Hacker News new | past | comments | ask | show | jobs | submit login
Firefox 7 Live: Download Here (mozilla.org)
156 points by lforrest on Sept 27, 2011 | hide | past | favorite | 98 comments



Desktop highlights:

* Drastically improved memory handling for certain use cases

* Added a new rendering backend to speed up Canvas operations on Windows systems

* Bookmark and password changes now sync almost instantly when using Firefox Sync

* The 'http:// URL prefix is now hidden by default

* Added support for text-overflow: ellipsis

* Added support for the Web Timing specification

* Enhanced support for MathML

* The WebSocket protocol has been updated from version 7 to version 8

* Added an opt-in system for users to send performance data back to Mozilla to improve future versions of Firefox

* Fixed several stability issues

* Fixed several security issues

http://blog.mozilla.com/blog/2011/09/27/mozilla-firefox-sign...

--

Android highlights:

* Android-style Copy & Paste

* Ability to quit

* Use Firefox in your language

* WebSocket API

http://blog.mozilla.com/blog/2011/09/27/firefox-for-android-...

[Edit] Updated the list of highlights



>* The 'http:// URL prefix is now hidden by default

I hate that:( I hate that they removed the rss button too:(


As said in the other thread, you can change it back by going to about:confing and changing browser.urlbar.trimURLs to false.


It is not in the url bar but you can right click on the toolbar select customize and then drag a rss button back onto the toolbar.



That's weird -- the icon is still there, but not default, and it just takes about two clicks to restore it.

So what's the purpose of the addon?


You mean the icon in customize? You cannot put that one inside the urlbar.


Aha, all is now clear. (I really liked the idea of the RSS button when they added it, but never actually ended up using it...)


websocket protocol bump to v8? isn't v10 the one with the security issues fixed?


Version 8 is the same thing as Draft 10, which is the latest draft. Yes, it's confusing.


Man I never knew about text-overflow: ellipsis. I'm gonna have to start using that.


IMHO quite importantly, this version also includes a list of Microsoft fonts that don't render well with DirectWrite, and will switch back to GDI font rendering for those.

On my machine, this fixes the "fuzzy fonts" issue that existed since Firefox 4 if you enabled hardware acceleration.


Thanks for the information! I first noticed the improvement in Aurora and Nightly builds, and was wondering what they did to fix the problem.

Here is some more information I just found for anyone interested:

Main bug report:

https://bugzilla.mozilla.org/show_bug.cgi?id=661471

The two new preferences:

gfx.font_rendering.cleartype_params.force_gdi_classic_for_families

gfx.font_rendering.cleartype_params.force_gdi_classic_max_size

And a post talking in detail about the issue:

http://blog.mozilla.com/nattokirai/2011/08/11/directwrite-te...


Why would I go there to download a new version of Firefox? I went to 'About Firefox', it downloaded a 10 MB update, restarted, and here's Firefox 7.


Consider the case of not having Firefox already installed.


As a user of Firefox since 0.x versions, I am happy FF still keeps up with nice updates in 2011 and onwards.


I did this to check which version I was running. It started downloading FF7 automatically, I just had to hit 'Apply update'. Awesome.


It's awesome, but it's still not nearly as frictionless as Chrome. My understanding of Chromes upgrader - as a user who only occasionally uses the browser since firefox is still my main - is that the browser auto-upgrades the first time you load it after the new version comes out. Now, I thought this was how the new Firefox worked, but since I did it manually today even after rebooting once for a plugin issue (Netflix doesn't fullscreen properly after you load it a few times - apparently it's an issue with the firefox plug-in container and/or silverlight), now I'm not so sure.

Either way, Chrome installs cleanly without the user even having to know it happened. Firefox, OTOH, causes the UAC in Win7 to halt everything for the upgrade. It's not as transparent as maybe it could be.


Firefox does update automatically. By default it checks for updates every 8 hours, downloads the update in the background, and installs it the next time you start Firefox. If you haven't restarted Firefox for 24 hours after an update is downloaded, it'll prompt you to restart. If any of your add-ons will be disabled by the update, Firefox will list those add-ons and ask whether to upgrade or not.

We're working on refining all of this; here are some of the things that may change in the default update behavior in future versions: https://wiki.mozilla.org/Silent_Update


Thanks for the clarification. I did think the Firefox update was automatic, but I think I confused myself by doing it manually today.

My main point was really meant to be the UAC, since it was a bit jarring to see today (First upgrade on Win7; just bought this computer a month ago). I'm glad to see that this is being looked into.

On the bright side, I had absolutely no issues with add-on compatibility this time around. In the old days, I would wait a while - sometimes literally months - before upgrading, because of issues with add-ons I considered "important". I actually ended up backing up my profile before any upgrades, and often had to roll back the version. I didn't even think of that today, I just clicked apply and everything worked. It's a serious improvement.


