade: "Added X11NativeGdkEventExpose class to be used for Invalidat..."
Commit e211da2 on master by Diorcet Yann: "Windows: Add Exe wrapper to generated msi"
Commit 4bfc802 on master by Diorcet Yann: "Add Cab function and files templates"
Commit 7d4d67e on master by Richard Bateman: "Merge pull request #63 from diorcety/tmp
Commit 6346cef on firebreath-1.7 by Gordon Smith: "Support side by side installation of 32/64 bit plugins.
Commit f858be7 on firebreath-1.7 by Richard Bateman: "Merge pull request #80 from GordonSmith/FB_PLATFORM_ARCH_64
Commit 372487d on firebreath-1.7 by Gordon Smith: "MT and MD switching needed for C files also.
Commit b619489 on firebreath-1.7 by Richard Bateman: "Merge pull request #84 from GordonSmith/WITH_DYNAMIC_MSVC_RU..."
Commit ea63ee0 on firebreath-1.7 by Gordon Smith: "X11/WIN/OSX Mouse Events
Commit 2168701 on firebreath-1.7 by Richard Bateman: "Merge pull request #81 from GordonSmith/MOUSE_EVENTS
Commit 01c1ab2 on firebreath-1.7 by Gordon Smith: "Tweaked how "WITH_SYSTEM_BOOST" works.
Commit 0223e82 on firebreath-1.7 by Richard Bateman: "Merge pull request #86 from GordonSmith/SYSTEM_BOOST_SUPPORT..."
Commit a5613a2 on master by Gordon Smith: "Suppress Erase Background message on windows.
Commit eb23dcb on master by Richard Bateman: "Merge pull request #82 from GordonSmith/WIN_ERASE_BACKGROUND..."
taxilian 18:09 alright, I think I'm done for now
g'night everyone
reichi 18:09 n8 taxilian
FireBreathBot 18:09 Commit e211da2 on firebreath-1.7 by Diorcet Yann: "Windows: Add Exe wrapper to generated msi"
Commit 4bfc802 on firebreath-1.7 by Diorcet Yann: "Add Cab function and files templates"
Commit a5613a2 on firebreath-1.7 by Gordon Smith: "Suppress Erase Background message on windows.
Commit 355611d on firebreath-1.7 by Will Roberts: "FIREBREATH-171: initial implementation of handleUnsolicitedS..."
Commit ebdc3a1 on firebreath-1.7 by Will Roberts: "FIREBREATH-171: more information about unsolicited streams
Commit 90a6402 on firebreath-1.7 by Richard Bateman: "Disable PCH by default"
Commit cc071ca on firebreath-1.7 by Richard Bateman: "Fixed compilation sans PCH on windows"
Commit 308c1cd on firebreath-1.7 by Richard Bateman: "Update URI.h to support + char and add query params"
Commit 9977bbb on firebreath-1.7 by Richard Bateman: "Improve constness of browserhost methods"
Commit acc85df on firebreath-1.7 by Richard Bateman: "More minor fixes to PCH support"
Commit 6b24801 on firebreath-1.7 by Richard Bateman: "change default models on mac to Cocoa/CoreGraphics"
Commit 974b3eb on firebreath-1.7 by Richard Bateman: "Partial update of the BrowserStreams features"
Commit 4e4020f on firebreath-1.7 by Richard Bateman: "Fixed compilation w/out PCH for IDL compiler; breaks PCH"
Commit dbeae2d on firebreath-1.7 by Richard Bateman: "BrowserStreams refactor essentially done, but untested"
Commit 56b437c on firebreath-1.7 by Richard Bateman: "Fix so things build both w/ PCH enabled and disabled"
Commit 372487d on master by Gordon Smith: "MT and MD switching needed for C files also.
Commit a26e795 on firebreath-1.7 by Richard Bateman: "Add SynchronousRequest method and document SimpleStreamsHelp..."
Commit a73761c on firebreath-1.7 by Geoffroy Couprie: "generate a DMG installer for Mac plugins"
Commit 01c1ab2 on master by Gordon Smith: "Tweaked how "WITH_SYSTEM_BOOST" works.
Commit db17e1a on firebreath-1.7 by Geoffroy Couprie: "remove the default installer's background picture"
Commit f858be7 on master by Richard Bateman: "Merge pull request #80 from GordonSmith/FB_PLATFORM_ARCH_64
Commit 983ecc1 on firebreath-1.7 by Richard Bateman: "Add tweaks to DMG installer to be more flexible on program p..."
Commit 2168701 on master by Richard Bateman: "Merge pull request #81 from GordonSmith/MOUSE_EVENTS
Commit b619489 on master by Richard Bateman: "Merge pull request #84 from GordonSmith/WITH_DYNAMIC_MSVC_RU..."
