Firefox 3.1 Beta 3 Released
ink writes "Mozilla has released the third beta for Firefox 3.1 (which may become Firefox 3.5). This beta includes the new location bar, Mozilla's new JavaScript engine Tracemonkey, new HTML5 features and many other enhancements. It looks the same on the surface, but there are many changes under the hood."
They changed the location bar again.
Now I can watch people flip out about it on the interwebs for 6 months as well as being personally annoyed with re-getting used to how it functions.
You can't take the sky from me.
I was going to download this using Mozilla Firefox, but Microsoft told me it would be faster downloading, and it's returning a exception. Too bad I wanted to use it.
It's all fun and games till someone divides by 0. Then it's hilarious.
Maybe I'm missing it, but I don't see a release announcement (not even on a Planet Mozilla blog entry). It's also not available on auto-update yet.
Did Slashdot just interrupt their release process, again?
How pretty much everything we do uses JSON and until now there has been no love from the browser.
My question is, will all these new JavaScript goodies (both in Firefox and in IE8) get rolled into jQuery? That way if jQuery sees the browser can do JSON serialization, or timeouts on XHttpRequests, it will use the native stuff instead of emulating the behavior?
I'm gonna have to play with the VIDEO thing. The big problem such a new feature will have is codec support. Nobody is gonna transcode their streaming content to use this thing when they can just use flash player. That and I really dont want "normal people" trying to find codecs on google--most of the hits for "$AWESOME_CODEC" are usually just spyware installers.
...It looks the same on the surface, but there are many changes under the hood."...
Will Joe Public be in position to notice them? The new engine might be indeed faster but I wonder whether an ordinary user will see a difference.
What's the new location bar? Is it something like the old location bar, aka the UnAwesomeBar? I'm pretty much sick to death of the awesomeness of the present location bar, what with Slashdot being listed as "Server 500: Internal Error" in the dropdown because about 4 months ago I got a 500 error message?
More importantly, have the Firefox devs realised that downloads can fail yet? Or does it still report downloads as successful if they are interrupted before they reach Content-Length bytes, or how about reporting as successful because they successfully downloaded a 404 error page? Supporting incomplete standards and reimplementing an already existing feature is all well and good, but reaching version "3.5" of a web browser without having the most rudimentary capability of detecting a failed download is nothing short of obscene.
Choking would require said wang to be long enough to pass the teeth, I think firefox is safe from your wrath.
You can't take the sky from me.
I thought Firefox was going to be implementing the same type of preemptive threading and memory protection that Chrome and, I think, IE 8 have?
So far the latest FF beta all seem horribly slow with multiple pages. The more tabs the worse the overall performance.
Also, the latest FF betas still have the awful performance rot where overall performance degrades over time as you continue to open and close tabs.
After using Chrome for a while it is hard to keep using FF when I've been able to keep Chrome open for a couple weeks and it still feels as lightning quick as it was when first started up.
I know there is a tendency among some people to think of version numbers as decimal, since they use decimal points. I know I did when I was younger.
It's kind of annoying when major projects make this mistake though. It leads to all sorts of confusion when people see results like version 3.1.150 being after 3.1.50 and don't know why that's the case (".5 is more than .15!", which in the case of the Firefox release mentioned in TFS would be accurate, but in the case of properly-numbered software wouldn't), or other people truncate 3.1.50 to 3.1.5.
I wish major projects at least would use the traditional "increment by one" method. If it can be done for the X-Men 2.1 DVD (after nerds no doubt complained about the "X-Men 1.5" DVD), it can be done for Firefox et al too :).
"...always new atoms but always doing the same dance, remembering what the dance was yesterday." -Richard Feynman
is back!
It looks like they did. Firefox 3.1 beta 3 is still not available on the All Betas page, and when you click on the Download Now link on the Release Notes page, you get Firefox 3.1 beta 2.
The release linked to in the summary may not be the final, completed version, as Firefox 3.1 beta 3 has not been officially released yet. Download it at your own risk. You should wait until it's available through the links I give in this post.
What a fool believes, he sees, no wise man has the power to reason away.
I'm looking forward to them resolving the bit where the *nix Firefox builds performed slower than the win32 builds, supposedly due to Profile Guided Optimizations in javascript:
http://www.tuxradar.com/content/benchmarked-firefox-javascript-linux-and-windows-and-its-not-pretty
I just downloaded and am using it now .. it definitely feels faster, however, will it crash less on a Mac? :(
I love firefox, and I use it everywhere, but man, is it awful on a Mac.
Please don't flame, just an honest opinion from a long time firefox user/supporter/evangelist.
AirSpeak - http://itunes.com/apps/AirSpeak
Hey everyone - glad you're excited about the new beta, we're pretty excited to release it. We actually haven't finished the QA on the download page, the update snippets, etc, yet. What you're seeing here is that last night we started sending out the final bits to our mirror network. So yes, you could go get it directly off the FTP servers, but that can overload mirrors and make it hard for other people to download it.
We'd prefer if you waited a few hours until about 2pm PDT when we'll be ready to update:
http://www.mozilla.com/firefox/all-beta.html
which uses our mirror-rotation script to ease the load of downloads.
Mike Beltzner
Director of Firefox Development
Stop WINE'ing about Linux!!!
So after shoving a freaking DATABASE into Firefox 2, they're now adding a freaking VIDEO playback feature?!
Sorry, but first off, Firefox has supported video - well, since it existed. It's called a PLUGIN. That means that if I don't want it, I can not install it and save myself the bloat. No more.
Secondly, video?! From a web browser?!! I want something that can browse the web, I don't want Firefox to pretend it's freaking iTunes. If I want video playback, I'll install Flash. At least that will actually work with video on the web, rather than the bloated and useless Theora crap they're dumping into the browser.
On the upside, it's nice to see Firefox is finally supporting JSON. I was unaware that eval() didn't work in Firefox. Right, I know, "security." Except that you're already limited to loading JSON from the original domain, and if the original domain is serving up malicious script as JSON, you're already hosed.
That totally depends on the site - for someone such as yourself who is usually only using the web for viewing Microsoft's site, Firefox is the fastest, according to Microsoft themselves. Other people's mileage may vary, but hey, speed only matters if you don't care about things like viruses, worms, general crashes due to poor code, and overall crap. (Which is why I won't be using this beta until it's finished, any more than I'll be using IE.)
Why is it that every time a new version of Firefox comes out, we need to have a news story about it? Same with any other browser, but at this point it seems like the nightly builds are being posted about. It's getting ridiculous.
http://CryoLANparty.com/ A lan I'm staff on!
Everyone remembers FF devs flaming people in those FF memory leak stories from a few years ago. The anger comes from the fact they know they have a huge problem with the way FF is architected. Lashing out is a very common reaction from developers who are aware of some fundamental problem with their code that they know would require massive amounts of work they are unable or unwilling to fix.
The FF devs got away with it because they were compared to the horrible mess that IE was back then. Now IE has really gotten its shit together now with it great leaps forward with javascript performance, threading, and memory protection.
With Chrome and its incredibly clean and modern code base and extensions soon to arrive and the Linux version rapidly maturing, the only reason to keep using FF will be misplaced lingering fanboyism from the "IE sucks! I use FF so I'm cool" days.
Finally..finally!
Now I think I an transcode my snapshot video footage into a format that I don't have to worry about for ...at next 5-10 years.
Not feeling so smug anymore are you Firefox devs?
You had years to get your shit together and fix the stinking pile of shit that is the Firefox code base. Instead you clowns sat around in forums running your mouths off and flaming anyone who dared complain about the absurd resource leak and performance rot problems with Firefox.
So while you sat on your asses, Microsoft has gotten their shit together and Google has come out with a technological masterpiece with Chrome.
So now both Chrome and IE 8 have modern preemptively threaded and memory protected tab architectures and Firefox is the same stinking pile of monolithic shit it was years ago. One tab crashes, everything goes down. One tab sucking up performance, the entire UI degrades in performance, one tab leaves behind leftover memory and other resources and it never goes away until you quit the entire app.
Congrats retards!
Why don't you dimwits go grab the Chrome source and see how a modern browser is written. Dump the piece of shit Gecko layout engine that no one but you wants. And learn how a modern app implements threading and memory protection.
