|IRC Nick||Time (GMT-7)||Message|
I am looking for making a plugin for chrome and firefox in windows. The plugin need to get information information from a mapped memory create by a C# application.
and he has to edit the content of some webpage selected (they are listed on the mapped memory)
Is it possible to make this with FireBreath ?
(when I talk about content, I mean autocomplete Login and Password)
|didu||06:12||the plugin has to be always open without called by a webpage|
|jshanab_||06:12||PLugins are loaded when a web page has an element specifing it's mime-type. While I think you can access the mapped memory I cannot see how it can be always open without a web page. For this you will need an extension not a plugin.
But extensions are different for each browser. I do not know if there is a comparable cross browser extension framework.
Extensions can have plugins in them. Maybe an extension can insert a plugin into the dom of a page?
|didu||07:12||Thank you for your answers
The plugin while just open the mapped memory if it was create by the C# application before
Plugin can edit the content of webpage ?
and read it ?
(I think your idea of extensions with plugins in them is a good)
|jshanab_||07:12||IF the memory can be accessed from c++, i think a plugin can do it.
A plugin to my knowledge cannot manuipulate the DOM directly but an extension is different on every browser and can deal outside the page, with tabs and such so it's lifetime at least allows it to still be around.
Going only by what I have seen two extensions give hope. firebug which definatly reads the page and restlet which I think varis the page for the request type
|didu||07:12||Really thanks you for all you answers !
I was thinking access to mapped memory will be easy but It look like I have to code more than a simple extension :P
Extension are so limited for such things :(
|jshanab_||07:12||I have only looked at FireFox extensions and they are cool in that they can manipulate the browser interface, but they are a form of XML the heavy lifting I think is left to be done with the c/objectivec/c++ inside a plugin.|
|didu||07:12||yes it is true :P, but Firebreath use boost who can manage xml
I am very lost :( I will wait the other people like you say, they are from which country ?
|jshanab_||07:12||Well the defenitive answer will come from taxilian and he is on west coast of US I think. I forget where he lives bu I know he is 2 hours behind me and I am in Chicago.|
|RenJuan||07:12||pretty sure there's nothing above Firebreath in generality, or I guess beside it in the matter of a xbrwsr extension builder|
|didu||07:12||Ok I like Chicago :P, I was living in Detroit when I was young :) Now I am in Saint petersburg (Russia) for my study ^^. There is approximately 12 hour jet lag but I will wait|
|RenJuan||07:12||sfaik, extension, plugin, add-on are all as far as common sense if not details of implementation, the same thing, API for customization
nothing I know of, doubtless is something
RenJuan. I am sorry if I missunderstood your statements, I have to read them 2 or 3 times and I am still not shure what is statement and what is question.
|RenJuan||07:12||you ignore IE?
or do you say the IE plugins are NPAPI?
|didu||08:12||I already have IE plugins, it is not NPAPI
hummm it is called BHO for internet explorer and you can manage the memory
and edit the content of webpage
I understand why chrome and firefox don't want that extensions can't acces memory for being more secure but I need it :(
|jshanab_||08:12||No, For plugins that is one of FB's best tricks It has the active-x equivelent to allow you to write the code once and it just works on active-x browsers too. I need to look that up|
|didu||08:12||Oh that wasn't for me sorry :P|
|jshanab_||08:12||The claims about one tech being more or less secure are market-speak. NPAPI became a defacto standard, even IE supported it for a while. But Microsoft wants you to drink their koolaide, in for a penny, in for a pound. If they can get you to use their tech, then you end up with, after hours of development, something that "only works on ie" The apple crowd is no different.
The development environments out there that really work well are what determine who gets software first. Microsoft's Visual studio is great. Xcode is great. Each have their own language and libraries they push but it means desktop apps come first on MS for houses that write MS and first on apple for those that write apple software. Cross platform stuff goes to *nixes and bubbles up to the...
|didu||08:12||haha I agree xD|
|jshanab_||08:12||Having writen this videoplayer as a browser plugin and as an iphone and android app and trying to port it to mac as well. I gained a new perspective. It is very obvious why iphones get apps before droid|
|didu||08:12||I think chrome and firefox make everything that it is not easy to use code of extension from one browser to an other.
