updates Meet the browser: Firefox Next

[Christophorus_Hatchenson]

Firefox Metro is scheduled to role out when Firefox 26 is released. Look decent so far. I wonder how the add-ons system will carry over if it will at all?

 

Sexy. Makes me want to buy Windows tablet. Is it installed separately, or included in standart Firefox installer? Can I launch desktop app by clicking tile?

[The_Decryptor Veteran]

It's handled like IE, where if you start it from the tile you get the tablet app, but if you start it from the desktop you get the normal app.

Addons will follow the Firefox Mobile route, you don't have a normal XUL interface to hook into, it's all JS API calls to add UI stuff.

[+Zlip792 MVC]

From Inbound:

 

Golden old bug related to autocomplete: https://bugzilla.mozilla.org/show_bug.cgi?id=135019

Ion Monkey: https://bugzilla.mozilla.org/show_bug.cgi?id=905091

Ion Monkery (performance improvement on benchmarks): https://bugzilla.mozilla.org/show_bug.cgi?id=864220

JS performance regression fix: https://bugzilla.mozilla.org/show_bug.cgi?id=903802

OdinMonkey (aka asm.js): https://bugzilla.mozilla.org/show_bug.cgi?id=904809

Content Security Policy in XSLT: https://bugzilla.mozilla.org/show_bug.cgi?id=663567

Content Security Policy: https://bugzilla.mozilla.org/show_bug.cgi?id=802872

Check Spelling issue in Editor: https://bugzilla.mozilla.org/show_bug.cgi?id=905176

Off thread JS loading when scripts running from XUL: https://bugzilla.mozilla.org/show_bug.cgi?id=897655

ES6 related code cleanup: https://bugzilla.mozilla.org/show_bug.cgi?id=892549

JS Engine performance improvement: https://bugzilla.mozilla.org/show_bug.cgi?id=862848

Box2D regression fix: https://bugzilla.mozilla.org/show_bug.cgi?id=905760

[cetla]

Interesting tip for Aurora build (ff25), it looks like Mozilla adopted the F12 shortcut key to toggle the dev tool from IE. I like it.

[+Red King Subscriber²]

I seriously don't get why Firefox is pushing cheap homebrew solutions for PDF and (please let me be wrong on this) soon SWF

 

Just read all the one star reviews here to get just how cheap this homebrew actually is,

https://addons.mozilla.org/en-US/firefox/addon/pdfjs/reviews/

 

I had the issue where government forms were not rendering properly - wasted lots of time to install latest greatest version of Adobe Reader, then half hour later, it hit me, Mozilla blocks Adobe Reader by default.

 

EDIT:

Just did a test in Nightly.

Adobe Reader visible installed and ACTIVE.

Nightly still uses cheap homebrew.

2GB RAM to open a PDF.

Text visibly gets populated on a 4.5Ghz Quad Core.

EDIT: 2.5 Gb

EDIT2: Crashed - apparently installer still doesn't detect proper CPU architecture - 32 bit mem limit reached.

[+Zlip792 MVC]

I seriously don't get why Firefox is pushing cheap homebrew solutions for PDF and (please let me be wrong on this) soon SWF

 

Just read all the one star reviews here to get just how cheap this homebrew actually is,

https://addons.mozilla.org/en-US/firefox/addon/pdfjs/reviews/

 

I had the issue where government forms were not rendering properly - wasted lots of time to install latest greatest version of Adobe Reader, then half hour later, it hit me, Mozilla blocks Adobe Reader by default.

 

EDIT:

Just did a test in Nightly.

Adobe Reader visible installed and ACTIVE.

Nightly still uses cheap homebrew.

2GB RAM to open a PDF.

Text visibly gets populated on a 4.5Ghz Quad Core.

EDIT: 2.5 Gb

EDIT2: Crashed - apparently installer still doesn't detect proper CPU architecture - 32 bit mem limit reached.

 

I think what Nightly has pdf.js, its newer than what offered as add-on, I could be wrong. But I rarely use Firefox as a PDF Reader, I go with PDF-X Change Viewer.

[The_Decryptor Veteran]

What on earth kind of PDF file are you opening that takes up 3GB of RAM?

[Menge]

Wow I have never had issues with PDF viewing with the built in PDF.js viewer!