Don't worry, the Google guys working on Chrome are very friendly and will help you out with the hard to understand bits you Firefox devs obviously are too incompetent to handle.
Right you are. The decision to change version numbers was made in a meeting last Friday, and it was already pretty much a sealed deal going in.
Will they keep adding crap in that has remote relevance in order to make their product appear fresh until the whole thing (which already is a pretty good resource hog) balloons and becomes as slow as old IE?
Do I need a location bar? WTF is it anyway?
You do have a good point. There has been talk about supporting additional image formats (JPEG 2000, TIFF, MNG) using imagelib extensions. They could do the same for different video codecs, as well.
I notice that Firefox is an open source project, so all it takes is someone to come forward to do the work. I also notice that the Google Summer of Code will be starting over the next several months. Are there any students out the that want to make some extra $$$, get great software development experience, and add a significant new capability to one of the most popular browsers?
What a fool believes, he sees, no wise man has the power to reason away.
Here's the feature I want: bug fixes! Everytime I turn around there's a new Firefox packed full of new features I don't need. I wondering how rock solid it could be if they spent half that energy on fixing bugs. No new features until the bug queue is empty!
Don't blame me, I didn't vote for either of them!
The Minefield beta is really fast.
Why is it so hard to only have politicians for a few years, then have them go away?
Only when Google decides to shoot its revenue foot and release adblock, I might consider Chrome.
Firefox by itself - I'd be grateful if they scaled back. Do you remember the origin? There was this bloated hog called Mozilla Suite, and there was this little-known neglected wild branch called Phoenix, which was meant to be the Mozilla engine with a minimalistic, customizable frontend - cut on all the bloat.
And suddenly people switched en masse to the small, lean "just a browser" thingy while the monstrosity died.
Now Firefox becomes the new monstrosity full of bloat. It really needs another "phoenix branch" - something that will take all the lean mean backend stuff and do away with "awesome bars", "intelligent bookmarks" and all this cruft people don't give a shit about, and move it ALL to extensions, from which you'd take what you like while not being encumbered by all the rest.
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
1. Speed: Slowest browser on the planet. ... fun to ride until your friends see you on one.
2. Reliability: 1 tab crashes, they all crash -- and they will crash, guaranteed.
3. Coolness: FF is like a fat girl or a moped
I do like the massive plugin support, themes and such, but eye candy doesn't make up for the reasons I don't use FF anymore. I'm using Chrome right now, and on my work PC I am stuck with Windows, but I was able to get IE8, which seems (very subjectively) more robust to me.
Its a web browser, not the holy grail of computing. Its no more important, in the big picture, than an IM client.
Yep, I fully expect to be modded down for stating my sure to be very unpopular opinion here.
Couldn't they provide 64bit binaries? That would be very useful, at least to me...
I use Firefox on Linux as my primary browser. I'm having a huge problem with random slowdowns, however. It seems to be fairly random, exacerbated when multiple tabs are open, and possibly related to Flash. When the slowdowns start occuring Firefox will start eating 99% of CPU and become unresponsive. A strace will show dozens of gettimeofday() calls every second.
A google search for "firefox getttimeofday" will show many people with similar problems.
This is on CentOS 5.2 with the latest packaged firefox...
Anyone seeing similar problems?
the only reason to keep using FF will be misplaced lingering fanboyism from the "IE sucks! I use FF so I'm cool" days.
that, or the wealth of useful add-ons that makes FF more useful than the others you listed, esp Chrome.
The Firefox devs got away with it by fixing their memory problems. They made Firefox use less memory than other browsers. What was it they were unable or unwilling to fix again?
What a fool believes, he sees, no wise man has the power to reason away.
uh...looking at the Google Chrome team page, I can immediately pick out the following people as being ex-Mozilla employees or contributors: Ben Goodger, Darin Fisher, Brett Wilson, Peter Kasting, Mike Pinkerton, Jonathan Haas, Pam Greene, Jungshik Shin I'm sure there are more that I'm not aware of, but those are all certain.
The initial goal of the Firefox project was to create a user oriented browser. Making it small was a side goal, but it was never a primary goal for the people involved.
Nerd rage is the funniest rage.
Looks like the ubiquity add-on crashes beta 3 (which, I suppose, is to be expected). I had to disable it just to get Firefox to open up properly... Just an FYI in case anyone else is using ubiquity.
I don't remember any Firefox developers flaming people about memory issues. I remember fanboys doing it, but there's a big difference there. The only "official" response I know of from any Firefox developers was Ben's "It's a feature, not a bug" blog post from 2005, which is long-obsolete. Ben doesn't even work for Mozilla anymore (ironically, he's on the Chrome team now).
Give me a browser that can run on a Commodore 64-sized computer times the 8-to-64 bit word conversion..... .....somewhere around 1/2 or 1 megabyte in size. That would be slick.
Yeah I know. "Impossible." (sigh). Right now my FF3 browser is using ~150 megabytes and I don't know why it needs all that room just to display one single page.
"I disapprove of what you say, but I will defend to the death your right to say it." - historian Evelyn Beatrice Hall
So, what is it like being completely retarded?
really brings up the MSN page fast. No really. Really really fast.
A lot of the bullshit surrounding Firefox is the reason I avoid it where I can. Even if that means just using the unofficial or Iceweasel branding. I don't particularly like Mozilla Corporation's obsessiveness with income (see the other Mozcorp story on the front page - "Hey, look, we might be able to get more money from someone else!"), their stance on trademark names/icons (see Debian vs. Mozcorp), or the way their developers tend to come off as, sometimes, elitist (such as the memory leak "feature" discussions).
Now it's been released. It's available for everyone through the links I gave, and through Check for Updates... for Firefox 3.1 beta 2 users.
What a fool believes, he sees, no wise man has the power to reason away.
The only thing funnier than that comment is the fact that the parent post got modded "Informative". :D
Let q be a radix > 1. I am in ur base-q, killing 10 d00ds.
My interest in the new Firefox betas is its official support of cross-site HTTP requests (documented at https://developer.mozilla.org/En/HTTP_access_control). It's following the new W3C spec (http://dev.w3.org/2006/waf/access-control/) for allowing the XmlHttpRequest to communicate with an external domain without the use of the filthy "script get" hacks. I've just spent some time implementing a proof-of-concept for this stuff, and am impressed with how well it works. It even allows POST requests so you're not limited by the usual GET length limits.
It does require server-side modifications, but they're mostly simple.
I see this as the best new feature of Firefox and plan on adding support for this method of XHR into my applications, with failover to the old "script get" stuff. I only hope that other browsers also embrace this new functionality in the near future.
Beg to differ.
http://waynepan.com/2008/09/02/v8-tracemonkey-squirrelfish-ie8-benchmarks/
Pretty sure Lynx will.
WORKSFORME!
With Chrome and its incredibly clean and modern code base and extensions soon to arrive and the Linux version rapidly maturing, the only reason to keep using FF will be misplaced lingering fanboyism
It's easy to have a clean codebase when...
* No fullscreen mode.
* No detection of click-through
* Cut and paste uses icon-shape style instead of dragging an image
* Can't grow selection using cursor
* Not cross platform
* History is just a list of titles (can't even get URL info)
* History looks like a webpage, but you can't do text search or select or right-click on links
* Downloads looks like a webpage, but same problems as history
* Closing a window with multiple tabs nukes them with no warning.
* No 'view page info' showing links, media, etc
* No 'page style' css choices
* Poor handling of many tabs (they shrink forever).
* Can't control what sites are in the screenshots on start page
* Can't search inside and outside a text field at once (either or)
* Can't see pages that are in the cache (work offline mode)
* Print... just silently does nothing if no printer installed
* No rss support at all
* No multiple profiles
* With lots of bookmarks, it doesn't remember where you were in the list so you have to scroll to the bottom again to click more than one
* Can't allow/prevent pages from choosing their own fonts
* No whitelist for cookies
* No clearing of cookies on closing browser
* No separate proxy settings, have to use OS ones
* No settings for enable/disable Java, Javascript.
* Can't restrict Javascript behaviors, such as moving windows
* Can't disable image loading
* Can't modify MIME type mappings
* Can't set max history time in days or entries
* Can't set cache size
* No master password
* No whitelist to avoid site warnings
* No support for security devices
* Can't control update behavior
* Poor accessibility
* No autoscroll (fixed?)
* Can't clear all transfers (have to remove one by one)
* Buggy UI, for example Text Encoding menu doesn't autoscroll up despite having arrows (have to click arrow, can autoscroll down if wiggle mouse)
* No firebug equivalent.
* No mouse gestures.
* Plugins perform badly and/or fail
* Has bad rendering on many non-perfect sites (same with all WebKit browsers)
Oh yeah, and they stole the name 'chrome' from Mozilla, which is pretty scummy. They don't even give props to Mozilla for the name.
Let me know if these are outdated... I don't have my Windows vmware image handy.
Plugin support?
No, codebase isn't clean because lack of features. It's clean because it's designed well. You can add tons of features easily to a clean codebase and keep it clean.
nsPyContext.o: In function `nsPythonContext::Deserialize(nsIObjectInputStream*, nsScriptObjectHolder&)': /home/savages/mozilla/mozilla-1.9.1/extensions/python/dom/src/nsPyContext.cpp:728: undefined reference to `PyMarshal_ReadObjectFromString'
nsPyContext.o: In function `nsPythonContext::Serialize(nsIObjectOutputStream*, void*)': /home/savages/mozilla/mozilla-1.9.1/extensions/python/dom/src/nsPyContext.cpp:682: undefined reference to `PyMarshal_WriteObjectToString' /usr/bin/ld: libpydom.so: hidden symbol `PyMarshal_WriteObjectToString' isn't defined /usr/bin/ld: final link failed: Nonrepresentable section on output
collect2: ld returned 1 exit status
make[1]: *** [libpydom.so] Error 1
make[1]: Leaving directory `/home/savages/mozilla/mozilla-1.9.1/extensions/python/dom/src'
Maybe an early version. The recent Lynxes got quite monstrous.
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
Hmm. I bet there will be a mini-flood of "Oh Shit! Some bastard has worked out how to make the browser crash (yet again) (or leak the sandbox) by doing something truly evil and unexpected with JavaScript again."
It's hard to get that sort of thing really secure. If they do it truly securely, well... my hat's off to them.
If they support JSON shouldn't they release the beta tomorrow on Friday the 13th ?
----------------------------------- My Other Sig Is Hilarious -----------------------------------
Lots of spinning beachballs and hangs when scrolling compared to beta 2.
* No multiple profiles
--user-data-dir='path/to/profile'
You can even simultaneously run two instances using different profiles. My partner and I use this on our shared desktop so we can stay logged in to all those sites we don't care if the other person sees.
http://www.chromeplugins.org/tips-tricks/how-to-create-profiles-in-google-chrome/
Plugin support would be nice. A set of useful working plugins is even better.
Konqueror has plugin support. But where are the good plugins?
What is a nonconsole text editor, and what makes it so?
And I can't use anything else (like, Chrome) other than FF because of the awesome bar (wish they had not come up with that awful name though).
Took me just two days to get used to it, and now I don't have to remember / bookmark more than 90% of the sites I use regularly.
Sorry to yell but Firefox is damned near perfect for me, bar 2 minor issues.
1. a crappy web page which locks up a tab, I want to be able to ignore that tab / process and continue browsing, without locking up the entire browser.
Full honesty: I'm not a coder so I can't appreciate how difficult this might be, but I don't care - I'm an end user, I feel this would be a very very welcome addition to Firefox.
1a. performance increases from 1, ideally also
2.
Applications and plugins 'stealing focus' of firefox keyboard controls.
Control W
Control tab
Control shift tab
Control shift N
etc etc
Some of these awesome and handy keyboard controls simply don't work properly when using FFox on pages such as youtube and other flash based pages, the plugin itself 'takes focus' of the keyboard and for all the ALT D, ALT TAB or any other combination of keystrokes you simply can't get the browser to listen to the commands, you HAVE to click the current tab up the top and re-highlight it / select it (try it now, you'll see a faint outline circle the tab)
I really wish that could be addressed as I'm a 'high end' keyboard user and it really gives me the niggles to have to go up and click on that.
Anyone know a way around it?
I tried to document the issue on youtube but it's hard to put in to words, or visually.
http://www.youtube.com/watch?v=SLk0MBSxb-A
I thought Brett Wilson and Pam Greene were already at Google while they worked on Firefox (on Places, to be specific)? And I have no idea who Jonathan Haas is - his LinkedIn profile (helpfully on that page!) says he was ex-Microsoft / ex-Bungie? (Were you thinking of Josh Aas?)
Goodger, Fisher, Kasting, Pinkerton, and Shin indeed did all work on Mozilla before. I think they were all ex-Netscape? Also, I'm pretty sure Mark Mentovai worked on Mozilla too.
I think Tony Chang also worked on Places for Firefox 2, but Chinese names are common enough that it's hard to tell...
You're assuming Firefox's horrid code base is an inevitable result of its many features. I believe it's a direct consequence of it choosing technologies with bad trade-off like XUL and XBL. Get rid of them both and I suspect the code base would become way cleaner.
A lot of what's in your list are not difficult at all to be added to Chrome without uglifying the code base, so they're irrelevant (e.g. URL in history, cookie whitelist, clearing cookies, max history times, close tab warnings, clearing transfers, RSS support, multiple profiles, printer check, image loading disabling...).
And then there are other features on your list that I'm totally happy for them to get rid of to have a cleaner code base.
I call them the "who-the-fuck-cares" features for non-developers:
> * Cut and paste uses icon-shape style instead of dragging an image
> * Can't control what sites are in the screenshots on start page
> * Can't see pages that are in the cache (work offline mode)
> * Can't set cache size
> * No autoscroll (fixed?)
> * No firebug equivalent.
> * No mouse gestures.
A lot of what's in your list are not difficult at all to be added to Chrome without uglifying the code base, so they're irrelevant (e.g. URL in history, cookie whitelist, clearing cookies, max history times, close tab warnings, clearing transfers, RSS support, multiple profiles, printer check, image loading disabling...).
A lot of those things need their own UI elements. That means translating to other languages. Making sure the UI elements still look ok with 40 pixel-wide asian words vs 40 character wide German words (I kid ;-) vs right-to-left words. And so on.
Firefox has 54 full, localized versions. Chrome has what, a handful of robo-translations?
Maybe if you aren't a programmer you can't appreciate it, but even lots of "easy" features can end up taking a lot of time and effort (which is why Chrome doesn't have them yet), and cause the code to go downhill really fast. Programs always tend to become more and more disorganized and chaotic over time and the more code you have, even if it's "easy" code, the more this happens.
You can always start over from scratch, like Chrome, and do things the 'right' way. But it always ends up as a huge mess unless a huge amount of effort and care is taken. On top of the natural tendency toward chaos, the requirements change over time and invalidate assumptions in the code. Like for instance having each tab be a separate process is great... until something changes like for instance if each page starts needing to keep a much larger amount of state around and duplicating this dozens of times becomes a huge problem (like running a dozen JVMs on your system... how wasteful and slow is that!).
As always, you can try it out by downloading the portable version - Mozilla Firefox, Portable Edition - from PortableApps.com. It won't affect your local 3.0.7 install, so you can try out the features of the new beta without worrying about it affecting your extensions or settings.
http://portableapps.com/news/2009-03-12_-_firefox_portable_3.1_beta_3
Portable versions of Firefox, GIMP, LibreOffice, etc
I have to disagree. Autoscroll is huge and "regular users" use it. The rest can go (I'd really like mouse gestures...but I can admit that it's not for the masses).
hmm, I think I was thinking of Dave Haas. Oh well. Thanks for clarifying.
Comment removed based on user account deletion
However, the original poster assumed that Firefox's horrid codebase is primarily *due to* it's features.
It's not the case here. If you strip off from Firefox all of these features that don't exist in Chrome, the codebase is still crap.
Reason? 3 letters: X-U-L. This technology doomed the Firefox codebase right from the start, even before all these features were added vs. Chrome's clean codebase.