That's because the release is closer to 4.3 than 7. Faster, smaller changes rather than slower, bigger ones.


By default, Chrome/Win installs in your profile dir instead of the global Program Files so it doesn't need admin to update.


Chrome also uses bdiffs, so updates are significantly smaller.


Firefox uses bsdiff to generate binary diffs for updates: https://wiki.mozilla.org/Software_Update:MAR

It's not quite as efficient as the "Courgette" algorithm used by Chrome, but it's much smaller than downloading a full installer. Hopefully we can use Courgette if the patent issues* are resolved.

* http://ipwatchdog.com/2009/10/31/google-sued-for-patent-infr...


That's not working for me. Help->About offers an update, downloads it, installs it, then informs me that the update it's just installed (6.02) is old, and I should go to the website to download v7.


Why am I told 'firefox is up to date' (6.0.2) when i do this? Is this done geogprahically? (UK here)


The designer sitting next to me is most excited about this: >* Added support for text-overflow: ellipsis


This one drove me nuts. Many developers were achieving the same result in firefox 3.6 and earlier using an XUL hack - but in FF4, the ability to do that XUL hack was removed for (valid) security reasons - but without addressing the text-overflow: ellipsis issue for which they'd had an open bug tracker entry for about 5 years. There were many complaints about this months before FF4 was released but they went ahead anyway.


i'm pretty excited about that one too. i was just cursing firefox last week for that.


* Drastically improved memory handling for certain use cases

^^ apparently not the use cases I use. FF7's been running for about 12 hours on my win7 box and is using over a gig of ram just like normal...

It started off good but then just kept growing over the hours. Nothing really improved here. :(


> FF7's been running for about 12 hours on my win7 box and is using over a gig of ram just like normal...

Are you looking at resident set size (aka real memory), versus virtual set size (aka virtual memory)?

The memory VMs in both Windows and Linux kernels will happily let an app malloc all they want until they hit the userspace address limit, with no affect on performance. It's only when those apps use the memory that there's a perf hit.


Be sure to enable this:

* Added an opt-in system for users to send performance data back to Mozilla to improve future versions of Firefox

Some of these issues are very hard to reproduce for the developers. If users like you turn it on, there's hope of actually getting it fixed.


What add-ons are you using? 95% of the time, problems with Firefox (speed, crashing, oddities) can be blamed on a slow add-on.


Open about:memory and see where the memory is going. And if you think the use is excessive, please file a bug including the about:memory data and any details you know that would aid in reproducing the problem.


Is this memory that is needed for some other application you are running? Is your system running low on virtual memory? If not, what is the problem with FF using it? From what I've seen (at least with FF6), when system memory gets low FF will start to release memory back.

I'd rather have FF use the memory for cache to speed up performance than to have it sit unused.


Try going to "about:memory" -> "Minimize Memory Usage"


I thought the point of Mozilla's new release scheme was that we don't have to download anything anymore. The updates just come to us, as if by magic. https://support.mozilla.com/en-US/kb/Updating%20Firefox


They do. These posts are for those who've never installed it before. All 5 of them :-)

Seriously, the release notes and changelog are what we're after.


But extensions are still version dependent.

I was just about to update my installation, but it told me that my Roboform extension isn't compatible with v7.

If they allow the application to update when it wants, they're going to quickly render me unable to use Firefox, since I won't be able to log into secure web sites.


I was just about to update my installation, but it told me that my Roboform extension isn't compatible with v7.

How do you get to that information prior to the upgrade?

I've only realized extensions were incompatible after updating, because Firefox disables them.


Did you receive the update automatically, or did you download and install it manually? If you are using automatic updates, Firefox should ask before updating if any of your add-ons will be disabled.


My experience is that if I click "Check for updates" and an update happens to be available, then there's no way to prevent that update from being installed, even if it causes add-ons to be disabled.


By the way, Roboform has been updated to support Firefox 7: http://www.roboform.com/support/news


Go to Help > About Firefox and click Update. It should update automatically after some time anyway.


Yep. You only have to download the update if you don't want to wait the up to 8 hours before your browser pings the update server and discovers there's an update.


Great. So now I have to be two versions behind if I need to use GWT hosted mode.

(Grrr... yet more evidence of G's gradual sunsetting of GWT.)


I came here to see if anyone had anything to say about this. Hadn't heard rumblings about sunsetting GWT, but the conspiracy portion of my brain wondered if Google not trying too hard to keep the plugin up to date was a subtle push for GWT dev's to move to Chrome. :)

Last I checked the plugin for FF6 can be found here: http://www.google.com/url?sa=D&q=https://dl.google.com/d...

