Firefox 28 Release Notes | Hacker News Reader

Firefox 28 Release Notes

mozilla.org

252 points01Michael10158 comments

Interesting to see that now the current stable versions of Chrome and Firefox have dropped spdy/2, which is the only version of spdy supported by the current stable version of nginx. And nginx is the server used by more than 90% of hosts supporting spdy[1]. I presume this means it is mostly a Google/Twitter/Cloudflare thing for now since they seem to all be running spdy/3.1 in production.

[1] http://w3techs.com/technologies/segmentation/ce-spdy/web_ser...

I don't think people that are running Nginx with spdy are doing so with the stable version. I started using spdy back when it was a patch you had to manually apply, before compiling Nginx.

Count me as one of those people who doesn't like to have to work around backwards compatibility breaks when installing security patches on production servers. :-)

(edit: clarified wording)

There's always HTTP 1.1 available for the people that don't want the burdein of being on the leading edge (yep, I'm on that team).

Anyway, doesn't spdy/2 have a huge security flaw in certificate management? I don't remember it well. If that's the case, it's good that they droped it.

These days you just have compile it with --with-http_spdy_module in configure step. (for the current 1.4.6-stable). We were about to go into production but will hold it off.

Also the current mainline 1.5 version has implementation of spdy/3.

Chrome has also dropped support for SPDY/2 https://en.wikipedia.org/wiki/SPDY#Protocol_versions

Speaking as a Chromium SPDY & HTTP/2 developer, we are very much focused on standardization. SPDY is an experimental protocol meant to drive the standards process, not become a de facto standard itself. Therefore, it's critical for us to kill off old SPDY versions.

For more discussion on this topic, please refer to these discussion threads: https://groups.google.com/forum/#!topic/spdy-dev/_uvxTJkeCP0 https://groups.google.com/forum/#!topic/spdy-dev/XDudMZSq3e4

And note that your reference is slightly misleading (I don't think you intended this). It's true, the vast majority of hosts supporting SPDY are running nginx. In practice though, the vast majority of these hosts are Cloudflare or WordPress.com hosted sites. Both run newer versions of nginx with SPDY/3.1 support.

Right, so my point stands that it is interesting that current stable browsers don't support current stable web servers. So unless someone has the resources to test and deploy unstable web servers, it effectively means you shouldn't bother with spdy at this point.

I think it's a fair assessment that supporting experimental technologies requires more engineering resources. Everyone has to do the cost/benefit analysis themselves.

You seem intent on arguing about something I have never contested.

I just made the (implied) observation that anyone deploying nginx shouldn't even bother enabling spdy in 1.4 (stable) because effectively nothing will use it anymore. It was something that people who use "stable" software were able to benefit from starting in May 2013, but it is now no longer and won't be again until nginx reaches 1.6. That is, unless spdy/3 is dropped in favor of spdy/4 by then.

(edit: fixed date)

I think for people who actually want spdy3 support, they will evaluate whether nginx 1.5 is stable enough for them. Stable is just a word, and unless it's a straightjacket "stable or nothing" culture you're referring to, yes people will have to make their own decisions about whether 1.5 is stable enough for their needs.

As the posts below, "mainline" is considered stable enough for production.

A rose by any other name?

Well, the difference is that breaking changes are made to 1.5 so when a security update comes out (like today) you have to do regression testing instead of just grabbing and building the latest version.

Er, I thought I was agreeing with you. What do you think I'm arguing? To be clear, I view these two statements as grounded in the same logic, although perhaps one is more strongly worded than the other:

Yours - "So unless someone has the resources to test and deploy unstable web servers, it effectively means you shouldn't bother with spdy at this point."

Mine - "I think it's a fair assessment that supporting experimental technologies requires more engineering resources. Everyone has to do the cost/benefit analysis themselves."

Let me rephrase then:

Don't even bother building nginx 1.4 with spdy support or configuring it since no one can use it. From May 2013 until now there was a benefit to end-users. There no longer is.

If you want to provide the speed benefits of spdy to users, you now need to run unstable nginx or mod_spdy.

Right. I think the point is that you shouldn't be relying on experimental standards for one's web architecture unless you recognize its experimental status.

As an outside observer, it seems like this kind of pattern happens pretty regularly on HN. It's always kind of funny. There seems to be two distinct flavors:

Pattern 1:

    > Statement
    >> Counterstatement!
    >>> You're saying the same thing.
    >>>> I am? Oh. I *am*. Whoops!

Pattern 2:

    > Statement
    >> Agree with tone of disagreement.
    >>> Disagree on disagreement, while offering up facts to somehow agree *harder*
    >>>> Rebuttal, optional restatement of disagreement, optional statement(s) further
    >>>> reinforcing the common point on which we just can't agree to agree

The 'mainline' branch of nginx which supports spdy/3 is not considered unstable and can be used in production.

Source: http://forum.nginx.org/read.php?2,221377,221390#msg-221390

Except when a security patch comes out and the API changed between security releases and you have to potentially update configuration and fix third-party modules.

Nginx has had spdy3.1 support since late Jan, 2014, starting with nginx 1.5.10 (don't use that version, read on).

The current nginx 1.5.12 has patches to fix an overflow issue with the spdy support so it's gotten at least some attention paid to it since the feature was released. I expect spdy3.1 support will only get better as more and more people move to the latest version of nginx.

Check http://nginx.org/en/CHANGES for the changelog.

From what I can tell about the exploits found at Pwn2Own, there were a total of 4 critical. Looks like FF28 addresses all 4. That was a very quick turnaround.

The fact that FF runs in a single process seems to be a major security issue since the web views are not sandboxed. I know Mozilla is working on their multiprocess project Electrolysis https://wiki.mozilla.org/Electrolysis that is experimental in FF 30. Hopefully, this project will help harden FF security.

"The fact that FF runs in a single process seems to be a major security issue since the web views are not sandboxed."

Only to people who don't really understand security. This whole "multi process" "sandboxed" meme started by chrome guys has really got out of hand.

I suppose I was mislead by the mozilla wiki link I posted which states at the top "The goal of the project is to run web content in a separate process from Firefox itself. The two major advantages of this model are security and performance. Security would improve because the content processes could be sandboxed"

The effect is overstated & overrepeated, especially with facilities like e.g. ptrace being available to processes on linux

No, ptrace() is not available to Chrome's sandboxed processes. Not in the SUID sandbox and not in the BPF seccomp sandbox.

This stuff is described at http://code.google.com/p/chromium/wiki/LinuxSandboxing - if after reading that you still don't think it provides a significant additional layer of security, it would be interesting to hear what flaws you see.

> The fact that FF runs in a single process seems to be a major security issue since the web views are not sandboxed.

Running content in multiple processes, by itself, doesn't help security at all. It does help stability, since if one process crashes the rest of the browser need not; that's particularly important if running plugins like Flash, which crash in a stiff breeze. But if untrusted content manages to exploit a separate browser process that doesn't have any additional sandboxing applied, there's nothing standing between it and the rest of your system.

Additional security, however, comes from then applying sandboxing features to those separate processes, following the principle of least privilege.

I suppose there would have been a better way to phrase that sentence. The goal of the Electrolysis is to render the web content in separate sandboxed processes to improve stability and performance.

Unfortunately the bug reports are still 'secret' - https://bugzilla.mozilla.org/show_bug.cgi?id=982906

The autoupdaters are still throttled so not everyone is on Firefox 28 yet.

Running in a single process (not strictly true; addons are in another) has a lot of performance benefits though. I can get Firefox up to ~1.2GB with a few hundred tabs; Chrome uses that much memory for about 20.

Also, splitting process is bogosecurity like "I run SSH on a non-22 port". I thought the general consensus here is that security by obscurity is bad.

> I thought the general consensus here is that security by obscurity is bad.

How is running web views in separate processes "security by obscurity"? I agree that there is nothing inherently secure about it, though it has other benefits, but it doesn't increase obscurity as such (you could argue that the need for well-defined message-passing semantics leads to more decoupled code, decreasing 'obscurity').

Eagerly waiting to ditch Chrome & its forced notifications. Perhaps with Australis (Firefox 29), the UI would be at par.

Edit : Chrome feels subtle, lighter & perceptually faster (specifically : tabs, scrollbars). Although after updating Firefox today, I can notice that there's improvement in Firefox.

One major (okay maybe not that major but an important factor in my decision) issue for me with Australis is the separate boxes for URLs and search. Every other browser merged them and I don't know of any reason to keep them separate - is there one?

It's admittedly a corner case, but I've searched phrases that contain URLs and have had them get treated as a URL rather than search phrase in Chrome. In that case, I just drop to google.com and issue the search from there.

I actually run into this one (or its opposite, a URL getting treated as a search) fairly frequently, which is pretty annoying every time.

ctrl-k works as "go to search box" in both browsers. For Chrome, that means focussing on the address bar and presetting it to "? ".

Nifty. That would certainly do the trick. Thanks!

If you prefix your query with a ? it will be sent to your default search provider instead.

Thanks for the tip. I was unaware. But of course, if I have to think about it, it's failed its primary function of pretending I shouldn't have to care :-P

You know you can remove the search bar, right?

Right click in the tab area, select "Customize" and then drag it to the available icons panel. Searching works from the address bar too.

It is also probably there still because you can select where you want to search, so you can pull down the menu and search in Twitter, or Amazon, eBay, Google translate, Google maps, The Pirate Bay, Wolfram|Alpha, YouTube, ....

If your search bar is your URL bar, then all your keystrokes (even when typing URLs) go to your search provider to provide autocomplete of search results. In Firefox you can type URLs without having them sent to a third-party. We think this is important. (You can get autocomplete search results by typing in the search bar.)

You could handle this with configuration, as others have.

Edit: judging from the other comments, maybe you have already.

No, you can not.

If you don't have a separated search bar, your only options are enabling or disabling autocomplete. There is no option of using autocomplete just when you want it.

Also, the local autocomplete of the address bar is great. Count me as another one that does not want to miss it.

I was letting him slide with exactly that: an on/off switch. But with additional configuration one could use a 'g ' prefix to activate autocomplete via Google, for instance.

This is why it surprised me that, as I understand it, the Firefox OS homescreen search bar seems to take exactly the opposite approach: every keystroke is sent to some obscure for-profit private company in Israel. Do you know what's up with that?

There is. Mozilla doesn't want data from users (unknowingly) being sent across networks to a server by default.

The urlbar is already capable of searching though. It just doesn't give you suggestions for the reason above.

A few areas of impact, likely many more.

Presumed privacy - Some prefer to not send their direct traffic and history title strings through a google search input.

Usability - search engine autocomplete/suggest can interfere when typing a page title from your history.

Money - traffic volume referred to search engines is significantly increased by merging URL and search inputs.

Edit: wow, parent had zero responses when I started typing this!

This is to protect user privacy as explained by a Mozilla dev: https://news.ycombinator.com/item?id=6756582

Otherwise every keystroke you made would be sent to Google.

Uhm it could just search when hitting enter. No need to autosuggest.

This already work. Remove the separate search bar and you are ready to go.

dude! :D

If you remove the bar and press Ctrl+k you are taken to google in the current tab... not cool

But why would you press Ctrl+K if you've removed the search bar?

Stuff like this is why I continue to use Firefox.

You can turn autosuggest off

And you can install Omnibar or Instantfox if you want a combined address bar and search bar.

Or just customize the UI and remove the search bar if it bothers you so much. Right-click the toolbar in <29, click the menu button and choose "Customize" on the bottom left in >=29.

Assign keywords to different search engines.

This lets you do queries like such:

Type "g hello" into the url bar - googles "hello"

wa integral from 0 to 1 of ln(x^4)dx - can be set to wolframalpha

w France - my wikipedia one

Thanks glad to see an explanation and a good one too.

