Hacker News new | past | comments | ask | show | jobs | submit login
Firefox 50.0 (mozilla.org)
384 points by conductor on Nov 15, 2016 | hide | past | favorite | 193 comments



"Set a preference to have Ctrl+Tab cycle through tabs in recently used order "

Finally something that is an actually improved UI feature!

Not some "removed status bar and instead hover its info over text you want to read sometimes" or "moved refresh button to different place than before just to annoy you" or similar thing :)


I have this feature in sublime text & Atom and absolutely hate it.

I feel like ctrl + tab has to move to next tab & ctrl + shift + tab should move to previous tab (cycle if it reaches the end).

I feel like that's more predictable behaviour.


It's all personal preference, but since Opera 12 I've been addicted to ctrl-tab for most recently used switching (Opera 12 had a nice tab switcher that made the interaction very understandable/manageable), and then alt-tab for switching in "physical" order. (I run a large number of tabs, and quickly switching between two (or a small number) that I've used recently seems to be a common interaction for me - but then being able to look at favicons and traverse quickly to a closely located tab too (is less crucial, but important enough that I don't use FF (because I can't remap that alt-tab interaction simply).)

And I love it :-p It is literally the reason (in combination with "open new tab next to current") I use the browser I do (Vivaldi; would consider Opera, but it doesn't/didn't do "open new tab next to current").

To each their own :-)


Check out TabMix Plus extension for firefox - you can set up literally everything about tabs behavior.

(It is literally the reason I will not be leaving Firefox after this Ctrl-Tab moves to most recently used tab abomination :))


I was trying out the new Opera with its built-in VPN and ad-blocking which I thought were nice default features but the tab switching was driving me bonkers! I could not understand it, it was flicking through recently used first you say? I would never have guessed. To me it just felt random. I want ctrl-tab to cycle sanely left to right and shift-ctrl-tab to cycle sanely right to left. The browser equivalent of alt-tab for app switching. I do not want any application grabbing alt-tab from the system.

I'm joking of course that I think to desire otherwise is insane. Except I'm not. Or am I? :)


Fair point about alt-tab; I didn't quite make it clear, but I'm using DWM with the windows-key as meta, so meta-tab switches between two tags (do that infinitely often (ie, not cycling, just swapping)) and meta-[0-9] for general tag jumping.

Hence, alt was just hanging out being useless :-p (I'm like 95% terminal or browser, so it's a great use of alt to me :-) )

Amusingly, last time I used OSX or Windows properly, I got super grumpy about the fact "(alt|mac)-tab" seemed amazingly inconsistent - I think cause I'd try and switch consistently between two windows, and got surprising-seeming results... I somewhat think I had gremlins :-p

(As I wrote this I actually did a few ctrl-tabs and a few alt-tabs... I hadn't quite realise I used both as often/evenly...)


a feature that I've loved about Vivaldi (and vimium on chrome) is the spotlight like search for tabs. When you run a lot (~30-40) of tabs, that's an killer feature right away!


Aside from the fact that the regular URL bar in Firefox does this, Tab Center (https://github.com/bwinton/TabCenter) puts tabs on the side with a filter-search at the top.


The search and hide/unhide function seemed interesting to me, but in my opinion the Tree Style Tab add-on is superior. Perhaps that is because TabCenter is aimed at a less technical audience [1]. Hierarchical tab structures are more valuable in my opinion, because it implicitly groups related tabs together. The structure in which tabs are opened reflects my workflow while opening them and allows me to e.g. close all tabs that were related to a certain problem. It does however lack a search function and some sort of versioning, as I find my tab-bar clutter over time. Perhaps a aging approach could be helpful for me, but I have yet to come across something better than Tree Style Tab.

[1] https://github.com/bwinton/TabCenter/issues/498#issuecomment...


Oh, yeah, I'm not pitching this as an alternative to Tree Style. I like Tree Style too.


This is brilliant! I wasn't aware of tab center. The best part is that it works really well with tabgroups too.

However, I am unable to use it through the keyboard. The undocumented C-S-l seems to trigger a search over there. But there is no way of navigating to a matching tab without the mouse. Any tips?


Ctrl-tab seems to work for me.


With that filter I wonder if we could combine tabs and bookmarks?


Pentadactyl for Firefox also allows that (press 'b') and it's absolutely amazing. Way quicker than searching for the tab visually and manoeuvring the mouse to it on a notebook.


Actually, Firefox has it natively since 48. Just enter the title in the URL bar, and it'll suggest going to that tab.


Firefox has had it native since firefox 4. See this article: http://kb.mozillazine.org/Location_Bar_search#Location_Bar_s...

You can enter "% string" into the location bar, and it will suggest switches to open tabs which contain that string either in their title or URL.

I've been using this little-known location bar feature for a long long time.


Hm, then 48 just changed the way it works, because I've never seen it before that one.


The problem with this feature is that it depends on hidden UI state. With the "last used" behavior there's no way at a glance to know which tab will be activated when pressing ctrl-tab, so you either have to maintain the state in your head, or guess-and-check how many times to hit the key combo.

If you only have a few tabs open, it can be useful in conjunction with Ctrl-[1-9], but I am also still strongly in the next/prev tab camp.


Pressing Ctrl-Tab once without letting go of Ctrl shows you a graphical overlay of most recently used tabs. If you don't like what you see, just add one Shift-Tab to your already pressed Ctrl.