More discussion here: http://groups.google.com/group/google-web-toolkit/browse_thr...


Well, the main evidence for me can be found in the leaked "Dart/Dash" memo, where GWT is referenced a few times:

"The JSPrime effort was begun to unify and be a (single!) successor to GWT and Closure/JSCompiler"

"What is the future of the JSCompiler and GWT? JSCompiler and GWT were already on a merger path. This effort gives us a direction for that unification around the Dash language. We will actively support teams for a long time on the current generation of JSCompiler and GWT and provide fantastic co-existence and migration tools to Dash."

See: http://markmail.org/message/uro3jtoitlmq6x7t

(My emphasis added. I read that as "maintenance mode".)

(Edit: thanks for the heads-up on the FF6 version of the plugin - but I guess I have now missed the two-week window where both FF6 and the compatible version of the plugin where available to download!)


Whoa! I had no idea, thanks for sharing. This needs its own submission, so that more people can see it.


>The 'http:// URL prefix is now hidden by default

This can be changed back with browser.urlbar.trimURLs in about:config. They should've provided an option for it.


No, they really shouldn't. If you care, you already know about about:prefs.


This is one example of a devastatingly common category of assumptions. Never assume that the set of people who care about something is the same as the set of people who know how to fix it. Similarly, distrust statements of the form, "Anyone who is/does/has X already knows about Y."

I myself have been bitten by this assumption when I discovered a new subject for study (e.g. math, entity formation, physics, a new programming language, etc.), but found most of the web pages related to the subject assumed that all those interested in the subject were already fully knowledgeable about it.


Note that if you copy/paste the contents of the URL box, the pasted text includes http:// or https:// or whatever was relevant.


I've had problems with this using the X buffer copy / paste on Ubuntu. It also makes copying just parts on the URL less intuitive.

Personally I don't see the benefit of removing the "http://. If this trend continues we may end up with something similar to AOL Keywords.


I don't want to install FF7 to test this, but how can a user switch from a http to https version of a site? Can a user still type the URL prefix to override what is hidden?


Yeah, I just tried it, and apparently typing a protocol in front replaces the hidden 'http://.

As long as you can assume that the protocol is always http:// when hidden, I think I'm ok with this.


After Firefox 4 came out I had to install so many extensions to make it feel like 3.6 that it became unstable and incredibly slow. Because Firefox apparently tries to be more like Chrome I decided to switch to the original and I haven't looked back yet. I only fire up Firefox when Chromes built-in developer tools lack functionality compared to Firebug, which is rarely the case. The way Firefox decided to take makes me really sad, as I was a big Firefox supporter, until they screwed me over with the new UI. I hope this is not too off topic, but I never got the chance to comment on the changes.


Huh? The only way it's like Chrome is how you can move the tabs above the other controls on top. 3.6 feels very slow now, I can't use it and have never felt that I wanted to.


I actually still run 3.6 since I like the UI on it better. The speed of the browser means very little to me, I'd rather have the familiar UI. If you liked FF 3.6, why not stick with it?


And FF 7 (which I've been using for about 5 months now) is not even noticeably faster than 3.6 on my usage pattern. In fact it seems to spend more time with an unresponsive UI than 3.6.


after I've worked with a fast browser like Chrome I just can't go back


What fundamental new is for new UI comparing with 3.6? I use Firefox for some way, only for firebug mostly. I think its just evolution of person, now I use what is most serves my needs.


Tab Groups are cool, but I don't think anyone uses them. If you're on Windows 7 it hides the menu bar for more vertical space. It has support for lightweight extensions (called "JetPacks" internally) which don't require a restart to install or uninstall. There's no statusbar, like Chrome, URLs pop up when you hover over a hyperlink.

The main advantage over Chrome is that it doesn't freeze up when it's loading a new tab.


I use Tab Groups mainly for task related stuff, as well as using a tab group to allow for a read later style of interface. But no, it's not a major thing for me, I've used Chrome for the past few months and not missed Tab groups.


Tab groups would've been much better if they showed all the tabs across all windows (like Expose on Mac does for windows). Unfortunately it's just a gimmick in its current implementation.


Firefox also hides the menubar on linux, although it can also be disabled. It is also posiblse to enable an 'extension bar', which takes the place of the status bar for extension icons.


what bothers me most is that they made the whole top part (including title, menu bar etc if you choose to enable it) transparent. I personally just find that horrible. Screenshot: http://0.tqn.com/d/browsers/1/0/Z/c/-/-/firefox-5-show-menu-...

Here is a screenshot comparing FF 3.5 to 4: http://cache.gawker.com/assets/images/4/2009/12/500x_firefox...