You can do searches directly in the URL bar and it's not exactly the same, but if you want OmniBar-like behavior there's a plugin for that (when I used it, it worked well): https://addons.mozilla.org/en-US/firefox/addon/omnibar/

What I like about Firefox (versus Chrome specifically) is that the AwesomeBar gives much better suggestions from your history, doing full-text searches in titles - within Chrome I can rarely find something in my history without opening the history view and seriously, most searches that I do are for things that I've already seen and it saves a lot of roundtrips to Google (hint, hint), plus searching for things you can barely remember on Google, which you end up doing in Chrome, is like searching for the metaphorical needle in a haystack.

Plus, as others said, I like it when the browser cares about my privacy. On Android on the other hand there's only a single bar due to a lack of space, but it asks you if you want to activate Google's search suggestions.

> within Chrome I can rarely find something in my history without opening the history view

Yes. And bookmarks, for me. Back when I had switched from Firefox to Chrome, I thought it hadn't imported most of my bookmarks. Chrome would hardly ever show me the things I expected/wanted.

Having switched back to Firefox, I'm delighted with the experience of it actually suggesting the things I want, again.

The only time I use Chrome these days, is for something Google can't be arsed to make work correctly in other browsers (e.g. photo editing on G+, label dropdowns in Gmail, ad nauseum).

the awesome bar might be the single biggest technical reason for me to use firefox. its fast and the search is very accurate (since its searching sites you previously visited)

Like many others I have long since removed my search box. I also use keyword search [1]. If I want to search google I write "g hacker news", if I want to search Wikipedia it's "we thing", for stackoverflow it's "so great", and so on.

[1] http://kb.mozillazine.org/Using_keyword_searches

There is, yes.

You have three somewhat simple options for the search/url thing:

1) Have a single box, do not offer search suggestions. 2) Have a single box, offer search suggestions for the stuff users type in it. 3) Have separate boxes; offer search suggestions for the stuff typed in the search box but not the stuff typed in the url box.

#2 involves sending everything the user types in the url bar to the search suggestion provider to check for suggestions

Firefox does #3, because #2 is not considered acceptable from a user privacy perspective.

There are probably more complicated options here like only offering search suggestions if the string typed so far doesn't look like a URL or something that might protect user privacy better than a straight up #2. But it gets easy to screw it up and leak private user data to the search provider....

Another option is to prefix searches. If I type "google " then start offering search suggestions. Chrome even has this, and a nice UX for it. It just unfortunately also has unprefixed searching.

We can also make it bash compatible, and have all the GNU tools available at the address bar... (find www.wikipedia.com --title GNU)... But I don't think most people will be able to use it.

>only offering search suggestions if the string typed so far doesn't look like a URL

How would this work? Lets say I wanted to search for Wikipedia, perhaps because I know there's news about it. How would Firefox know I was going to type "Wikipedia" and not "Wikipedia.com"?

Yes, it's things like that that make this approach complicated and failure-prone. Which is why Firefox is not doing it.

Personally, my reason for preferring them separate is that I use the address bar to search my browsing history, which doesn't work if it's cluttered up with search results and prior search history.

The perfect examples I use for this: when I'm trying to use a local hostname from my hosts file (e.g. 'devserver') and it googles it unless I type http:// in front, or when I'm trying to google an IP address and I have to go to the Google home page first.

You can add special characters to your address bar search to restrict it to only browsing history, or bookmarks, etc. There is a list of them here: https://support.mozilla.org/en-US/kb/awesome-bar-find-your-b...

For example, "^ hacker news" will only search your history, and "* hacker news" only in your bookmarks.

I have the omnibar add on and have Firefox (nightly actually) configured so that searches in the bar first look at bookmarks and the history. If nothing shows up there, then it searches DuckDuckGo, and if I don't see anything there, I fall back to Google.

I have custom search plugins for work stuff that are absolute lifesavers when combined with the following addons:

Context Search (adds a context menu dropdown when you right click highlighted text; think plaintext words you search in trouble ticket type stuff) https://addons.mozilla.org/en-US/firefox/addon/context-searc...

Add To Search Bar (adds an option to search-engine-ify any submit text box) https://addons.mozilla.org/en-US/firefox/addon/add-to-search... (might as well mention https://addons.mozilla.org/en-US/firefox/addon/organize-sear...)

Bonus productivity favorite: https://addons.mozilla.org/en-US/firefox/addon/tree-style-ta...

So yeah, you'll pry Firefox away from my cold dead hands.

What problems do you have with the current firefox UI?

Chrome feels lighter & perceptually faster (mostly psychological) because of a flatter UI (specifically the tabs, scrollbars). Although after updating Firefox today, I can notice that there's improvement in it.

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

I switched from Chrome to Fx this week. So far the only annoyance really is that flash elements on a webpage can takeover keyboard focus. I'm used to KB navigation and opening a youtube tab can really mess with your flow.

The first thing I look for in a browser is a way to switch through tabs in last-used order. If doesn't work out of the box, then I personally don't have a reason to investigate further. If this was a checkbox feature in Chrome (or Firefox, whichever's first), I'd probably never try anything else.

Can you explain why you want window manager features in your browser? If you use windows-in-a-virtual-desktop instead of tabs-in-a-window, they're LRU by default on pretty much every OS.

Because I rarely dedicate a virtual desktop to Firefox alone, and I enjoy the flexibility of a deeper hierarchy (virtual desktops - windows - tabs). Everyone has their own needs and usage patterns. Are you also opposed to file browser and terminal window tabs?

Unfortunately I cannot really justify my interest in philosophical or design-oriented terms. I had simply gotten used to the tab switching behavior in Opera, and have missed that behavior a great deal since having to switch (for reasons that should be obvious).

Actually, this feature works out of the box in Firefox. You just need to turn it on by changing browser.ctrlTab.previews value to true in about:config.

Actually, I'm not sure that it does. Or perhaps I'm not adequately describing the behavior I'm looking for. Thanks in any case.

Is that feature being out-of-the-box, as opposed to available through an add-on, really that important? I love Tree Style Tab (https://addons.mozilla.org/en-US/firefox/addon/tree-style-ta...), but I don’t really mind that it’s not built in to Firefox. It’s easy enough to install on a fresh computer, and the default tab bar is useable enough for short periods.

Anyway, as someone else said, setting browser.ctrlTab.previews to true in about:config should work. For me, it successfully changes the behavior of Ctrl-Tab to most-recently-visited order, and it causes hitting Tab while keeping Ctrl held to pop up previews of the tabs, just like Alt-Tab does on Windows. If you still can’t get that to work, there is also an no-restart add-on for what you want, MRU Tabs: https://addons.mozilla.org/en-US/firefox/addon/mru-tabs/.

I'm generally alright with add-ons that are Mozilla (or Google) 'branded', and those that are particularly well known otherwise. I also appreciate that the two you mentioned are listed as having been 'preliminarily reviewed' by Mozilla. I wouldn't go as far as to call the situation ideal though.

You're right about browser.ctrlTab.previews in any case - I must have only just tried option-tab, not restarted, or something similar when I responded to the other comment. How one is expected to dig up that option is a mystery to me though. Is it really such an unusual request?

Australis is a huge step backwards for firefox.

If you want chrome (or opera, which is little more than a chrome clone these days), you know where to find it. Firefox is supposed to be the customisable browser that gives the user the freedom to have what they want and not what Big Google decides they do.

And: https://developer.mozilla.org/en-US/Firefox/Releases/28/Site...

They have more detailed information than the release notes.

Well, there's also stuff that you can't see in the "release notes".

I found that Firefox 27 had huge speed improvements in canvas 2D compared to Firefox 26, at least in Linux without hardware accelerated canvas.

I couldn't find anything related to that in 27's release notes, sooo (still happy though!).

In case you're curious:

http://arewefastyet.com/

https://areweslimyet.com/

Any idea why the "slim" numbers are moving steadily in the wrong direction?

I'm wondering too. FF13 seems the best with "slim" numbers.

Changes in browser engine and method/elaboration in memory measurement. And possibly a lot of regressions...

Problem is, this does not convince Average Joe about the slimness of Fx. I asked them to set up a main chart using only the stable ones with the latest changes (allocation), but it was ignored. IMHO that would help them see the what the end user would see over the long course, and A. Joe would see what actually effects him (as he is unlikely to use dev channels).

Edit: the charts also do not show the improvements in Garbage Collection (IGC in Fx16, upcoming ICC and GGC) and the memory management with add-os (Hueyfix in Fx15).

Tiny regressions are hard to track down, but over time they add up.

Some of it is the move from single to multi process/thread tab and rendering model. And if I'm remembering correctly a few of the new HTML5 elements added have caused significant (+20-30mb) leaks as they are, as of yet, unoptimized in the rendering code.

There's no multiprocess content code enabled by default yet, so that's not it.

Not process but I could hsave sworn there are multithreaded processes enabled by default in Aurora, image processing for example.

It's enabled by default in version 30, which is Aurora right now.

Controlling memory usage is a battle that requires constant vigilance, and it's hard. It's difficult to measure memory usage in a reliable, repeatable way such that we could use it to prevent regressions. Additionally, small regressions can add up over time, making it hard to pinpoint any one thing that could be backed out to improve memory usage.

Partly it's just that the browser is getting more features. Things like WebRTC are a ton of code.

More generally, AWSY is useful for identifying some regressions, but it's one specific, unrealistic workload. You can't use it to judge Firefox's overall memory consumption.

For example, looking at AWSY you'd think Firefox 13 had the best memory consumption. But I guarantee you that Firefox 28 has better memory consumption. In particular, Firefox 28 has fewer bad cases where memory consumption spirals out of control, and those are the cases that really hurt performance and stability. Using 10 or 20% more memory at start-up, for example, isn't a big deal in comparison.

Nonetheless, the increase isn't good. I filed https://bugzilla.mozilla.org/show_bug.cgi?id=986323 about this, and have started investigating.

Anyone have a good breakdown of the differences in these benchmarks? i.e. Why does Chrome dramatically lead the(ir) Octane benchmark?

Because Octane was written by Google. So it's either:

  -representative of things Google knows it's faster at
  -representative of things Google thinks should be fast and prioritizes... and is thus faster.

or a little of column A little of column b in a feedback loop.

Octane is the continuation of V8 benchmark (and contains all of the V8 benchmark tests) — V8 (the JS engine) was specifically designed to provide good performance on those tests (this is why V8 had a generational GC from launch, for example; though somewhat amusing last I checked Carakan (Opera/Presto's last JS engine) was quickest at splay, primarily a GC benchmark, despite having a very naïve GC!). The tests new in Octane don't display such a clear disparity, as V8 wasn't designed for them.

hint: Splay is a benchmark that penalizes generational GCs. If you have a fast mark-sweep GC with an aggressive growth policy you are going to totally rock it.

I may slightly be showing how long it's been since I looked at Splay. :)

My memory was that originally V8 was out-performing everyone else primarily because the initial allocation (into the nursery) was way quicker than anyone else did. Or maybe it was freeing the nodes that die young?

Certainly I remember both SpiderMonkey and Carakan spending huge amounts of time on GC on Splay, and Carakan's massive speed-up came when the object representation was changed (GC actually asymptotically regressed as a result, but in almost every practical case cache-locality (and reduction of memory accesses) outweighed it).

In regards to rendering performance, we have this internal app that's pushing a lot of points into timeline charts rendered with HighCharts. When all the datapoints are rendered, Chrome has a lot of latency but is still usable, whereas Firefox chokes. I see that in Australis it's much improved, but still less usable than Chrome.

This is awkward since in general I find Firefox's performance to be on par with Chrome, sometimes slightly better, but these charts make me open Chrome and it's a pity since I prefer Firefox.

Maybe HighCharts is not optimized properly for Firefox and of course, those charts being a work in progress are not optimized. But I do wonder, what's the difference here? Is it something about hardware acceleration?

I've found Chrome's graphics performance in general to be better than Firefox's, specifically with regards to canvas and SVG. Not sure the reason, but my guess is just lack of optimization.

Chrome can make graphics calls in a separate thread, whereas I think Firefox still makes graphics calls on the main thread.

Internal apps are hard for Firefox devs to debug, but I've had success in the past debugging problems in internal apps by using a test account where the developers have emailed the login details to me privately.

If you are able to provide such a test account, please file a bug at https://bugzilla.mozilla.org/enter_bug.cgi; use "Core :: General" in the "find product" box; and don't worry too much about most of the fields, the title and description are the most important; and also add ":njn" to the "CC" field. Say something like "I can provide login details privately".

If you can do that, I'll try to find the right person to address it.

Multi-line flexbox. FINALLY! I shall drink champagne today.

Includes important security fixes for vuln's disclosed at Pwn2Own, like this arbitrary heap write from JS:

https://www.mozilla.org/security/announce/2014/mfsa2014-31.h...

Super excited about better codec spread. What do they mean by volume control, though? We've been able to control the volume of HTML5 video/audio for a while now.

Edit: here's a comparison of the audio UI between versions: https://mediacru.sh/2528762308f7, looks butt-ugly imo

Not in <audio> default controls in Firefox. <video> had it, though.

There was no UI for it in the default browser controls for the <audio> element. Compare http://cs.iupui.edu/~aharris/n351/notes09/HTML5AudioDemo.htm... in Firefox 27 and Firefox 28, for example.

The relnote is presumably talking about https://bugzilla.mozilla.org/show_bug.cgi?id=649490

Looking for Firefox 29 beta. And nice to see that VP9 implemented in Stable release.

Is Australis due out in the next version?

Had swapped back to using the default version from nightly in anticipation, but it seems that was a bit premature on my part.

It goes Nightly -> Aurora -> Beta -> Release

Yes. Australis will ship in Firefox 29, which will be pushed to the Firefox Beta channel this week:

https://wiki.mozilla.org/Features/Release_Tracking#Almost_De...

Yes, it's due with 29.

Why Firefox removes spdy/2 support? They don't want to support spdy/2 servers? Why?

PS: I'm not expert on servers :)

It got discontinued, and Chrome dropped it as well. Both support the newer SPDY/3.1

https://en.wikipedia.org/wiki/SPDY#Protocol_versions

https://bugzilla.mozilla.org/show_bug.cgi?id=912550

https://code.google.com/p/chromium/issues/detail?id=303957

While all browsers support http 1.0 websites almost three decades later.

Spdy really shows how fast people are willing to put immature technology in production as long as Google claims it better. Amazing.

What's even wise is the attitude towards abandoning published protocols and contracts. This is not how a lasting and stable internet was built.

This clearly shows why spdy can't be a candidate for http 2.0

Dude, you are comparing two entirely different things and no, it doesn't clearly show anything.

Spdy is not an established standard, it's an experimental protocol developed by Google to improve the transmission of data on top of HTTP. So yes, it is normal to have Google and Firefox drop the older versions to push forward on the iterations, it's an experiment to see how they can improve Spdy and in the end, how HTTP 2.0 can benefit from those results.

It is more irresponsible to keep and use an older experimental protocol that nobody will use.

HTTP 1.0 is an established standard agreed by everybody and that's why it hasn't changed in 30 years, something Spdy is not.

HTTP 2.0 working group is already establishing majority of Spdy's experiment results as a working draft for HTTP 2.0.

SPDY was an experimental protocol that's only used by evergreen browsers and servers, so there's no need for backwards compatibility.

To go along with this release, I just pushed out a new version of Firefogg (https://firefogg.org) with support for creating WebM VP9/Opus videos.

Is VP9 supported by all WebM-supporting browsers now?

(i.e. is it feasible to completely drop VP8 and use VP9 instead?)

The vp9 ncoders are too slow for this to be praticable.

Last version before they ruin their UI with Australis.

Best way to delay the pain (and that of their general too-frequent breakage) is to switch to the ESR version: https://www.mozilla.org/en-US/firefox/organizations/all/ , which should give enough time for addon writers to undo the australis damage.

Come on. It's not that bad. I know design is a subjective thing, but calling it a disaster is a bit too much, IMO.

I wouldn't care if they made it an option so people could choose "I want a chrome clone that doesn't violate my privacy" and get the chrome-ified UI, or choose either the 'Firefox menu' Office 2007-ish one, or the classic interface. Instead, they removed just about every customisability-related feature they could.

Spent 45 minutes playing with it after reading your comment and I absolutely love the new UI. Everything is so much more fluid and intuitive. The dev tools especially.

It might turn off some Firefox users but I really dig it.

Improved dev tools are a separate issue to removing every customisability option they could (small icons, status bar, the fscking ability to reorder toolbar items...)

<input type=number> didn't make it in?

https://developer.mozilla.org/en-US/Firefox/Releases/28

"< input type=color > and < input type=number > have been implemented, disabled by default."

9007199254740992 running in "random" mode with 0ms delay is still way faster in Chromium 32 than in Firefox 28. http://www.csie.ntu.edu.tw/~b01902112/9007199254740992/

Same there (try 10000 sites, click "Benchmark"):

http://www.raymondhill.net/voronoi/rhill-voronoi-demo1.html

The difference has narrowed though, FF used to be much, much slower than chromium.

I can definitely see a difference on this one, yes, about 30%. A question. Does "rendering not included" mean not including painting times, or not including the canvas API call times either? My profiler is showing a bunch of fill(), rect(), stroke(), moveTo() calls dominating the wall-clock time, but I'm not sure whether those are what's being measured here or not.

That benchmark measures only the computation of the whole diagram -- which is the sole purpose of the library. The rendering, if any, is left to the caller after the diagram has been wholly computed, so there was no point for me to measure that part.

Gotcha. I filed https://bugzilla.mozilla.org/show_bug.cgi?id=985262 on this; if there happens to be a standalone version of this benchmark that skips all the canvas stuff that would be pretty helpful.

Hmm. I see about the same thing with Chrome 35 dev and Firefox (28 or nightly) on Mac here, performance wise.

I am pretty sure he was joking. But there are cases though where the speed does matter. Cool d3 animations for example.

This was on Xubuntu 14.04 - Chromium was racking up score on random quite a bit faster than Firefox was. This is a squishy benchmark for all sorts of reasons ... but the difference was glaring.

It's entirely possible that platform-specific drawing paths matter here. :(

What's OpenSearch on mobile?

Can we finally add custom search providers like desktop version from http://mycroftproject.com/. It really seems to be an odd exclusion since sync takes care of everything else.

Anyone have issues using Tweetdeck on Firefox 28?

I get a scripting error, and it won't load it!

OSX notifications are going to be great for our internal web apps!

For other desktops, there is this: https://addons.mozilla.org/en-US/firefox/addon/gnotifier/

Firefox should allow you to type "45/3" in the address bar and be taken to Google with the calculation.

It does, when you prepend "google " in front of "45/3", which is in my opinion the best solution (I don't want to route all my URLs through google).

You can prepend it with just ? and it will go to your default search engine.

try it with a space, like "45/ 3" and it works

Thanks, that works!

If you type "45/3=" into the search bar and it will return the result without having to load the google page.

I’ll say it once. Multi. Line. Flex. Box.

Multi-line flexbox is finally in!

I wouldn’t mind if this was just version 27.1.

Firefox 29 hits the beta, that means MacOs users get tabs in title line, finally!

That makes me wiling to try Firefox again, since amongst the Safari/Chrome/Firefox it was the browser that consumed most vertical space, which is an issue on mobile (less than 15") laptops - at least for me.

Give it a try .. it looks great in OS X native full screen mode.

Reply on news.ycombinator.com