Can be fixed in Sublime Text as follows:

{ "keys": ["ctrl+tab"], "command": "next_view" }, { "keys": ["ctrl+shift+tab"], "command": "prev_view" }


Ctrl+shift+P then type "key bindings user"


Cmd+Shift+P for MacOS.

Also Preferences -> Key Bindings works too but I'd encourage everyone to get used to the command palette for all sorts of commands like Set Syntax and Package Control


I agree. It's the worst!

I don't use ST, but at least in Atom, I know you can easily override it in your keymap:

  'body':
    'ctrl-tab': 'pane:show-next-item'
    'ctrl-tab ^ctrl': 'unset!'
    'ctrl-shift-tab': 'pane:show-previous-item'
    'ctrl-shift-tab ^ctrl': 'unset!'


Is there a way to disable this ctrl+tab behavior in Sublime Test? I can't stand it either.


I guess people are down-voting you because it's a simple google search away. However, to answer your question:

On Atom, I've set ctrl + tab to remap to pane:show-next-item, and ctrl + shift + tab to pane:show-previous-item.

(view your keymap, and see what ctrl + page-up/down is bound to, and then bind it to ctrl + tab /+ shift)

Edit: here are my settings:

   'body':
     'ctrl-tab': 'pane:show-next-item'
     'ctrl-tab ^ctrl': 'unset!'
     'ctrl-shift-tab': 'pane:show-previous-item'
     'ctrl-shift-tab ^ctrl': 'unset!'


Seems to be a subjective preference indeed then :)

Usually when programming, the most relevant are the last source files I viewed, e.g. a source file and its header file. I love ctrl+tab switching between most recent files this way.

Same in a webbrowser with related articles. Don't see a use of a next/prev in the list tab shortcut myself, it's not like tabs are ever in some particular order for me :)


Yep. I want to be able to predict where ( Ctrl + Tab ) x 3 where take me.


When browsing, I actually use both behaviors in different cases. C-<tab> and C-S-<tab> to cycle linearly, and then I use vimium shortcuts for more advanced navigation. So I survey tabs by tabbing through them, but when I'm going back and forth between two or three (citing things, for example) I use LRU-type features. Forcing people to use one or the other, regardless of which, is going to make them angry.


Yes. In iTerm2 as well, (shift+)ctrl+tab jumps across recently used tabs. Holy hell how it's confusing and I am used to use (shift+)ctrl+tab in Chrome so switching between Chrome and iTerm2 while web developing sometimes is frustrating.

If I want to switch between 2 tabs quickly - I set them side by side and it solves the problem for me.


I don't know about Firefox or Atom, but in Chromium you can achieve that via Ctrl-PgUp and Ctrl-PgDn.


I know I can ctrl-pgup/down in Firefox (it's the best alternative to alt-tab afaict) but on one of my laptops it's incredibly impractical (pgup/down is way up the top right), and on the other pgup/down is in a practical, but very different, location to tab so it's "disjointed" in my mind (ie "do I want to go mru? ctrl-tab; do I want to go ltr/rtl? alt-tab" they're still in the same area, and hence it seem logical/simple to switch between those options often, whilst pgup/down seems very different?)

(And then meta-tab is also hanging out there, being simple and ready for an alternative context switch (mostly browser <-> terminal).)


Which is a pretty dumb shortcut to use one handed and compared to alt+tab.


I have never seen a keyboard which doesn't have a Ctrl on the right side.


Sure, but there's way more distance between Ctrl and PgUp then Ctrl+Tab (I'm taking a classic desktop keyboard in the account now). There are quite some more compact models where PgUp/Down are between the NumLock and - on the right side with the Numpad.

Not to mention you'd need to set the hand from one end of the keyboard to the middle, just to press something you'd otherwise move 1 key.

As I said, it's dumb.


Tons of laptop keyboards. Of course lots of those have no PgUp either.


It exists for ages already, though. You use the same shortcuts to cycle tabs in dialog boxes. And that goes back to at least Windows 95, if not further.


On Mac at least you can use Command Shift [ and ]. Which is the same as in the browsers.


Probably pretty nice for left handed mouse / trackpad users though.


Every non-macOS browser I've used supports those shortcuts. They're old enough to be universal among tabbed browsers.


I use Ctrl+Tab to switch between two tabs I'm working on.

If I want to go to specific file, I'd just Ctrl+P and type in the filename.

Using Ctrl+PageUp/Down to scroll through all open tabs is predictable but kinda useless since I have 20+ tabs opened after a while.


I hope I can disable this feature in Firefox about:config.


Ctrl + PgUp , Ctrl + PgDn


This has been possible for a very long time, as a side-effect of enabling browser.ctrlTab.previews in about:config

Glad to see they are paying more attention to this though.

Thousands of people have requested that feature for Chrome. To this day there hasn't been (or I haven't found) an official answer and the only way to do it is to use an extension, using an alternative keybind as extensions aren't allowed to interact with ctrl-tab.


Chrome doesn't add settings for things when extensions can do the job because doing so adds to our support burden and slows further development - it's better to appeal to fewer people and have a simpler and easier/faster-to-maintain product than to appeal to more with higher maintenance complexity.


It seems as if the feature is just directly using browser.ctrlTab.previews, toggling it also toggles the new option in the Options menu.


I'm glad you like it. This is one of those personal UX preferences that has you either one or the other camp. Am I right to assume you liked this from your past Opera use?