[ViperAFK]

I haven't had any issues with it either, but I don't use pdf's often and usually the ones I do open are simple. pdf.js not handling large complex pdf's well does not suprise me, its only meant to be a basic viewer.

[+Red King Subscriber²]

What on earth kind of PDF file are you opening that takes up 3GB of RAM?

Scrolling a 500 page UNM catalog.

It is just text BTW and well you know should like 50 megs to open and view.

 

But it is more of a problem is that Firefox does not properly detect Adobe PDF Reader plugin and enable it causing these problems.

[ViperAFK]

Yeah, pdf.js is nice to have when the user doesn't have a pdf reader installed, but if the user does have a pdf plugin installed it would make more sense if firefox had that take precedence over pdf.js.

 

Its not too hard to disable pdf.js though, I believe you can just set pdfjs.disabled to true in about:config.

[(Account no longer active)]

Its not too hard to disable pdf.js though, I believe you can just set pdfjs.disabled to true in about:config.

Tools (menu) --> Options --> Applications (tab) [easier user-friendly method] is where it can be changed also.

 

I personally find pdf.js slow (when compared to SumatraPDF and even Adobe Reader). Users of Firefox should have been notified better about the change and be presented with an option to easily switch between pdf.js and their default PDF plugin.

[+Zlip792 MVC]

In this list, I censor many bugs which affect Release Management, Build Config, Firefox OS, Android, Developer Tools, Thunderbird, Test fixes, Test failure during try builds etc.

If I miss some please forgo them, I am human and can make errors. LOL!

From mozilla-inbound:

 

Electrolysis bug - Loading about:newtab in remote process - https://bugzilla.mozilla.org/show_bug.cgi?id=907342

Electrolysis bug - MacOSX only Focusing on content area - https://bugzilla.mozilla.org/show_bug.cgi?id=908453

Electrolysis bug - Geolocation Permission box - https://bugzilla.mozilla.org/show_bug.cgi?id=908692

JS Engine - Old bug - https://bugzilla.mozilla.org/show_bug.cgi?id=517765

HTML5 Parser bug - Old bug - https://bugzilla.mozilla.org/show_bug.cgi?id=528863

WebRTC - https://bugzilla.mozilla.org/show_bug.cgi?id=906105 , https://bugzilla.mozilla.org/show_bug.cgi?id=880067

Crash fixes - https://bugzilla.mozilla.org/show_bug.cgi?id=904147 ,

Exact Rooting for Generational GC - https://bugzilla.mozilla.org/show_bug.cgi?id=908483 , https://bugzilla.mozilla.org/show_bug.cgi?id=908825 , https://bugzilla.mozilla.org/show_bug.cgi?id=908891

Layout bug - overflow:hidden - https://bugzilla.mozilla.org/show_bug.cgi?id=906199

Header cleanup - https://bugzilla.mozilla.org/show_bug.cgi?id=908671 , https://bugzilla.mozilla.org/show_bug.cgi?id=908180 , https://bugzilla.mozilla.org/show_bug.cgi?id=908724 , https://bugzilla.mozilla.org/show_bug.cgi?id=908778 , https://bugzilla.mozilla.org/show_bug.cgi?id=908746

DOM Workers - Reducing workload from main thread - https://bugzilla.mozilla.org/show_bug.cgi?id=903772

DOM Cleanup - Binding related - https://bugzilla.mozilla.org/show_bug.cgi?id=887533

Regression fix [bACKOUT] - Fix scrolling of about:addons content - https://bugzilla.mozilla.org/show_bug.cgi?id=908006

Transition bug to C++11 standard - https://bugzilla.mozilla.org/show_bug.cgi?id=784739 (few more patches landed)

WebGL 2 Bindings cleanup - https://bugzilla.mozilla.org/show_bug.cgi?id=908662

Theme fix under special circumstances - https://bugzilla.mozilla.org/show_bug.cgi?id=898955

Tabbed Browsing bug fix - https://bugzilla.mozilla.org/show_bug.cgi?id=894048

Odin Monkey (asm.js) - https://bugzilla.mozilla.org/show_bug.cgi?id=908813

JS Engine Crash fix - https://bugzilla.mozilla.org/show_bug.cgi?id=908472

JS Engine bugs - https://bugzilla.mozilla.org/show_bug.cgi?id=899712 , https://bugzilla.mozilla.org/show_bug.cgi?id=900144

AES GCM cipher suite - https://bugzilla.mozilla.org/show_bug.cgi?id=880543

navigator.DoNotTrack fix [bACKOUT] - https://bugzilla.mozilla.org/show_bug.cgi?id=887703

HTML Input Element fix - https://bugzilla.mozilla.org/show_bug.cgi?id=905240

Cache GLContext - https://bugzilla.mozilla.org/show_bug.cgi?id=905161

Tokenizer fix - https://bugzilla.mozilla.org/show_bug.cgi?id=662669

Fixing #include dependence bug - https://bugzilla.mozilla.org/show_bug.cgi?id=908351 , https://bugzilla.mozilla.org/show_bug.cgi?id=908530 , https://bugzilla.mozilla.org/show_bug.cgi?id=908576 , https://bugzilla.mozilla.org/show_bug.cgi?id=908837

[bawz]

a lot of those Landings only affects Windows an not Linux yer? 

[The_Decryptor Veteran]

A lot of that stuff looks cross platform, so yeah Linux would get it.

[+Zlip792 MVC]

a lot of those Landings only affects Windows an not Linux yer? 

As Decryptor said they are cross platform.

To judge more precisely, I might have to read all bugs in details especially their comments which I avoid most of the time. LOL!!!

[bawz]

 

Chrome recently got some support for extensions to access the downloads API. The designer of this API at Google kindly informed us about it.

It now seems possible to do a rudimentary, albeit stripped down version of DownThemAll!, using Chrome?s new capabilities. Advanced features, like segmented downloads, proper checksumming, resuming etc. still do not seem feasible due to technical limitations. However, stuff like filters + OneClick seems to be within reach now. Basically, this new API enables extensions to use the Chrome download manager, but doesn?t give control over the request, data streams and/or low-level details themselves.

This still doesn?t mean you can expect a Chrome version anytime soon, if ever.

Right now there are simply more pressing priorities like getting 3.0 out of the door, and maybe doing a Firefox for Android version.

Also it is not clear if doing a Chrome extension is possible at all, without first actually experimenting with the new Chrome API(s). And after all, doing a chrome version means coding something up from scratch ? code sharing is pretty much not directly possible, if alone for the lack of Javascript esnext support in Chrome which the DTA code base already uses extensively in the Mozilla extension ? in an environment we?re not very familiar with.

Having said all that, it should be noted that I?m still very much a mozilla fanboi? in the light of recent revelations maybe more than ever.

http://www.downthemall.net/latest/still-no-google-chromechromium-support/

[bawz]

As Decryptor said they are cross platform.

To judge more precisely, I might have to read all bugs in details especially their comments which I avoid most of the time. LOL!!!

thats good, hopefully it speeds up Firefox on Linux

[cetla]

So is the ultimate goal of Servo is to replace Gecko?

[+Zlip792 MVC]

So is the ultimate goal of Servo is to replace Gecko?

Might be..

But there are several benefits in doing so.

Like your own language standards... memory management like you want no more reliance on stuff like jemalloc, tcmalloc etc. Similarly no more compilers issues...

[The_Decryptor Veteran]

I get the feeling Servo is more of a test bed for highly parallel work, stuff like CSS selector matching in multiple threads, etc. It might end up replacing Gecko but it's way too early to tell (Can't even do Acid1 yet)

 

Also, the nightly builds finally fix the most annoying bug, where the password manager wouldn't attach to "script generated forms", which are becoming more prevalent for whatever reason,

[cetla]

@the_decryptor

 

Actually, Servo passed Acid1 just a couple of days ago and moving on Acid2.

 

https://groups.google.com/forum/?utm_medium=App.net&utm_source=PourOver#!topic/mozilla.dev.servo/WzBBITnZWMc

[The_Decryptor Veteran]

Heh, I was reading that news group like 3 weeks ago, didn't realise they'd move so quickly.

[bawz]

http://www.mozilla.org/en-US/firefox/24.0beta/releasenotes/ if anyone else is looking for the changelog here it is for 24. 

[.stan]

https://people.mozilla.com/~prouget/appmanager.webm