When I was looking for communication of interprocess, firefox have some tech that chrome don't have and chrome have techno that firefox don't have.
|jshanab_||08:12||Chrome is going there own way. I have been using ZMQ in my aps and was thinking about putting it in my plugin. To make a browser plugin that can communicate with a service either dameon on linux or windows service. Abusing the browser as my cross=platform windowing toolkit|
|kylehuff||08:12||between Opera, Safari, Chrome and Firefox... I officially hate extension development|
|didu||09:12||Don't say me such things, I just began yesterday I don't want to give up so fast xD even if I am totaly lost|
|kylehuff||09:12||there does exist a few frameworks that abstract extension development between browsers. Kango (local framework you download) and Crossrider (web-based application). Neither support NPAPI bundled plugins (since some browsers do not support that), but you could maybe add your NPAPI to the versions that do.|
|didu||09:12||OK I check Kango|
|kylehuff||09:12||I've not used either, so I can't comment on their functionality or license. just passing on that they exist.|
|jshanab||10:12||OMG. I just discovered a nasty c++ (at least on MS) "bug" If you cast the this pointer of a class to an interface you forgot to inherit from. (oops Git Stash) It compiles, it links, it runs. but it crashes because it creates a pointer to the destructor ROTFL
"silent but deadly" ;-)
|didu||10:12||diorcety you are french ?|
|kylehuff||11:12||he also talks to himself... =c )|
It's better to ask question here
|didu||11:12||I try to make a extension or a plugin for chrome and firefox that are able to read xml on a mapped memory create by an other application|
where is the issue?
|didu||11:12||the extensions of plugins has to modify some part of the web page with the content get from the mapped memory
I look at the beginning to create a extension for reading the mapped memory but the google or firefox API are not able to do that or not completely
There I find NPAPI that can be able to read mapped memory
but I see too that plugin has to be call by webpage
but my plugin has to be always activate to check every open page
and see if there is something to edit
But I completely don't know how to do it and if it is possible to edit content of web page from a plugin
Therefore if someone has understand what I said (sorry if I badly write it :P) and can told me if it is possible :)
you can handle the document in the plugin
|kylehuff||11:12||didu: the plugin can be always running in the extension background page (overly in Firefox) -- and the content scripts can interact with the plugin via message passing.
loading the plugin on each page could be problematic
|diorcety||11:12||kylehuff: using firefox plugin functions it must be possible
kylehuff: in the same way that Adblock+
|kylehuff||11:12||I am not saying it isn't possible, just problematic. It would be easier to house the plugin in the background (per instance of browser) and handle the interaction between the webpage content and the plugin through the extension.|
|didu||11:12||How can I make the plugin running in the background ? did I have to make a extension to do this ?|
|didu||11:12||Oh ok therefore the extension will be like a proxy
he get the content and send it to the plugin
|kylehuff||11:12||exactly. and your extension code-base could be pretty much the same for both firefox and chrome. the only difference would be the extension build/control files for each specific browser|
|didu||11:12||Ok therefore I have to make a extension loaded in the background, who load the plugin and who get the document that will be send to the plugin for editing ? or the plugin can do everything him self after loaded (his is called when a page is open) ?|
|kylehuff||11:12||you could probably do it either way, depends on the data I guess. The extension route also has the benefit of handling your plugin distribution.|
|didu||11:12||Ok I see how it have to work now :)
Next step I have to find how to do it :P
|didu||11:12||Thanks you for all :)|
|jshanab||12:12||kylehuff. i tried and I love the way the extension handles the plugin install. I just don't know how to make that work in chrome and IE|
|kylehuff||12:12||In IE, I don't know. In chrome it is easy. you just need to include the plugin binaries in your extension directory, and reference them in your manifest.json file. I have a current chrome extension you can reference: https://github.com/kylehuff/webpg-chrome/tree/master/extension|
|kylehuff||12:12||the only catch regarding bundled plugins in chrome, is that you have to reference any OSX binaries before any linux binaries.|