I was excited to sudden see this surprise in the patch notes, now I am the opposite here thinking, "How did they mess up such a simple feature?" The preview bar that pop ups is slow to work and actively delays changing tabs.(If it even attempts to switches tabs.) Additionally, I do not need the preview portion.

FLST: https://addons.mozilla.org/en-US/firefox/addon/fww-flst/

It just works. Tab switching is instant, no unnecessary preview, and it has never messed up remembering tab order.


I'm with you and cannot stand the way Opera does it but it also has a quick tab switching mode. I haven't encountered a single switch-with-preview implementation that didn't leave me with forced slowdown and delay of the action I was about to do. I also configure Firefox to not close the window when it's the last tab, which is hard-to-impossible to emulate in Chrome but is the default in Opera and Vivaldi.


Actually I like it from IDE's and text editors :)

As well as the alt+tab desktop behaviour


Years ago, this used to be the default in Firefox. Since Ctrl+Tab has not been the default, I've been using Ctrl+Tab extension. Other browser like Chrome do not provide APIs to allow extensions access to override tab order. Accessing tabs by order of last used, is a requirement for my daily web browsing, which is why I've stuck with Firefox for so long. I'm glad this is now part of the preferences, and does not require an extension.


Added things with hip names like Pocket or Share that you will never use and inserted it right next to your address bar.


Personally, I wouldn't be comfortable with that activated. It happens in Atom and it's really quite annoying.


on that note: I really wish they would move the refresh button back to the modular interface. I hate where the refresh button is (so do lots of people) and just wish it wasn't the ONE thing on the toolbar you couldn't config. (yes I've replied to and opened a ticket)


Any firefox devs here? You guys broke your own VTT polyfill with this update: https://github.com/mozilla/vtt.js/issues/354

Only reason I'm mentioning it here is that it looks like the repo is dead.


I'm making some inquiries. It looks like the old code just ended up setting "position" to 0 when an object was assigned, while the new code (correctly per spec) throws. The real bug is assigning an object to a non-object-valued property...



Thank you!


Excellent, thanks a lot! If I find a fix I'll PR it


JS source maps support is a big one for devs.

https://developer.mozilla.org/en-US/Firefox/Releases/50#Chan...


Also to note: Firefox can now upload directories.

Granted, it's done by mimicking Webkit APIs, but that has the advantage that existing sites work without changes.


> Granted, it's done by mimicking Webkit APIs,

That's ... not a bad thing. Chrome creating nonstandard APIs (IIRC this was for Drive?) on its own is a bad thing. Coming together and speccing (https://wicg.github.io/directory-upload/proposal.html) the API is a good thing. They seem to have specced more or less what Webkit had already implemented (plus some promise based stuff), but usually when a nonstandard API has been out there long enough it's best to build your standardized version on top of it instead of having two APIs for it. This is a common practice. This isn't "mimicking".


> They seem to have specced more or less what Webkit had already implemented

The problem is that the WHATWG always just standardizes "whatever Chrome does".

This leads to the HTML living standard actually being just like Office Open XML, with one entity controlling all of it.


> WHATWG always just standardizes "whatever Chrome does".

It really ... doesn't. When the situation has arisen that Chrome has implemented an API for long enough for it to become entrenched in the web, it then discusses standardizing it (see also: https://compat.spec.whatwg.org/). But it's not blindly following Chrome, it's only in these special cases.

I've recently been involved in a couple of issues where browsers disagree with each other and/or the spec. Each time there's a discussion, and the best version of the behavior is chosen and worked into the spec. This best version is not always the Chrome version.


I would postulate that the cause of this is that, as a provider of services, Google kind of knows what's difficult to do in a browser and what apis are needed to make it more native-like.

So it's not a terribly bad thing, unless other people's standards are just being ignored


> So it's not a terribly bad thing, unless other people's standards are just being ignored.

It's actually an unfair advantage.

"Oh, let's standardize what Chrome already has implemented and Google is using". This immediately sets Firefox behind the new "standard", and website that don't work on FF technically "follow current standards".


Actually, no. Google is bound to have its own favorite frameworks (perhaps even in house), its own coding style, and in general its own way of doing things. Most large companies do. Google would end up creating APIs that are most suitable for their use cases and their frameworks. These need not be the best APIs for all.


I was confused by what this meant - so perhaps some others are too. This is what a few minutes of research got me:

    Uploading individual files in a directory = was possible
    Uploading an entire directory (eg: a folder to a cloud storage service) = was not possible
Personally my workflow means I've uploaded the contents of a directory to a directory I make myself, so I've never encountered the problem.

Am I missing something or was that right?

Example of error: http://superuser.com/questions/909112/does-firefox-support-f...


You can't upload subdirectories, so it quickly becomes painful to upload an entire directory of files if it has multiple folders within.


Outside of my personal use cases but I can definitely see how many, if not even most, people would need that at some point. :)

Thanks for the response!


Another reason is on macOS where many file formats are actually directory based. (Called 'packages', they are actually structured directories with a metadata bit set.)

For example the RTFD format. This is a rich text document with embedded images. It is actually implemented as a folder containing a RTF file and the images. In the Mac GUI however it appears to be a regular file.

Attempting to upload it on the web would fail without the ability to upload directories. This may well explain why WebKit (being from Apple) was the first to implement this. I bet they needed it for their iCloud web app versions of Pages, Keynote etc.


Obligatory "does it break Tree Style Tabs?" comment.


You might want to check out Test Pilot and the Tab Center addon : https://testpilot.firefox.com/experiments/tab-center

Unfortunately it is very slow compared to Tree Style Tabs if you have a considerable amount of tabs open (even using extensions like auto-unload, it is very very slow). Hope it gets fixed because I would _love_ a built-in vertical tab feature.


I have hundreds of tabs open, and use Tab Center, and I haven't experienced it being slow.

However, past a number of tabs (enough to fill the "tab bar" vertically), it doesn't show thumbnails, maybe that's the difference?


I used to have those kinds of hundreds of tabs in multiple windows until I found Onetab: https://addons.mozilla.org/en-US/firefox/addon/onetab/

Still using Tree style tabs for their sub tabs feature which helps keeping things organised for me.


On my machine (Intel Celeron, 8g RAM) it's noticeably slower than TST even with only 50 or so tabs


I'm using nightly version 53 and Tree Style Tabs works fine


Using nightly here and it's kind of buggy. New tabs don't appear unless I click on them, and the very first tab is overflowing on my bookmark bar


I'm on tree styles tabs 0.18.2016090802 and firefox 50.0 and haven't had any problems.


If they could just add a decent profile management option, I would switch back from Chrome. Keeping my work and personal accounts separated is the only thing keeping me using that battery killer


If you're willing to try Firefox Nightly, check out the containers experiment: https://wiki.mozilla.org/Security/Contextual_Identity_Projec...

While everything still happens within a single profile, sites in different containers get different storage (cookies, localStorage, IndexedDB, etc.) and cannot see each other.


Cool.

I've been using a combination of NoScript,Self-destruction coockies,ublock and a personal vimperator script to wipe out all that nasty stuff when I close the window.

https://github.com/liloman/dotfiles/blob/master/vimperator/....

https://github.com/liloman/dotfiles/blob/master/Scripts/Scri...

By the way firefox is pretty nasty by default and you must do a lot of hard work to evade tracking, in special by google:

https://github.com/liloman/dotfiles/blob/master/vimperator/....

Of course you must use custom fonts if you don't want to be tracked by google for every single webpage you enter.

I think I will play with containers for firefox when I try firefox 50. Something like a new container for every tab, I reckon It must be pretty easy with vimperator(or whatever plugin you like) to make .

Put your vimperator/penta/vimium/X to work for you. :)


That looks way more powerful than what Chrome is doing. I'll give it a try. Thanks for the suggestion


I think this is great, but is there going to be a simple "sign in" feature similar to Chrome? People just love having their bookmarks / toolbars / etc follow them wherever they go.


You can sign in to a Firefox Account to sync your bookmarks, passwords, history, etc.:

https://www.mozilla.org/firefox/accounts/


What's wrong with `firefox -P`?


Compared to Chrome profiles, it's much less usable in regard to launcher icons and window management. There are workarounds (system-specific), but none I saw worked as nice as Chrome's implementation.

Container tabs are (IMHO) better than Chrome profiles, because they don't require one to have a separate window just to open a single site.


You mean compared to just right clicking the top right corner in Chrome? A lot.


You can also run several profiles in parallel with the no-remote flag.


yes, this! I've tried so hard to like FF profiles, I really wish they'd make this better. I still refuse to go to Chrome, but I'm getting lost on these small features.


I use 'firefox.exe -p "test" -no-remote' to open a second window with a different profile. or do you require to have it in the same window ?


>Added download protection for a large number of executable file types on Windows, Mac and Linux

It seems archives (such as .bz2, .gz) are treated as executable files. What is the reason for that? https://hg.mozilla.org/mozilla-central/file/054d4856cea6/too...


My best guess would be that there is still some risk with these archive types since the browser-client has built in automatic decompression for them.


> View a page in Reader Mode by using Ctrl+Alt+R (command+alt+r on Mac)

Took them bloody long enough. Now if only i was not stuck on ESR because GTK3...


If you have a hard requirement to use a thoroughly obsolete GUI toolkit, you can for now fairly easily compile your own copy of Firefox for GTK2; it takes around 30-40 minutes on a laptop with a Core i5. (On Gentoo, there is a USE flag for it.)


If it is that bloody simple, why can't Mozilla be bothered to provide an official download?!


Possibly for the same reason they don't ship Tk compiled versions. Why should they target every GUI toolkit and complicate their build process which they then have to deal with supporting?

If it's that easy, you can probably find some third party that compiles and offers it. I'm not sure why you think Firefox needs to do it themselves.


For the same reason they haven't made 64-bit the default on Windows and provide two versions. There are many users for whom only one variant works reliably. If upstream, like in the case of cairo-gtk3, decides GTK2 is not supported anymore, it's a matter of chance whether your cairo-gtk2 build of Firefox works at all. For instance, for me 49.0.2 gtk2 local build on Arch crashes anytime I try to use the file dialog, while of course the file dialog of both GTK2 and GTK3 doesn't crash in other gtk apps on the same machine.

Somebody requested this and the reality has been ignored so far: https://bugzilla.mozilla.org/show_bug.cgi?id=1268234


On windows, Firefox is the main supplier of binaries. I would hazard that on most Linux/BSD systems, the distribution is the main supplier of binaries. The distributions can easily build it as GTK2 if they want, or provide a GTK2 variant (and as pointed out above, some distributions seem to, or at least make it easy to build it yourself).

If you are on Linux and your distribution provides Firefox, this is a complaint to be leveled at your distribution, not Mozilla (who apparently already makes it easy to build the variant). I'm not sure how we got to a position where people feel justified in criticizing a company providing an open source product that's updated often and provides umpteen different binaries for different platforms and different build for those platforms for not building one more special configuration for what it likely a very small group of people, who can easily do so for themselves.


In the past the GTK3 code path was tested only by Red Hat as the primary developer of all things GTK and GNOME. At that time, there was only a GTK2 build from Mozilla, and now instead of doing the same as Windows 32-bit/64-bit we're presented with just GTK3 support with the promise to obsolete the GTK2 code, while seemingly not considering the regressions of GTK3.

Why do we use Mozilla binaries of Firefox on linux distro where there's Firefox builds in the package repository? Many reasons:

1. fast access to security fixes

2. access to EMEfree builds

3. access to different channels

While it is easy to build with cairo-gtk2 as the backend, that code path, as I wrote in a sibling comment, reliably crashes for me anytime I try to do File-Open. I'll try to find out if that AUR recipe or Gentoo ebuild do something different that makes it stable, but the fact remains that GTK2 is about to be unsupported by Mozilla while GTK3 hasn't gotten stable yet, which will make life for anyone that tries to build a GTK2 variant very hard. Therefore, I wouldn't really say it's an easy choice.


> While it is easy to build with cairo-gtk2 as the backend, that code path, as I wrote in a sibling comment, reliably crashes for me anytime I try to do File-Open.

That points towards GTK2 support possibly not being functional, which could be a good reason why they don't provide a build for it, even if they wanted to.

It sounds like the real problem here is that Mozilla is changing stuff that some people don't want changed. I don't think the solution is to ask Mozilla to produce a GTK2 build, but to ask them to fix anything they've broken. If they've decided to move towards GTK3, then that's their choice, and presumably they have reasons for that choice. I don't think it's out of line for them to expect to support a single working build for an architecture/OS combo. That said, they can and should be notified and pressured to fix any regressions. Additionally, if the change is something people don't want, Mozilla should be notified of that (although I suspect there are technical reasons for the change that most people are ignoring).


Good points. Just finished building firefox 50.0 with for cairo-gtk2 and trying it out now. Hope there are no crashes.

I agree with what you say and want to add that there two issues here. First is the GTK3 port of Firefox not working as a native Wayland GTK3 window. Second is the themeing issues with GTK3 that people ran into, which is something you can live with. Finally, it's the serious regressions of GTK3 itself which go unnoticed or ignored as evident in the gnome.org bug-tracker and the tickets I've linked to in a sibling comment.

The most important argument for Mozilla providing GTK2 builds is that otherwise the likelihood of the code bitrotting is very high as it happened with the Qt port. The interesting aspect of the Qt port was that at that point in time GTK2 was the best choice, but now for substantiated reasons major applications chose to rather port to Qt than GTK3. Therefore, if the Qt port were to be revived it's much more likely to be of interest and maintained than back when the GTK2 port was good enough that nobody cared about Qt.

I know X11 and Wayland are not Mozilla's main platforms of interest and I'm grateful that they do support with the feature set they do. I'm surprised at the seemingly isolated echo chamber perspective of the GTK3 devs. It's an interesting behavior to observe since without non-GNOME users it could just as well be rolled into GNOME itself. Interesting times, having lived through the times when jwz in 1998 was debating whether GTK1 was any good for Linux.


> First is the GTK3 port of Firefox not working as a native Wayland GTK3 window.

Is that officially supported yet? I see that back in June/July experimental support was released. Possibly is the current work in an effort to get that working, but it's not ready yet? I just tracked down what looks like the feature tracking bug[1], and it doesn't appear to be ready, but I could be reading it wrong.

> Second is the themeing issues with GTK3 that people ran into, which is something you can live with.

Sure, but from my own problems with themes in Chrome, it's not something you want to live with, especially if you used themes to signify instance information and they stopped working, so I feel people's pain there. :/ Preferably Mozilla would have kept this gated in a branch or experimental build until these issues were worked out. That said, earlier today I did see something about plugins built using GTK2 being loaded into Firefox with GTK3 and the symbols loading wrong, so perhaps it's a much harder problem than it seems, and if it requires theme's to rebuild, then perhaps the quickest and easiest way to make that happen is to force a little breakage. Then again, the distro build probably makes sure this isn't a problem.

> I'm surprised at the seemingly isolated echo chamber perspective of the GTK3 devs.

I imagine that's somewhat to do with where their focus lies. I'm under the impression that a lot of the funding comes from Red Hat, so there are likely complex motives at multiple levels from the enterprise to the funded developers (who might want to justify their paycheck).

In the end, it's one of those things that's hard to accurately critique as an outsider, because there's a lot of specific information that goes into a decision like that. Is a GTK2 to GTK3 migration easier than a GTK2 to QT (or some other toolkit) migration? Probably, but by how much? If this was happening a few years ago, we might be complaining that Firefox uses a toolkit (that is, underneath their own toolkit) instead of X directly. Now we have X and Wayland, so that wouldn't have presented a similar situation anyway. In the end, unless you have some upstream vendor willing to put lots of time and money into making sure you have a performant, backwards compatible API to call (e.g. Microsoft), you'll probably want to make changes at some point in any long-lived project.

1: https://bugzilla.mozilla.org/show_bug.cgi?id=635134


The "EME-free" build differs only by the default value for the "Play DRM content" checkbox. You can put a normal build in the equivalent state by unchecking the checkbox.


Good to know. I thought it doesn't include and prevent the download of some DRM blobs.