(3.5 is right)


Arguably, having the menu visible is not the default setup, is it? The default fold-out "Firefox" menu simply doesn't have that issue and is not transparent.

I agree it's a problem in Thunderbird, which has a menu by default.

I think that if you try to tweak Firefox 4+ to look like 3.6 you're going to run into suboptimalities. It evolved. I think that if you get used to the changes, they're actually changes for the better (new menu giving more vertical room, tabs-on-top, etc).


If you tap the Alt key, the regular old menu comes up. It's transparent.


Enable the menu bar and disable tags on top. Then install a plain persona like this one: https://addons.mozilla.org/en-US/firefox/addon/white-glossy-...

It's not exactly like 3.5, but it's not transparent anymore.


I did that and installed a couple of extensions that add old functionality and Firefox now crashes pretty much randomly


Great! Love the new canvas renderer. Anyone know when they're planning to add generational GC or other improvements to the GC? It's making some HTML5 games choppy as far as I can tell.


Smooth animation would require shorter GC pauses, i.e. incremental GC.

https://wiki.mozilla.org/Platform/Features/Incremental_GC

The implementation seems to be in active progress, which means it's going to take at least a few more versions to make it to stable.


Incremental GC is mostly done, so it might make it to trunk soon (so maybe in FF10 - 4 months from now). Generational GC and other improvements are also in the works but will take longer.


If you're looking for an updated version of Firequery (http://firequery.binaryage.com/) I patched and submitted a pull request: https://github.com/davemo/firequery/tree/firefox7

Until it's officially accepted you can fork, build it with rake and have yourself a working copy for Firefox 7 :)


Still no 64-bit version? I can't find a clear page on the status of Mozilla's x64 stuff.


Firefox reflects the general state of x64 adoption: http://en.wikipedia.org/wiki/Firefox#64-bit_builds


why is the channels[1] page never up-to-date? the 'download beta' link still sends 7.0 - if they don't want to confuse people, they should try to be more consisten across their site (which is in general not very well thought-out in my opinion)

[1]: http://www.mozilla.org/en-US/firefox/channel/


The Aurora and Beta channel updates are released a few days after the stable channel update. See the text on https://wiki.mozilla.org/RapidRelease/Calendar - or for even more details see http://blog.mozilla.com/channels/2011/05/11/merge-dates-vs-r...


"Code is not always released to users on the same day as the branch migration. The release to users may be a few days later, to allow for manual testing and sign-off."

I think this is confusing. Either you synch the switch (i.e. when you release the "release" version to users you release the "beta" too), or it doesn't make a whole lot of sense.


I upgraded, but Fast Dial is broken on 7.0, so I downgraded to 6.0.2. Hopefully it'll get fixed soon.


I like this BUT addons need to catch up and for many it's impossible due to such frequent changes.


99% of the add-ons that are hosted on http://addons.mozilla.org are compatible and had been automatically updated to be compatible. Can you list which ones you use that are not?


Earlier, I read a pretty interesting mozilla post about the problem with non-AMO addons.

> it was quite eye-opening when I learned that only 25% of the 600 million add-ons in use every day in Firefox 4 and later are active on AMO.

http://blog.fligtar.com/2011/09/26/add-on-compatibility-prog...


My most important one, roboform is coming as 'will be disabled.'


This may be because roboform proper isn't hosted on addons.mozilla.org, so they can't vouch for it or bump the version for you. The Roboform Lite that is on addons.mozilla.org is properly marked.

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


I have been having terrible freeze issues with 6.x so hopefully this update fixes it.


It is still happening with 7. I am using Vista with the following: http://www.reddit.com/r/firefox/comments/hbqag/my_firefox_4_... (I disabled Firebug)


what OS? what extensions do you have installed? I'm very interested in your freezes. can you say more?


I've had these as well. Firefox 6 on win7-32, with latest Adblock-Plus, Firebug(usually disabled), Flashblock, Greasemonkey, Noscript(in allow-everything mode), Foxit plugin, latest Flash, no antivirus, no other media plugins.


Just read this and thought of your comment...

http://blog.bonardo.net/2011/09/30/is-your-firefor-freezing-...


That sounds like my bug. Hopefully this fixes it! Thanks!


Nice to see Web Timing support make it into this catching up with Chrome and IE.


Anyone know a reason the auto updater goes from 5.0 -> 6.0.2 -> 7.0?


It looks like some of the old configurations on the update server have not been updated to reflect the latest release; some details at http://bugzil.la/689004


Looking at the features, it looks like they're fashioning the UI after Chrome's. The reload/stop button and tabs on top are new, and what is an "Awesome Bar" but an omnibar?


An omnibar that predates the omnibar.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: