User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:20.0) Gecko/20130212 Firefox/20.0
Build ID: 20130212042017
Steps to reproduce:
Following on from the recent announcement by Opera, http://www.opera.com/press/releases/2013/02/13/ and upcoming changes to Firefox:
We might as well join the club, stop innovating and make a monoculture out of it.
Created attachment 713443 [details] [diff] [review]
Damn, I had a patch!
We could just run WebKit in a tab using emscripten.
Clearly belongs in Core.
Maybe not. We could port WebKit to B2G using Emscripten.
Adding azakai for a feasibility assessment.
Well, if you can port Qt, I guess you can port anything...
But, please no.
I have faith in Emscripten!
Needs more discussion of possibilities!
Comment on attachment 713443 [details] [diff] [review]
Review of attachment 713443 [details] [diff] [review]:
Good start, but needs some testing.
I this requires some prior Mozilla experience.
Maybe we could get Google to fund this for Summer of Code?
This is clearly add-on material.
If we used the NPAPI instead, we could get out of process embedding for free.
@jdm Do you want to compile NPAPI with Emscripten?
(In reply to Tim Taubert [:ttaubert] from comment #13)
> This is clearly add-on material.
You are a genius.
All we need to do is make Chrome Frame compatible with Firefox and then redirect all users to http://www.google.com/chromeframe when their browser is opened. If users fail to download and install it we can refuse to let them navigate to any other URL until they do.
I think the simplest solution is to simply slap a Firefox sticker on Chromium and head to the bar. I'm buying. :)
(In reply to Lawrence Mandel [:lmandel] from comment #17)
> I think the simplest solution is to simply slap a Firefox sticker on
> Chromium and head to the bar. I'm buying. :)
I was actually going to suggest to Gerv that I found the solution all of our web compatibility problems in this bug. But that works too.
Unfortunately, many of our web compatibility problems would still remain. I've been privately informed by the Chrome team that if we change our User Agent to copy theirs (which would be necessary for full web compatibility) then they will sue us for trademark infringement.
I therefore propose that we developer 1024 possible different user agent strings which are similar but not identical to the Chrome one and test them by developing a random user agent string picker and shipping it directly on the release channel, using the relative volume of screams of anguish on input.mozilla.org to decide which one to use permanently.
We could sue them for using "Mozilla" in their User-Agent too as it's also trademark infringement ;)
Virtual_ManPL: Shh! Don't give away our legal strategy!
Any plans to adopt Blink instead?
(In reply to Please Ignore This Troll from comment #22)
> Any plans to adopt Blink instead?
We already are: see about:config "browser.blink_allowed" which is set to 'true'. Set this to 'false' if you want to switch back to using Gecko.
OK, I think the joke has run its course.