The "Play DRM content" checkbox controls both the EME API and the automatic download and installation of the DRM blob (the Google Widevine CDM). So you are still correct. :)


FWIW, 64-bit Firefox will be the installer default for new installs (and re-installs) on 64-bit Windows 7/8/8.1/10 starting in Firefox 53. Migrating existing 32-bit Firefox users to 64-bit will probably happen in mid-to-late 2017. Similarly, Google plans to migrate their 32-bit Chrome users to 64-bit sometime in 2017, too.

https://wiki.mozilla.org/Firefox/Win64


I know you're being facetious, but Mozilla shipped Motif builds for a really long time after most of the world had switched to GTK. Far longer than seemed reasonable at the time. Given that the majority of GTK-running systems support GTK2, and that the GTK3 builds of Firefox require GTK2 already, it seems very strange that the default distribution would bother with GTK3 at this time. I'm sure they have good reasons; I'm just curious about what they are.


Do you know how to run Firefox 50 with GTK 3.22.0 and the native Wayland backend? If I try it, it crashes instantly, and I was surprised to see gtk3 and gtk2 linkage in the xul libs. Any idea?


I think the third party situation is unlikely, unless they're shipping a fork like IceWeasel or stripping out all branding information. At the end of the compilation process in Gentoo it helpfully tells you (if you enabled the Firefox Branding) that you cannot legally redistribute a branded version of Firefox.


If there are a bunch of people that need this you could do the community a service by doing it and hosting it somewhere for all interested parties.


I do not have the time nor anywhere near the resources that Mozilla has to handle this. If i had either i would not be sitting here.


If you want to avoid GTK3 and you're on Arch, there's a firefox-gtk2 [0] AUR package that compiles Firefox against GTK2 only.

[0]: https://aur.archlinux.org/packages/firefox-gtk2


My local cairo-gtk2 build of Firefox 49 reliably crashes anytime you try to use the GTK file dialog. Does it work for you (arch or gentoo)?


That's interesting. I've been happily using firefox-gtk2 and it's been working quite reliably.

Not only I haven't had any issues with it so far, it even looks better than Arch's official package (that's built with gtk3).

The issues have been present with gtk >= 3.20, and they should be fixed with release 50 and 51:

https://bugzilla.mozilla.org/show_bug.cgi?id=1266914

https://bugzilla.mozilla.org/show_bug.cgi?id=1264079


After almost a decade of building Firefox from source I finally bit the bullet to switch to a precompiled binary last month. It's fixed all the issues that have started to crop up but leaves a bad taste in the mouth.


I had been using their binaries due to security updates and having to build at inopportune times for an hour or three (depending on machine), but like I wrote in a sibling comment, enabling the cairo-gtk2 backend resulted in crashes when trying to select a file with the gtk file dialog. So I bit the bullet and am struggling with the drawing regressions of GTK3.


> Now if only i was not stuck on ESR because GTK3...

What's the issue with GTK3?


GTK3 itself has various regressions and is slower than GTK2 despite the architectural refactoring lending itself to a snappier experience.

https://bugzilla.gnome.org/show_bug.cgi?id=771708

Firefox GTK3 has regressions too, for instance the bookmark manager not remembering where you were last which works with Firefox ESR (GTK2).

https://bugzilla.mozilla.org/show_bug.cgi?id=1267863

Also GTK3's file dialog is subjectively a huge regression, but others may prefer it, just like Apple's finder changed over time. I still haven't figured out how dconf and gconf work with GTK3, not using GNOME3 as a desktop.

There are more issues with GTK3 but these are the most visible for those of us who don't use GTK3 regularly as part of GNOME3 and have been forced to use it via Firefox. The way I read the GTK3 bugreport it seems like the devs don't test GTK3 outside GNOME3 and hence do not consider it a priority. That one dev has been stressing that a compositor is needed and multiple answers by the reporter that they're using a compositor seem to be missed during reading on the other side. It's a weird exchange. I'm with the reporter. If GTK3 is not supposed to or not tested outside GNOME3, then this should be communicated so that everyone can make an informed decision to use something else or revive the Firefox Qt toolkit code.


Serious comment:why bother with gtk anyhow and move to Qt?


I love GTK2, but I can do without GTK3, seeing we're at 3.22 and it's still not as stable or regression free as GTK2.

I'd be the first to build and use a Firefox where the Qt port was updated and made to work but GTK is the GUI toolkit used by Firefox outside Android, macOS and Windows. That said, GTK2 was and still is very good at what it does. It just works but doesn't support Wayland.

I cannot move to Wayland anyway until xterm or rxvt-unicode are ported since XWayland integration is still imperfect. Like they wrote in the bug report, while GTK2 doesn't have a Wayland backend, GTK3's backend isn't really production quality either with dialogs sometimes opting to zoom out rather than scale out or general stability issues. If you try to start Firefox under Wayland by telling GDK to use Wayland, it just crashes on startup.


After switching to GTK3 version (Arch) I noticed it is much more sluggish and feels slower than previous versions. There are also some inconsistencies or just breaks in interface (I assume it is related to theming).


It's fairly simple to build Firefox with Gtk2. There are some instructions here: https://www.reddit.com/r/linux/comments/5b0vtd/howto_build_f...


"CVE-2016-5292: URL parsing causes crash" made me curious, given the talk I've heard about rust-url being integrated into Firefox.

