IRC Log Viewer » #firebreath » 2013-06-28

IRC Nick Time (GMT-7) Message
wavey 01:06 niftylettuce: std::ofstream ma_file( "where_your_file_is", "wb" ); ma_file << "Write stuff"; ma_file.close();
http://www.cplusplus.com/reference/fstream/fstream/
you ARE writing an NPAPI plugin (since you are using firebreath)
Sebustian 02:06 Hi , Guys
I am newbie in browser plugin program . My goal is to invoke a javascript through NPAPI while every page onloading , is it possible?
Any answer would be great appreciated .
wavey 02:06 There are some hooks provided to the extensions from some particular browsers
If I remember correctly, firefox calls them "Context" plugins ....
or I might be wrong...
but they are not native plugins
so no NPAPI
I am not sure if with FireBreath you can create a *browser extension* and not just an NPAPI plugin
Sebustian 02:06 Hi , wavey . Thanks !
I tried to get page URl in NPAPI , referencing https://developer.mozilla.org/en-US/docs/Getting_the_page_URL_in_NPAPI_plugin
reichi 02:06 hmm that's a plugin
but a browser extension
npapi has to be embedded into html
http://colonelpanic.net/2010/08/browser-plugins-vs-extensions-the-difference/
Sebustian 02:06 Get it , thank you !
I have another question , is it possible to redirect some page(containing mimetype flash) to designaged page through NPAPI (npn_getUrl) in firebreath , if yes , how)
wavey 03:06 FB::BrowserHost::Navigate I guess
http://www.firebreath.org/display/documentation/class+FB+BrowserHost
found as m_host on your main JSAPIAuto plugin
reichi 03:06 well
no
you could actually react on the mimetype for flash
but there is no defined behaviour if two plugins request the same mimetype
or handle
first come first serve
basically
wavey 03:06 ah, that's also true
reichi 03:06 where first come means the order the browser is cycling through the plugin mime-types
but i gues across browsers the results will be entirely random
krishna__ 03:06 reichi, if you used nutty, would be glad if you rate it at https://chrome.google.com/webstore/detail/nuttyio/kmihcoocnipiclchikdmgfblhngdedem
reichi 03:06 i just played around with it ;)
krishna__ 03:06 i am trying to get the initial traction
reichi 03:06 but i guess it's enough for a rating :)
krishna__ 03:06 thanks :)
if possible write a review also
on the chrome web store
reichi 03:06 oh hmm
doesn't work :/
krishna__ 03:06 oh
reichi 03:06 (you'll still keep the 5 stars ;))
krishna__ 03:06 you are using chrome right? wondering why it does not work
reichi 03:06 yes chrome
hmm
krishna__ 03:06 you were able to give the stars?
reichi 03:06 dom excception in reconnectin-websocket.js:151
exception even
nutty doesn#t work ;)
krishna__ 03:06 oh!! :)
reichi 03:06 i've already rated it
because i think the idea is so cool
and it's open source :9
bugs can be fixed, we're all developers and know that, right? ;9
Uncaught TypeError: Object #<HTMLObjectElement> has no method 'ptycreate' master.js:331Uncaught TypeError: Object #<HTMLObjectElement> has no method 'ptyrowcol' master.js:52833Uncaught TypeError: Object #<HTMLObjectElement> has no method 'ptywrite'
hmm
the plugin starts to loud though
load even
hmm wait...
it's 64 bit here
oh damn sorry... gimme a sec...
works ;)
krishna-- 03:06 reichi, sorry got disconnected
using websocket or webrtc?
reichi 03:06 it works
sorry for the confusion...
krishna-- 03:06 ohh
great
reichi 03:06 i've enabled click2play recently
:/
for plugins
krishna-- 03:06 let me check what is click2play
ah ok
reichi 03:06 woth seem to work fine
well there was a small hint
krishna-- 03:06 great
reichi 03:06 that nutty.io want's to execute a plugin ;9
which i missed...
krishna-- 03:06 ah ok
yeah nutty is opensource, else i wouldn't myself use it ;)
reichi 03:06 am i right that you#re only doing the communication in npapi
and all the rest in javascript?
hmm no
i'm obviously wrong there :p
krishna-- 03:06 i use npapi to create the tty
everything else happens in JS
reichi 03:06 ah ok
so i'm right
i had a very quick look at the npapi code
and it was quite simple
krishna-- 03:06 yeah
reichi 03:06 oh
the youtube video is "shining through" in crhome
on the page
that's pretty annyoing ;)
krishna-- 03:06 shining through in the sense?
reichi 03:06 overlaying the console
after clicking demo once and dismissing it explicitly that's gone
krishna-- 03:06 got it ... hmm i had done some workarounds ... thought it had fixed that issue ... will look at it
that issue is not seen in FF though
it works fine on chrome, if you don't click demo and start the terminal
reichi 03:06 https://reichholf.net/cloud/public.php?service=files&t=5208b1bad0577d1e25143143643998d0
krishna-- 03:06 let me check
reichi 03:06 haven't clicked demo
just started the terminal
krishna-- 03:06 ok
reichi, damn ... will look at it
reichi, what is your chrome version?
reichi 03:06 28
krishna-- 03:06 ok, not released yet right?
reichi 03:06 sure
krishna-- 03:06 i updated chrome few days back and it was 27
reichi 03:06 oh
maybe i'm on the beta channel
dunno ;)
28 is the first version that comes with blink
krishna-- 03:06 ah ok!
wavey 04:06 krishna--: Did you have any trouble with google when you released an NPAPI plugin on the chrome store?
Because I saw the developer aggreement and they said that NPAPI plugins are only accepted in special cases
(when approoved by google)
reichi 04:06 i never did
so no ;)
krishna-- 04:06 wavey, i did not have any trouble
they took a week to approve it though
but when i subsequently uploaded the plugin, it reflected on the store immediately
wavey 04:06 nice ;) Thnx
krishna-- 04:06 wavey, would be of help you rate/review my extension :-)
https://chrome.google.com/webstore/detail/kmihcoocnipiclchikdmgfblhngdedem
yecril 13:06 taxilian: do you have access to the source code of Confluence?
taxilian 13:06 hmm. I might. I haven't looked
yecril 13:06 They say it is not open source and available to paying members.
taxilian 13:06 well, we're not a "paying member", but we have a free os source license
rather a free open source project license for the thigns
things
yecril 13:06 They say open source projects get runtime license and paying members get the source.
taxilian 13:06 then I don't have the soruce
yecril 13:06 Then the invalid HTML problem is not fixable unless I pay them $10.
taxilian 13:06 well, while I agree that it's not really ideal, is it really that big of an issue?
I'd just report the bug and hope they'll fix it soon
yecril 13:06 I am not in a position to report a bug without having a license.
taxilian 13:06 hmm. didn't realize you had to have a license to report a bug
well, write up what I should report and I'll send it to them, then =]
yecril 13:06 I reported it three times in their wiki and they deleted it three times.
taxilian 13:06 honestly not really surprised... if it's something they don't consider a major problem, then it's not something that they want advertised to the world on their wiki
yecril 13:06 I did but you deleted my comment.
taxilian 13:06 because similarly I don't see any reason to advertise to the world the faults of our wiki =]
yecril 13:06 So where should I put it?
The reason is to warn the reader that he may be expecting unpleasant things.
taxilian 13:06 just because it doesn't pass the w3c validator doesn't mean there will be unpleasant problems with the site
yecril 13:06 It is better to say "We are sorry for the temporary inconvenience" than to pretend everything is all right.
But there are, e.g. URL with spaces.
taxilian 13:06 as far as I can see, everything is fine, though perhaps not completely ideal
unless you know of any specific bugs or negative UX that results from the issues?
yecril 13:06 URL with spaces is certainly not fine
taxilian 13:06 again, do you know of any specific bugs or negative UX that results from a URL with spaces?
yecril 13:06 When you paste such an URL in a plain text medium, only the thing up to the first space counts, and the link is broken
taxilian 13:06 every browser that I know of deals with it just fine
but most browsers will automatically convert that to %20 instead of space
I don't know of any that don't, though I imagine there may be some
anyway, these are things that a few people might have trouble with, but the vast majority will not
yecril 13:06 Not if you copy to Clipboard.
taxilian 13:06 copied from the home page: http://www.firebreath.org/display/documentation/FireBreath+Class+Reference
using Chrome
'firefox and safari do the same
what doesn't?
yecril 13:06 Seamonkey
taxilian 13:06 that sounds like a seamonkey bug
and the percentage of our users that use seamonkey is very, very low
I'm not saying that it shouldn't be fixed
I'm just saying that when you have a small bug that very very few people know about or are affected by it usually doesn't make sense to broadcast to everyone that the bug is there
yecril 13:06 Seamonkey is a Web browser, it is free to do whatever it likes with nonconforming material
taxilian 13:06 yes, I know what seamonkey is
yecril 13:06 So, you told me and thus it is decided. Where do you want me to write?
taxilian 13:06 well, is what you want written just what you posted to the wiki before?
because I do still have that
you wrote "This page and probably all pages below are http://mandrillapp.com/track/click.php?u=13710631&id=04c67a85811747d092faee3853225bc3&url=http%3A%2F%2Fvalidator.w3.org%2F&url_id=e8199e454e5bf57ef07935308798669301c09fef&tags=_all,[email protected]  It should not happen in a project targeting Web developers."
heh. except instead of the url you put a link "Invalid"
it just pasted weird
yecril 13:06 Well, I have no idea what did it get that mandrilla from
I surely pasted a validator URL
taxilian 13:06 right; that's from the email server we use
sorry
yecril 13:06 Except for that, that is it.
taxilian 13:06 however, I will not send that to atlassian; they will ignore such a statement. It would need to be a clear explanation of what the problem is, specifically which elements you consider to be most problematic, and why they should be fixed. "because that's what the spec is" will not be considered a good reason by them
they are obviously using the invalid meta tags for their own stuff. My understanding is that invalid meta tags will just be ignored and I don't know of any browsers that will not safely ignore them
other things might be more relevant, but I still don't understand why this is a major problem worth warning users about, so I'm sure they won't
yecril 13:06 META tags are not a major problem, spaces in URL are.
But I cannot even get to that because validator says it is wrong and here are your 102 errors.
taxilian 13:06 where specifically are there spaces in the URL that caused you problems?
yecril 13:06 In the class browser, I shall try to paste an URL
taxilian 13:06 ahh
that might be my fault
but again, are you seriously telling me that seamonkey isn't smart enough to interpret that and put %20 in the url?
yecril 13:06 It puts %20 into the GET request, but not to Clipboard.
taxilian 13:06 that's too bad. well, there is another way you can link to the page
click "tools" -> "Link to this page..."
and it'll give you an alternate url
so the reason it does it that way is that I have a program that imports the doxygen docs
and it is what puts the spaces in
not confluence itself
you didn't mention before that you were in the class docs area
you put the warning on the root of the home page
yecril 13:06 http://www.firebreath.org/display/documentation/class+FB+BrowserPlugin+Members
taxilian 13:06 most likely the missing members haven't been documented
yecril 13:06 from the tool
taxilian 13:06 feel free to document them, test it (the doxygen config is in the source tree) and submit pull requests
yeah; that class hasn't been documented yet
yecril 14:06 http://www.firebreath.org/display/documentation/class FB BrowserPlugin Members
from SeaMonkey
taxilian 14:06 right
yecril 14:06 Observe how the link is not clickable any more
taxilian 14:06 yes, I'm not stupid
I can see what you're getting at
so, as I said
this isn't a bug in confluence
it's actually to do with the documentation importer
that importer is found in the firebreath codebase
under doxygen/doc2confluence.py
yecril 14:06 Should I start from ./Doxyfile?
taxilian 14:06 well, you may have to build with doxygen in order to understand the source format
but the python file is what actually changes the doxygen files to make it possible to use them in confluence
anyway, I'm actually really swamped today and I've spent mroe time than I can afford on this
yecril 14:06 Do you mean build with doxygen installed, or telling something to cmake?
doxygen is not mentioned in CMakeCache
nor in the Makefile
doc2confluence.py does not say how it should be run
taxilian 14:06 the challenge with debugging that
and the one that I'm trying to figure out how to overcome in this instance
is that using it requires priviledged access to the wiki
specifically, it requires a username and password that I don't generally give to anyone
also if you make mistakes while working on it it can completely mess up the docs on the wiki
it's not really friendly for troubleshooting :-/
this is how it's run
https://gist.github.com/taxilian/5887822
and now I really do need to get back to other things. Take a look if you'd like and see if you can come up with a solution to the problem; we can talk about it more later
yecril 14:06 Thanks. Maybe I will try to run it against a trial Confluence server.
sabotaged 15:06 anyone tried compiling firebreath on mavericks/xcode 5 yet?
taxilian 15:06 not yet =]
sabotaged 16:06 how about xcode 4