Commit 7bc3505 on firebreath-1.7 by Geoffroy Couprie: "Use FireBreath's logo as background for the DMG installer"
Commit 0223e82 on master by Richard Bateman: "Merge pull request #86 from GordonSmith/SYSTEM_BOOST_SUPPORT..."
Commit 076223e on master by Richard Bateman: "Merged fixes in master with 1.7"
Commit ea63ee0 on master by Gordon Smith: "X11/WIN/OSX Mouse Events
Commit 6346cef on master by Gordon Smith: "Support side by side installation of 32/64 bit plugins.
Commit 4792453 on firebreath-1.7 by Richard Bateman: "Improved documentation"
Commit 6929421 on firebreath-1.7 by Geoffroy Couprie: "Add a CMake function to build a Chrome plugin"
Commit 5416b42 on firebreath-1.7 by Geoffroy Couprie: "Add a CMake function to build a XPI package using XPISigner"
Commit 772825c on firebreath-1.7 by Geoffroy Couprie: "use a more explicit file name to store the passphrase"
Commit ab519e1 on firebreath-1.7 by Geoffroy Couprie: "fix add_chrome_package usage"
Commit b3410c1 on firebreath-1.7 by medusade: "Added X11NativeGdkEventExpose class to be used for Invalidat..."
Commit 076223e on firebreath-1.7 by Richard Bateman: "Merged fixes in master with 1.7"
Commit d715745 on firebreath-1.7 by Geoffroy Couprie: "use the plugin version from PluginConfig.cmake in manifest.j..."
Commit eb23dcb on firebreath-1.7 by Richard Bateman: "Merge pull request #82 from GordonSmith/WIN_ERASE_BACKGROUND..."
Commit 7d4d67e on firebreath-1.7 by Richard Bateman: "Merge pull request #63 from diorcety/tmp
Commit 2a3c37f on firebreath-1.7 by Geoffroy Couprie: "use the plugin version from PluginConfig.cmake in install.rd..."
Commit 5ce5806 on firebreath-1.7 by Geoffroy Couprie: "Remove debug message"
JuanDaugherty 19:09 so apparently what git gets is what you've recently pushed
imagine my surprise when I had cleared the resource thing to be looking at a trap in the debuger with sources for
pluginMain->handleUnsolicitedStream(streamReq);
*debugger
JuanDaugherty 20:09 maybe just because I got a pull from a little earlier, will find out later mover fwd with 1.6
*move
maybe a quarter of the items above are relevant to my app so will merge with it once app stable in 1.6
JuanDaugherty 20:09 also forgot I need to merge 2 sets of resources
3 w fb
JuanDaugherty 21:09 resources are super tedious and break the whole cmake scheme but she's gotta have em
and they look nice when done
at least if I deliver em to the client so they can maintain them in vs
taxilian 21:09 yeah; I never really knew how to work resources into the whole thing. if you think of any ideas for improvement let me know
JuanDaugherty 21:09 been maybe 13 or 14 years since i dealt with this stuff but it seems to be work same, better maybe a bit
taxilian 21:09 it's another of those things I don't personally need =]
.rc files never really change, I think
JuanDaugherty 21:09 first worked with in OS?2 PM
the windows stuff split off from that
taxilian 21:09 ahh, OS/2. that was a wonderful system
that was before I was doing any real serious coding, but my first web apps were written in REXX
JuanDaugherty 21:09 i worked in Boca at IBM as a contract in it's heyday
started there a bit before the first announcement
*contractor
last was at IBM Research c. '96
so it's not that bad, but think the resources but a break on running the prep script, they won't be able to run it after merging in their artwork
*put a break
taxilian 21:09 why is that?
JuanDaugherty 21:09 well we'll see
taxilian 21:09 the way the .rc is intended to work is you put the template in your plugin directory and firebreath will configure it w/ the PluginConfig.cmake values and use that instead of the built-in one
so there isn't any reason you wouldn't be able to keep using the prep script
JuanDaugherty 21:09 i wont run it again to move forward until all the GUI and solicited streams, registration stuff it doss is functional.
so unsolicitied, the 32/64 stuff, etc. and the 1.7 merge will finish it.
taxilian 21:09 you know that the build dir is machine specific, right?
putting it in source control is a waste of time...
so you'll need to be able ot run the prep script when you're done
JuanDaugherty 21:09 it has like a trial and stuff. I'll pm you with an URL for the trial page with it's viewable.
when I'm done, if they accept fb, they can handle it how the like
i'll present them with the one way that worked for me
taxilian 21:09 okay; just saying that it will only build on your computer
JuanDaugherty 21:09 *they
taxilian 21:09 or another with all exactly the same paths
JuanDaugherty 21:09 sure yeah, that's part of our deal
I provide them with a build host.
I don't want their name mentioned in this channel.
taxilian 21:09 that's fine
just making sure you understand is all
I get a lot of people who come in mad because the build directory they put in source control didn't work on another computer, when if they'd read the docs they'd know that the build directory should always be disposable