For anyone else interested, it appears that the patch is still being reviewed - https://bugzilla.mozilla.org/show_bug.cgi?id=1151899 (though I don't know if rust-url would have actually prevented the issue).


I wanted to try reader mode, so I went to a random Atlantic article (https://www.theatlantic.com/health/archive/2016/11/better-of...) and hit ctrl-alt-r. The graph related to the story was dropped. Would that be a bug, or is that how reader mode is supposed to work?

I don't have the equivalent Evernote plug-in anymore (I'm trying to get away from Evernote), so I have nothing to compare it to.


You can report a bug in Firefox Reader Mode here. Identifying the "important" content on a page is a hard problem. :) FWIW, Safari's Reader Mode doesn't include the article's graphs or pictures either.

https://bugzilla.mozilla.org/enter_bug.cgi?product=Toolkit&c...


I just wasn't sure if the image being removed is by design.

I found another article where it leaves the author photo but removes the main article photo (https://www.washingtonpost.com/opinions/david-camerons-wizar...) so I believe it is a bug.

BTW, I think this bug covers the issue (it's where I got the WP link from):

https://github.com/mozilla/readability/issues/219

Edit: Actually, this is a better bug report:

https://github.com/mozilla/readability/issues/299


I've noticed this behavior with things like Pocket as well.

Is there any evidence you've found of any publishers intentionally trying to load assets in a way that would break with this because it is stripping some ad impressions?


I don't think so. Medium, for example, loads images on demand. So if you never scroll down, I don't think it ever renders the bottom of an article. I don't think it's malicious, just conflicting goals.


Did something change in font rendering with Freetype on Linux? While 49.0.2 with FreeType 2.7 looked the best (subjectively) I've seen any font rendering (including Windows 10 and OS X Snow Leopard (been a while)), something off with 50.0's rendering. I haven't enabled any custom render options, just your typical archlinux freetype 2.7 desktop. Time to get ESR and compare with that but would be great to hear back from others with a similar environment.


Freetype 2.7 introduced a new rendering engine. I cannot link anything at the moment but the Arch Forums have pretty detailed help.


I love 2.7's rendering engine and 49.0.2 looked great, but 50.0 looks washed out. Tried ESR 45.5.0 and it looks great again. I'm confused as to what may have happened from 49 to 50 that caused this.


Built 50.0 with cairo-gtk2 backend using it right now but something different with the font rendering in 50.0. For instance there's a weird washed-out effect of the text in this very same text control I'm writing the comment in which comes and goes with the amount of text changing/wrapping etc.


Can you post a screenshot?


Sorry I can't because it's something that happens and doesn't stay. If you run Firefox 50 on Arch Linux, you will see what I mean.


I think I figured it out. Somehow Skia was enabled in Firefox and resetting that seems to have fixed the rendering bug. Given that 51 will turn Skia on by default I hope this will be fixed before the switch.


Not much for users but devs can expect lots of changes https://developer.mozilla.org/sv-SE/Firefox/Releases/50#Chan...


https://www.mozilla.org/en-US/security/advisories/mfsa2016-8...

PSA: Please update immediately. One of these is critical.


So can I sync the reading list and use it later offline, both on desktop or mobile? I've looked for the reading list in about:about but I can't seem to find it.


I find that Firefox has a considerable lag on Macbook 2012 compared to Safari. I wish that would change because I love firefox and don't want to use Safari but Safari is just nimble


I had to stop using Firefox, because the 64-bit version apparently exists only to consume all of the memory on my PCs, not just some of it, like the 32-bit builds would.

I'm using Iron Browser now, which has its own issues, but which at least handles a hefty amount of tabs and some essential extensions without destroying system performance.


I use firefox on android and sadly, I have to use firefox on desktop too because of bookmark/history sync, despite the fact that it is slow.


Breaks NoScript on launch day. That was an unpleasant surprise this morning...


Any reason you prefer NoScript to uBlock Origin's script blocker? Just curious...


I don't think they are comparable. I run both. NoScript is a security suite – besides blocking java, webgl, flash, silverlight, javascript, etc – it has additional defenses against XSS, ABE, clickjacking etc.

uBlock was to my knowledge never developed to securely stop scripts and deter drive-by attacks etc. It should be used for adblocking, not for security.


...or uMatrix with its fine-grained control.


The fact that NoScript supports script surrogates might be a reason. That means less websites will break when not enabling scripts that are expected but not really necessary.

http://www.ghacks.net/2015/11/24/noscript-script-surrogates-...

>When you visit a site in Firefox that loads the Google Analytics script on page load, NoScript intercepts that request and replaces it automatically with the replacement instructions (which basically tell the site that the Analytics script was loaded fine but does nothing in regards to user recording).


Does Firefox 50 break NoScript on all sites or a specific case? Can you share steps to reproduce the regression?


The update for Noscript last week fixed this issue.


I'm on the Aurora channel currently running 51.0 and NoScript works fine for me.


I'm on the Nightly channel running Firefox v 53.0a1 (2016-11-15) (64-bit) on a Linux box. NoScript works for me too


- Added Guarani (gn) locale :) Thank you


Tereiko porãke! Rojhayhû. ;)


Regarding the new Referrer-Policy header introduction, what happens when my network.http.sendRefererHeader is 0, network.http.sendSecureXSiteReferrer is false and some website sets a Referrer-Policy: "unsafe-url" header? Which setting has the priority?


The logic is basically like so:

1) Is the site requesting a "no-referrer" policy? Then send no referer.

2) Is network.http.sendRefererHeader set to a value that would prevent sending of referrer in this situation (e.g. 0 in all situations)? Then send no referrer.

3) All the other logic (but generally aiming to follow the most restrictive directive we have).

The "network.http.sendSecureXSiteReferrer" still exists in 50, but is gone in 52; see https://bugzilla.mozilla.org/show_bug.cgi?id=1308725


Thanks for clarifying, that's logical of course (user preferences should have priority).


It's great that Firefox finally has this option and I'm surprised that so many people dislike it; it's the single most important feature why Opera is my default browser.

I guess the reason why it's confusing to people is that Firefox (just like Sublime) doesn't have that little popup displaying a list of tabs while switching, it would make all the difference in the user experience, and make things clear.


What are people's opinions of it's developer tools features compared to Chrome? in particular, how does it compare in terms of running a HTTP trace?


The bottom of the page says "Congrats! You’re using the latest version of Firefox.", but Firefox's About Firefox box says 49.0.2...


Are you browsing with Firefox for Android? If you click the "Android" tab, it still links to the 49.0.2 release notes.


Ah, they've updated the page, now it's a "Download" button.


I'm in the process of switching to Pale Moon because Firefox is making changes that will break Pentadactyl.


Are you sure that this is the case? There's a lot of misinformation out there about Firefox's extension changes. I don't know about Pentadactyl, but Firefox should continue to support APIs providing equivalent functionality for most plugins. They will have to transition, but they won't be irreparably broken.


Are the Pale Moon guys able to keep up with the ever evolving web standards? Do complex websites work as expected in Pale Moon?


Don't see too many programs going to version 50 these days. Kudos to the team that keeps improving it.


With an aggressively "encouraged" monthly version treadmill, it's less impressive than it might look. If Linux used the same type of software force-feeding, it would be well past version 200 by now.


Er... Linux does use a similar time-based release schedule now, and the minor number has been reset a couple of times to keep it from "getting too high".


A version 2.0 is a feat these days. Reaching 50.0 is amazing.


I was hoping that the new tab spinner problems would be addressed in 50. Oh well.


Still no tabs on the left.



Thanks. The tab handling is the only reason I still use chrome.


Chrome has tab on the side now?


No, but I like the way it handles tabs much better than firefox and last time I looked into it, it was pretty cumbersome to customize firefox this way. Plus it has to be done on every machine.

This extension is even better IMO. And so far I'm enjoying having better tab handling and a much faster browser.


I don't understand how can one browse internet with tab not on the side :|


Wow, I'm so happy right now. It doesn't seem to handle parents/children though. (nesting tabs)


And works like a charm


[flagged]


> Jesus Christ why, just why. I'm so done with Mozilla.

I'm afraid "just why" is not enough. Please comment civilly and substantively on Hacker News.

https://news.ycombinator.com/newsguidelines.html


I imagine he actually meant that Firefox/Chrome/etc are prioritizing a different demographic of users now than they used to. Both are becoming more of an OS in which you run web "apps" rather than just a browser to display and read webpages. Web apps need a consistent base set of features that may not exist on all OS platforms. By providing it internally they signal who and what the browsers are designed for. Other people might consider these features bloat.

To answer his legitimate question: try Pale Moon. It's a browser that respects your freedom and does not include bloat.


Qutebrowser is also worth checking out. I haven't really given it a full blown test drive yet, but it's certainly less bloated than Firefox and Chrome, and has Vim-like keybindings.


I've vim keybindings on my Firefox through a plug-in... I almost forget I have a mouse


Is it Pentadactyl? Does it work on the latest versions of Firefox?


I use VimFx which doesn’t modify the browser chrome as much as Pentadactyl (apparently). It simply provides lots of convenient Vim-like keyboard shortcuts. I also use the ItsAllText! extension with GVim configured for editing text boxes. I'm not aware of any issues that either have with the new extension architecture.


Still relatively slow startup, even on my quite powerful computer (i7, SSD, 8GB RAM, the second start, everything already cached in RAM, no add-ons or extensions but the defaults, no sessions restored, open to empty tab) I have at least a whole second during which I don't see that the program is starting.

At least visibly better than the 49.


Firefox's reproducible automated testing begs to differ https://arewefastyet.com/


None of the tests there are startup time.


Of all the things to optimize, especially on the "modern" web, I doubt startup time is anywhere near the top.


May be, but this one is what matters to me, and I report it.


I started using Mozilla Firefox 1.04 many years ago. Everytime a new update was released I was so excited to install it, as the browser became better and better with each update. I remember when Firefox 2.0 was released and I could not wait for it to be available in the Debian package repository, because it was such a big improvement.

Now I absolutely HATE when a update is released, because either functionality is removed, appearance is changed or even more bloat has been added.

Firefox itself has become the problem it tried to solve.


What in this release would you class as functionality removal, appearance changes or added bloat?


This is in the the bloat category:

Added a built-in Emoji set for operating systems without native Emoji fonts (Windows 8.0 and lower and Linux)


Do you prefer that web pages using Emoji not render as expected for people on Linux and Windows XP/Vista/7/8?


Do you think that's a useless feature though? Seeing squares all over the internet isn't exactly a good experience, finding the right font sets to fix your emoji problem can be tricky on Linux too

EDIT: just got the update and it a) fixed some emojis I somehow didn't have, namely "thinking face", which is of course critically important and b) made the rest look a lot livelier.


I'll take squares over emojis any day.




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

Search: