Hacker News new | comments | ask | show | jobs | submit login
Firefox 65.0 released (mozilla.org)
724 points by theodorejb 19 days ago | hide | past | web | favorite | 414 comments

Recently switched from chromium to Firefox. Very pleased with its performance.

Android Firefox demolishes chrome too but probably just because I no longer need to download and render all the ads (thanks ublock.)

While I was at it I also took the chance to migrate passwords out of Google land into Bitwarden.

Session restore is amazing, my .cache is tmpfs but it still manages to instantly recover all my tabs on reboot. Tabs are addictive in FF and don't have the same usage constraints like in chrome.

The container system for managing different developer/test/personal profiles is a dream. No longer have to worry about Slack links opening in whatever random chrome instance had focus last. (Back up your user data, sync on these container settings doesn't work - yet)

Also enjoying different proxy profiles per Firefox profile as well.


Work profile + work containers + works SSH tunnel proxy.

Personal profile + personal containers + NordVPN.

Passwords managed by bitwarden and any required crossover is shared via a private organisation. This gives me minimal work access on my phone, and minimal private passwords on my work profile.(GitHub/stack overflow)

It's kind of complicated but it's such a quality of life change coming from 20 odd chrome profiles and the nasty sync issues that ensues. Most people can probably get by with a single profile. I just have multiple systems and need to keep work stuff separate but end up using work laptop as a daily driver most of the time.

I've pretty much made the same changes. I was luke-warm on the switch from chrome to Firefox until I started using containers and CookieAutoDelete with those containers. I have a 'Google' container that preserves google cookies for when I log into gmail (transition in progress)- but otherwise google cookies get automatically deleted. There are a handful of other sites I wish to stay logged into - like github, slack, spotify - and I have containers for them with CookieAutoDelete configured accordingly. Its also comes in handy for sites where I have a work account and a personal account - like AWS. Instead of having to open an incognito window to allow being signed into two accounts at the same time, I just have separate containers. Makes a lot of practical sense for me.

I also made the switch to Bitwarden - except for me I was using Lastpass before. Bitwarden has a much cleaner interface and a sane sharing scheme. Trying to keep shared passwords in sync with my wife in Lastpass was an absolute nightmare - to the point that I hated changing the passwords - which is one of the major points of a password manager. Lastpass has become stagnant and bloated since LogMeIn bought it.

You like the "Temporary Containers" extension. It automatically creates a new container for every site that doesn't already have an assigned container. The temporary container (cookies, etc) get deleted 30 minutes after you close the last tab for the container's site.


I use Sandboxie with a browser instance without cookies or cache. It automatically deletes files after closure. Only the browser is allowed internet access.

For specific services like my primary email, I have an isolated Firefox instance keep the cookies for that service only. I don't use that instance for anything but that service. Is there anything I'm giving them (besides what information you give by using the service at all of course)in keeping the cookie? Just want to be aware of any risk I may have overlooked.

I just found my new fave FF extension. Thank you for sharing this!

Was similarly lukewarm to the change. I actually only made the change because I was having resource issues with chrome and bookmark interactions would lock the screen for a few hundred milliseconds - tried everything. Eventually you just blame the vendor and switch.

But now, well, I don't really see me going back any time soon.

So Bitwarden is basically Lastpass except you can also host it yourself?

You can host it yourself - I don't. The interface is much cleaner and easier to use IMHO, especially if you use password sharing. At the end of the day its a password manager, and there are plenty to choose from. I used to be a big Lastpass fan - but their quality has continued to degrade and competitors have stepped up their game. If I didn't end up using Bitwarden, I probably would have ended up with 1Password.

Yea but you don't have to. I don't have a lot of experience with 1pass/last pass but I assume they're the same. I liked that Bitwarden was free, that I could do my own server, and that I can do dual account sharing via org setup - all free accounts.

What are you going to transition to from Gmail?

Not the original poster, but I would recommend to anybody that they move to Fastmail.

My only concern with Fastmail is they're Australia-based and the government there seems rather hostile to privacy.

This is exactly why the new encryption laws are so detrimental to the Aussie tech industry.

I've been on Fastmail for just over a year and it's been very reliable and effective.

another vote for Fastmail from me

I'd be careful with that one, due to recent developments in Australia, more or less banning encryption (and legalising subverting employees of companies without the company's knowledge, to break their encryption schemes secretly). Until we know more, not a single Australian product can be trusted and even Australian employees of companies working overseas can't be trusted anymore. It's horrific.

what about protonmail?

I have a protonmail premium account and I'm disappointed. First you can't use a native client in linux because the bridge isn't released yet (in the faq it's written it's planned to be released early 2018 lol). The android client doesn't like the fact that I desactived the google play service on my phone (how can a privacy focused email depends of the google play service? ). They hooked me up because they said that they where going to open sourcing protonmail, but for the moment neither the bridge nor the android/ios client are open source.

I unfortunately have to second this.

The bridge solution is a nice attempt at supporting open standards, but it's not on Linux or Android which basically means my email is silo-ed. I can't use it with my regular mail client and the ProtonMail client will never be all things to everybody.

The android client at least does seem to run just fine without Play Services, it just pops up annoying notifications saying it needs them.

At this stage I'm probably looking at migrating away.

I'd really like a mail provider that lets me forward a few addresses to the rest of my family since I own lastname.nz, but that doesn't seem possible on most mail hosts without a full x-user enterprise setup.

Protonmail limits the number of domains you can use. Fastmail has a much higher limit.

I'm currently giving ProtonMail a try - but I'm also using a custom domain. My current thoughts are that I'll switch everything over to my custom domain, and then I'll never be locked into an email provider again. If ProtonMail doesn't pan out, I'll switch to someone else - still keeping my custom domain. My reliance on google and my attempts to move out of it has really been an eye opening experience I hope to not repeat.

I went with Tutanota.com which on top of active development also communicates regularly with its users on what features are being worked on. Plus they allow you to have/create a bunch of different email addresses under the same mailbox.

Last I checked, containers weren't working properly, and I gave up on the idea because false sense of security is worse. My test website was web.whatsapp.com. I think this is the tracker, although I didn't really look too much into it (https://github.com/mozilla/multi-account-containers/issues/8...

Edit: fixed link

Have you tried that again since September 2017? I can’t seem to reproduce the issue on my machine. (By the way, the link includes an extra underscore. The correct one is https://github.com/mozilla/multi-account-containers/issues/8... )

Yes, I can still reproduce it on 64. I'm on linux though.

Pretty sure I figured this out, since I was able to reproduce it as well. It has to do with caching and something Whatsapp is doing to restore the page from local data most likely, as once you clear the local data for web.whatsapp.com, it correctly requests redirecting to the correct container.

If I had to guess, I suspect that Firefox triggers the container switch dialog on the first network request, and that page is so optimized that after the first visit it loads entirely from cache and/or localstorage data without any network activity at all.

If true, it might be that while it didn't switch, it actually wasn't leaking data between containers at all, since there was no network activity. I'm not sure if a background request would have triggered a container switch dialog, been blocked quietly, or have been allowed through some root page permissions cascade.

I may be far off, and this is trivially checkable, but I'm out of time.

I had a look at the github issue, and what worked was this comment: https://github.com/mozilla/multi-account-containers/issues/8...

You can try this if it affects you.

On Linux, works for me. Using add-on version if that helps. There is a bug that opening a url via e.g. xdg isn't able to prevent default tab opening on first startup.

Im also suprised by the firefox performance on android. Ive noticed it can render pages chrome cannot, which is so suprising considering googles investment in progressive web apps and how some of their new android go apps use chrome to render.

+1 for Bitwarden - it's been fun to watch that product unfold.

I’ve wanted to try BitWarden, but haven’t because I couldn’t find a way to use it as a standalone app without creating an account on its site or without having to self-host it. I prefer a solution that allows me to choose where to sync the password database.

Just run a bitwarden server locally, that's what I do.

Especially easy with bitwarden-rs or bitwarden-go (written in rust and go accordingly) instead of having to pull down the bitwarden blessed MS SQL stuff.


I made another attempt to switch from Chrome to Firefox only now, like I've been doing every year or so for several years now. Sort of a tradition.

This time it was the extensions that didn't cut it. One of the first things I do (after importing the bookmarks) is to install the same extensions, or find their equivalents.

I heavily use customized mouse gestures. The extension I use on Chrome works like a charm. On Firefox, I tried five in a row.

They would either be very poor in terms of options and customizability, or request 10+ bizarre permissions (which really have nothing to do with functionality I expect from them - such as "read the text of all open tabs", or "monitor extension usage"), or very cumbersome to customize. I've really got fed up after trying out several in a row.

Well - I'll wait another year.

It might be worth contacting the author of the Chrome extension and asking them if they'd be interested in porting; since Firefox's WebExtensions API is a superset of the Chrome extension API, it should hopefully not require too much effort.

There's also an extension called "chrome store firefoxified" that lets you install chrome extensions in firefox. Might be worth a try.

> I made another attempt to switch from Chrome to Firefox only now, like I've been doing every year or so for several years now. Sort of a tradition.

Same here! (except i'm using Brave instead of Chrome, it's also Chromium based and supports its extensions, without the Google tracking features.)

I don't understand why after so many years Firefox still doesn't support "pinch to zoom" on Macbooks/Laptops. For me it's such a basic feature, almost like a car without a proper steering wheel... not very smart if you want to attract more users.

At last, someone created an extension ("Multi-touch Zoom" https://github.com/haxiomic/firefox-multi-touch-zoom ), but unfortunately, it often doesn't work properly, like for example on this website...

On all my Linux systems (NixOS), Firefox lags/skips while scrolling all the time. It's a deal breaker for me so I use Chromium, which scrolls without lagging. I read it is because Firefox does not use HW graphics acceleration on Linux, in general. This is both on a computer with the NVidia driver and one with the open-source radeon driver.

Hardware acceleration is disabled by default because many graphics drivers suck and will break the browser if hardware acceleration is turned on. You can just go into the settings and turn it on (and then off, if you have problems).

Chromium disables hardware acceleration by default for a wide swatch of graphics drivers too (including all Nvidia GPUs).

On my current Linux system with NVidia drivers, HW acceleration is enabled according to chrome://gpu (there are many accelerations and some are disabled, but many are enabled including Canvas and Compositing). Regardless of all, Chromium works just great, including smooth scrolling and WebGL, and Firefox doesn't (laggy scrolling, slower WebGL than in Chromium). I don't think the problem is with drivers but with Firefox.

> Chromium disables hardware acceleration by default for a wide swatch of graphics drivers too (including all Nvidia GPUs).

From what I remember reading, it disables it for nouveau driver, not the NVidia driver which works. This is supported by the fact my Chromium is currently using HW accel on NVidia without me having forced anything.

I did try enabling HW accel in Firefox and it did not solve laggy scrolling. I don't think it even really got enabled, I think it's forcibly disabled even if you try to override it.

you're correct, it's definitely just nouveau that's blacklisted, not the proprietary nvidia driver.

Chromium also has hw accel enabled by default for intel drivers, which is by far the most used gpu driver on linux (and is pretty stable in my experience). Firefox doesn't even have it enabled for intel

Firefox has an option to uncheck Smooth Scrolling. So if you press the spacebar, the page instantly jumps down instead of showing the page quickly scroll. I don't know if this would help.

This is the exact reason I have not been able to switch back to FF, but on Windows. Enabling HW acceleration did not fix it.

I'm using FF on AMD and Intel GPUs, and works smooth

I hadn't heard about Bitwarden. Looks like it requires 2-4GB of RAM and a x86 CPU to run your own server due to relying on MS SQL server. Bummer!

There's an unofficial rust implementation[0] that runs in a docker image and uses sqlite for persistence. I don't know what the details are for storing data at-rest in the sqlite db, but it's supposed to be much better on system resources.

0: https://github.com/dani-garcia/bitwarden_rs


The 2-4GB RAM isn’t a hard requirement in my experience. I initially had the whole Docker compose stack running on a single $5 DigitalOcean droplet, and didn’t experience any issues for personal usage. Since they also distribute it to you as Docker containers, you can also set resource limits, specifically for the MSSQL container, to prevent it from hogging all the RAM.

I currently have my Bitwarden instance hosted on my Docker swarm cluster, which is just 3 $5 DO droplets and 1 with like 2GB and RAM and 2vCPUs. But, a quick look shows that currently all the BW containers are running on the $5 ones. I have the MSSQL container’s RAM limited to 1GB as well.

As for the x86 requirement, that’s likely the case, I’m assuming you mentioned it because you’d like to run it on some ARM device like a RPi. I haven’t tried running their containers on an ARM device so I can’t say they’ll be compatible, but I know docker’s ARM compatibility has improved so maybe someone has got them running on some ARM device.

I've been pretty happy with https://www.passwordstore.org/ Keeps passwords encrypted in git, and there's plugins and open source apps for just about everything. Throw your store in a private repo somewhere and you're good to go, no need to worry about Bitwarden or Lastpass or anyone else going out of business.

If you aren't hitting it hard (hosting more than a few connections) you won't notice much limiting to 1GB RAM... I run a couple projects with MS-SQL docker containers and 1GB limit...

    docker run -m 1GB --restart unless-stopped --name sql -h sql -e "ACCEPT_EULA=Y" -e "MSSQL_SA_PASSWORD=Let-Me-In" -p 1433:1433 -d microsoft/mssql-server-linux:2017-latest
I tend to change the name and port mapping as necessary depending on use, but that's the quickest way to get a server up for MS-SQL... can use SQL Server Management Studio (windows) or Azure Data Studio [1][ (electron, formerly SQL Operations Studio) which is cross platform[2].

[1] https://github.com/Microsoft/azuredatastudio [2] https://github.com/Microsoft/azuredatastudio/releases

You can run the Bitwarden Rust implementation on a Raspberry Pi with resources to spare.

The requirement is 2GB. Where do you see 4GB?

Can you expand on how you set up the profile + containers + vpn and how that works for you?

Think he means this addon: https://addons.mozilla.org/en-US/firefox/addon/multi-account...

At least that's for controlling them easily. I use it for Whatsapp web on 2 accounts

Yup that's for containers. It looks like development sorta died on it but I think they will circle back once infrastructure work has been done for sync.

For proxies you can use whatever vpn service or plugin[1]. I use paid Nord because I need consistent access to certain countries for testing. There is some tunnel bear/Nord like plugins that have lists of free proxies.

1. https://addons.mozilla.org/en-US/firefox/addon/proxy-switche...

Profile manager info:


See other comment for container/proxy info.

> Passwords managed by bitwarden

Is there a specific reason you don't use Firefox' password manager? (Read: Do I have reason to migrate out of it?)

It's obvious browser password managers are broken in a way it never locks its database, so anyone with physical access can log in to all the site you have passwords for, which also means changing password is easy as logging into your gmail also requires no knowledge.

To this day I can't believe none of the browsers actually try to fix this by at least locking the password database after a certain period like any password managers do. Safari at least disallows you from looking at the list of sites you have passwords for by requiring you to enter your account password but never asks you anything at each site's login.

For this reason, I never use browser based password managers.

You can protect your password DB with a master password on Firefox, though it only will be 'locked' when you close the program. AFAIK this is not too dissimilar from the model for most password managers.

If there was an option to import chrome passwords into Firefox I couldn't find it.

> Read: Do I have reason to migrate out of it?

No, but I suggest you try it out for a day or two, see what you think.

Doesn't Firefox give you the option of importing everything (bookmarks, passwords, etc.) from another browser the first time you fire it up?

IIRC, it also does this the first time you launch it under a new profile.

To log into mobile apps, for Wifi passwords, file passwords, etc.

How "safe" and separated are profiles?

For example, suppose I had a "Casual Browsing" profile and a "Banking" profile.

Could I have both profiles running at the same time and not risk any contamination?

If I were to pick up something bad in one profile, could it leak over to the other?

Unless 65 is faster it's not worth it to me... I work very fast, like Chrome, like other apps on my OS, and there's no excuse for Firefox to run slow like it does on my Windows and Mac still.

Are profiles something available on vanilla Firefox (no extensions)? Thanks!

You can open about:profiles and use it to manage/start them, I think

Yes but it's clunky compared to chrome.


firefox -p

Ordirect access:

firefox -p "account name"

I believe containers are available without an add-on using aboutconfig but that may have changed or it may not be full functionality.

And yet Firefox still doesn't have pull to refresh. How can they not support such a basic feature?

I don't particularly care for pull to refresh, given how frequently I refresh without meaning to because of it, but I do wish I didn't have to open the dropdown menu to get a refresh button to refresh the page.

Firefox is faster than Chrome on Android? For me, even scrolling HN (i.e. no JS at all) was jittery and laggy.

I think you're both right. On my Essential Phone (Snapdragon 835 running stock Android 9) pages tend to _render_ quicker in FF, but scrolling, pinch to zoom etc. is smoother in Chrome.

I also find that Firefox will bog down after a month and I have to manually delete out its cache in app settings. I haven't had to do this for Chrome since back in the Android 5.x era.

The latter might also be due to extensions

I could see that.

Not the issue in my case though.. I don't use any extensions on my mobile browsers.

Heh, maybe it's the opposite then! Like an accumulation of shitty trackers, because you don't use uBlock or privacy badger or the like. That would be pretty ironic

Trackers do not really accumulate though, no? Or can they install something persistent (web workers maybe?)

Store stuff in IndexedDB?

Scrolling in Firefox on Android is smoother, ie. more consistent 60fps, than Chrome for me for a few months now. It used to be very bad but suddenly improved dramatically.

Huh, your right. Scrolling in Firefox on Android was awful for years, and was the main reason i stuck with chrome. I'll have to give it a proper go now...

In my experience the minimum hardware baseline for Firefox is higher than Chrome but the optimal hardware baseline is lower.

I switched to Firefox on Android a few years ago and it was unpleasant until I upgrade my phone. At the time I was running a Nexus 4 and moved to a Nextbit Robin.

My partner switched to Firefox on a Google Pixel and regularly complains about issues with it's performance which makes me wonder if Google has done some optimizations to Chrome on the Pixel.

I feel like it has gotten a bit better with this update. Or, at least the release notes mention better scrolling performance and I'm imaging it.

If you really think Firefox is faster on Android, that you have not tried the optimized Kiwi Browser. Try it, and report back how much faster kiwi is compared to Firefox hehe


I’m a diehard Firefox user and supporter, but with every release there are still things I wait on the WebExtensions support to improve:

1. There’s still no full featured Tab Mix Plus possible with the current APIs. That’s been a big bummer for some years now.

2. The most downloaded session management extension now is still not as rich and good as the old XUL extension Session Manager (from mozdev).

3. I probably have to search again for a WebExtension equivalent for this one. Lazarus was a nice form saving extension in the past. Not sure if something similar exists or is even possible.

Best not hold your breath. You are an old-school power user and 99.9% of users no longer know or care about any of the functionality you're talking about.

I don't have any specific retorts to your points other than pointing you to https://addons.mozilla.org/en-US/firefox/addon/textarea-cach... for a Lazarus replacement (although I thought Firefox does this by default, and I don't understand why you need an extension at all).

Thanks for the pointer to this extension. I'll check it out. One of the things Lazarus provided (at the cost of adding some effort in maintaining privacy) was to save the history of form fields across sites and across restarts. In my knowledge, Firefox allows saving text fields, but not text areas. Periodic saving and history were quite helpful especially when writing longer walls of text like on HN comments or posts on other platforms.

I also come to the comments on every Firefox release to see if anything improved on that front. We've been waiting for years indeed, and I guess the only way to get the functionality back is to maintain a set of custom patches (forking basically)... Still disappointed about pushing web extensions with no alternatives available at the time, and if I had looked into the future, I'd have thought the crystal ball was broken as there is still no compatibility years down the line.

Have you looked at https://www.basilisk-browser.org/ ?

It's basically XUL Firefox w/o WebExtension crap (they played with WebExtensions and bailed).

Made by the same people who make Pale Moon, it's basically a successor.

The main downside, of course, is the XUL extension authors are unlikely to keep updating their extensions now the XUL is dead on OG Firefox.

I've tried the 64 bit version of Pale Moon several years ago and it didn't seem as performant as Firefox (could've been due to the setup and environment; I didn't bother to investigate deeper). The fact that many XUL extensions have been abandoned or have stopped development is a concern though. Thanks for the Basilisk mention. It may be a good alternative for older machines that the current Firefox doesn't support.

FF65 also supports WebP images and animations. Here's few test images: https://developers.google.com/speed/webp/gallery1

After 8 years, thank god! Image heavy pages load 40% faster now, like my photography index with lots more test images: https://shan.io/photography/

I wish they implemented the feature sooner. AVIF, an image format based on the new AV1 video codec with even better compression is already on the horizon.

Yes. As much as I was hoping they'd implement it back then, I kind of wish that they didn't implement it now. If they couldn't do it years ago, might as well wait for AVIF instead of letting webp proliferate.

Cool, thanks for the tip. Went to the page and was disappointed to see broken webp images - then checked Google play for updates, got the latest ff - went back and could see the webp images :) Finally.... We've been waiting too long...

I was intrigued by this statement:

> A better video streaming experience for Windows users: Firefox now supports the next-generation, royalty-free video compression technology called AV1

Since I don't care about Windows, I was curious as to whether this had already shipped for other platforms or not. At least using Firefox Developer Edition v66, it works fine behind the media.av1.enabled flag on Linux.


Also seems that https://www.mozilla.org/en-US/security/known-vulnerabilities... has yet to be updated.

> I was curious as to whether this had already shipped for other platforms or not.

AV1 support is available on all platforms behing the flag; this release just enables it by default on Windows [1].

[1]: https://bugzilla.mozilla.org/show_bug.cgi?id=1452146

AV1 also uses the brand new media codec sandboxing feature (RDD process). It has platform specific code, hence why only Windows was ready by 65 (but Mac + Linux should be available soon).

> Support for Handoff on macOS: Continue browsing across devices. Pick up where you left off with iOS (via Firefox or Safari) on Firefox on Mac.

OH YES! Finally!

This alone makes me consider Firefox again!

Handoff (when it works)is just so useful.

Still a little flaky even for apple apps so interested to see how it works on Firefox.

I tried both Safari and Chrome of macOS but went back to Firefox for one simple reason. When watching streaming content if the browser was not in the foreground it would reduce the quality of the streamed content which could take almost ten seconds to revert back when focus was given.

Firefox doesn't seem to know/care if its not top window

Safari user, always play videos in a background window, never seen this. I suspect it’s not the browser’s video playback but perhaps more to do with how it handles spikes and drops in connection quality?

I didn't realize Handoff was even possible for third party apps!

I suppose there is no way to make it work with Firefox on Windows and iOS? That would be awesome!

The architecture for Handoff isn’t available on Windows (and it would be much more complex than a simple library since so much of the networking and Bluetooth stacks are entirely different between platforms), so no. I think Windows has its own, similar thing but, of course, iOS wouldn’t know how to communicate with it.

I have turned on Webrender in FF66a (nightly) and it’s fantastic; it sped up a site I quickly threw together for a friend no end [1] and almost everything with lots of graphical painting is buttery smooth.

Even though they are only starting in nightly to make it the default on Nvidia hardware it seems to work without incident so far on this 2016 AMD Radeon’d MacBook Pro.

[1] https://uparchitects.co.uk/

[2] https://discourse.mozilla.org/t/have-you-tested-webrender/16...

Well the limit in nightly to nvidia on windows is less about saying “we think WebRender only works/is stable on nvidia on windows” than it is about saying “we want more people to try it and possibly send big reports but we don’t want to risk breaking the entire nightly population or have them all sending in big reports, so we will start with a small subpopulation where we think bugs will be due to WebRender rather than graphics drivers”

Ah fantastic, definitely good to stagger the releases. I do understand the driver/stability situation being much more fragmented on Windows, just nice to be able to confirm how big the improvements are to Firefox that are still to come; the fastest browser by a long way IMO!

Is there any way to enable it outside nightly builds?

Yes the same instructions should work but you might get a more buggy version of Web Render as it'll be an older less debugged version.

I don't understand why Firefox doesn't include an option to download the .deb package instead of tar file for Ubuntu and Debian during release announcements. The market share is already shrinking and why make it harder for people to install or upgrade Firefox? Chrome does it. Why can't Firefox do that as well?

If you install Firefox from the tarball, it a) doesn't require root permissions and b) autoupdates in place, so I think it's easier to install and upgrade than if you got it from a .deb.

(Chrome's .deb includes configuration to add Chrome's apt repository; I'm not sure if Firefox's does.)

This brings up an interesting point about releasing software in general. If you spend huge amounts of time programming cool things and tuning every last knob, but then it is difficult for your average user to find out how to install and start to see all the cool things you've produced, when it would take far less time to just make an easy installer than it took you to make all the cool things, then there needs to be some balance in usability. The pipeline to getting the software in front of someone is sometimes more important than adding new cool features. A basic thing that is easy to use will get more adoption than a mind-blowingly awesome thing that only sys-admins can figure out how to use.

Which would be a valid point, if Mozilla's definition of "every last knob" didn't include such basic features as "customizable keyboard shortcuts" and "allowing unsigned side-loading of addons", which the product already had for the 15 years ending 2016.

"Ubuntu and Debian" "average user"

I dont disagree with your point, not sure it is relevant however.

The pretense that all Linux users should be (or already are) technically capable is really harmful to the ecosystem. No, you should not expect Ubuntu or Debian users to figure out on their own (1) why the tar.gz file they download doesn't do anything meaningful and (2) how they will get this version of Firefox instead without any further instructions.

Running Firefox from a tarball is a very specific use case for a small subset of highly technical users. Even if you figure out how to do it, the lack of system integration gives a very poor impression of the way things work on Linux. Even the vast majority of technical users don't want to run Firefox this way. Yet that's still what the website offers you.

We should be able to recommend at the very least Ubuntu to average users. And people do that, sometimes quite successfully. But with the attitude that all Linux users can save themselves - "we don't need to think about the UX" - we can't really keep doing that.

I don't think we should be encouraging an average Ubuntu user to download and install anything from source that has an actively maintained package. Average users of Ubuntu/Linux should only use the maintained Firefox package that comes pre-installed on basically every major dist these days.

I think Firefox or anyone else would be doing a disservice to an average user to make them think they need to manually install this software outside their package manager.

Wouldn't the average Ubuntu user be best off by getting it through the Ubuntu package repo once it's updated there? Or doesn't Ubuntu have a channel that would provide that soon by default?

Yes. The Firefox website should ideally provide some instructions - including a way to open the Firefox entry in the Ubuntu/Gnome Software Center, instead of giving users a tarball most don't know how to handle (and if they do know how to handle it, is not necessarily what they're looking for). The tarball should of course still be available as a download option, but almost all interested users (yes, even the technical ones) should use their distribution's package manager.

Well, yes and no. I'm precluded from installing new Firefox via apt-get because my Ubuntu version fell off active support. Running a dist-upgrade is risky, and I don't feel like doing it until I know I have a spare free day to fix everything in case of a problem. A Firefox PPA would be a nice thing to use instead, for now.

Average user: "Pee pee what now?"

If Firefox offered a .deb for $OS on their homepage, and Grandma attempts to install it while running $OTHEROS, then likely she will either 1) fail or 2) break her system. That's a worse user experience than a .tar.gz.

Asking Firefox to maintain packages for every outdated Debian-based distro is unreasonable. Nice to have, yes, but unreasonable to ask.

> No, you should not expect Ubuntu or Debian users to figure out on their own

If someone knows what Debian is and has chosen to install it I think its fairly safe to assume they have some Linux knowledge.

Ubuntu on the other hand is the most likely to work without any tweaking. Its the default choice for less technical people wanting to try Linux as well as more experienced people who want a hassle free install.

Maybe I should not have started with the distinction between "technical" and "non-technical".

My argument is that even technical users, with the skill to install Debian [1] or far beyond, should not have to figure out this problem. Even if your users can figure out how to solve such problems, they shouldn't have to confront them with a thing to figure out if you don't need to.

The Linux ecosystem is so full of the attitude "oh well, my users can figure it out". It really annoys me, and it really holds stuff back.

[1] This really is not an amazing skill. I know plenty moderately computer literate people capable of getting through the straightforward Debian installer that are not up to date on Linux packaging systems. Just think of all the intermediately skilled Windows users looking to switch.

> The Linux ecosystem is so full of the attitude "oh well, my users can figure it out". It really annoys me, and it really holds stuff back.

Or worse, you get devs with super high egos that will say "Anyone who can't understand a tar file is a freggin' idiot". Note that this is significantly toned down. Ever wondered where Internet troll culture comes from? Linux devs.

Wouldn’t non-technical users have installed Firefox from their GUI, using the distro’s variant?

The UX of deb and of tar are equivalent, relative to a GUI.

hmm. I upgrade my VSCode pretty much all the time they do a release because they provide a .deb file. Installing the upgrade is as simple as double-clicking on the file. I am sure pretty much all the users of Ubuntu and Debian are capable of doing a Firefox upgrade if they really wanted to. But no harm in making it easier.

I have an older friend who used Ubuntu on his laptop, as he had never used Windows and just wanted a simple web browsing capable laptop (before the days of Chromebooks)

The number of times I've discovered umpteen copies of Firefox tarballs on his desktop, and every time he complains it's been a pain to 'upgrade Firefox'

Sure, I could have done more handholding, but yes, Firefox could certainly have made it easier back then and still today.

Also in my experience is always painful to have more than a version of firefox; a firefox-stable.deb and a firefox-nightly.deb would be appreciated

Yup I wasnt saying we shouldnt provide .deb files, I dont know why we dont and sounds like a perfectly reasonable request.

I was just replying to the comment making a point about being accessible to the average user, the average user of Firefox is not manually installing updates on Ubuntu

Hehe, point taken. That's the paretto principle too, spending a bunch of time on a small segment of your users wouldn't make much sense either, especially since most of them are fully capable of working with the system as it is.

Also, since it updates in place anyways, which is nifty, I guess this comment mainly applies to other software I've used which is a royal pain to install...

I think they should at least do some A/B testing on whether including a .deb package increases the downloads or upgrades significantly to justify the time and cost.

I'm just guessing here, but Chrome is not packaged by the distro whilst Firefox is.

Personally I like the tar. That's my preferred way to use Firefox in Debian. I untar on ~/apps/firefox, symlink to my ~/bin and it is ready to go. The auto-update of Firefox works perfect, so I don't need to use Debian packages for that.

Why not just use apt? It's already updated in the repository.

This may require updating your distribution.

and why hell do you keep updated your os ? I do a apt upgrade every week, and everything works fine.

I meant doing dist-upgrade, not apt upgrade.

just use the .snap package (they have all firefox channels as snaps stable, candidate, beta, edge)

channels: stable: 64.0.2-1 2019-01-10 (167) 216MB - candidate: 65.0-2 2019-01-28 (172) 217MB - beta: 66.0b3-1 2019-01-28 (173) 217MB - edge: 63.0b13-1 2018-10-09 (140) 206MB - esr/stable: 60.5.0esr-2 2019-01-28 (171) 213MB -

snap install firefox

Untar anywhere you like and run it from there. I recommend the home directory. You don’t even have to become root. This copy of Firefox should also update itself automatically.

It would be fantastic if this were more like .app bundles on macOS. It’s a special, self-contained folder that GUIs treat as the application binary (double click to launch, etc) but that power users can open as a folder or navigate inside of using a shell.

Then again, I though AppImage was supposed to be something like this?

i hide my Tars in a ~/.opt/ directory, because i dont want to mess with root owned dirs but i also dont want an ugly home directory. as well, i prefer the tars bc the updates are fast and in the background, and get updated faster then packages.

For 95% of users "just leave your Firefox alone you'll get an update in a day or two" is fine.

Dunno whether the Debian/Ubuntu team does anything with the packages, but on Fedora at least, the Firefox RPM usually gets held for a few days so that the Fedora QA team can do their own testing before releasing the package to the public.

Out of curiosity, don't most systems that use .deb files already have Firefox in the package manager? I was under the impression that Debian now has mainline Firefox (instead of Iceweasel) as of a few years ago

Yes, but if you’re not running either the most recent version of a Debian-based OS, you either have to hope there’s a version in the backports repo or potentially break your installation through a diet-upgrade.

Getting the version from the vendor can be more ideal, since there’s no dependency on one party to keep another party’s software up to date, especially where distribution versions are concerned.

You can update easily Firefox from tar file, it has local updates no need for official repository.

Rather than a .deb, provide a PPA.

This would not get automatic updates and would be bad for security.

I recently made the transition back to Firefox too, like a lot of other folks in here it seems. Containers makes working with all my AWS much easier, and the containers isolating Facebook are great.

One problem though, that I can't solve, is that every once in a while I want to log in with Facebook, like for AirBnB. But I can't, because the cookie only lives in the Facebook container, but I can't add AirBnB as a hostname to the Facebook container.

Has anyone solved this?

Edit: I just solved this. You have to disable the built in Facebook Container extension and then it works as expected.

If you use the Multi-Account Containers extension, you can set AirBnB to always open in the "Facebook" container so it'd have access to your Facebook cookies.

I do use that extension, but it specifically blocks you from adding new domains to the Facebook container, which is built into Firefox. I can't figure out how to turn off that built in one and use the multi-account one instead.

> I can't figure out how to turn off that built in one

I don't see the facebook container to be built-in. I have only the multi-account container extension installed, and I can make any website to open by default in the container I have created for facebook.

I think, you will have to uninstall/disable the facebook-container extension, and just have the multi-account container extension enabled.

FB container - https://addons.mozilla.org/en-US/firefox/addon/facebook-cont...

Multi-account container - https://addons.mozilla.org/en-US/firefox/addon/multi-account...

Would you happen to know how safe and separate the containers are?

Like, if I had a specific container used for Banking sites, would it be relatively safe from other containers?

Currently, I use Firefox for casual browsing and Chrome for email and financial related sites. But it would be nice to just use Firefox for both. I'm just worried it would be less safe.

Is there a way to configure Firefox so that a single-use container gets created when certain sites get visited, and discarded when they're closed?

I don't want to delete cookies by default (since I don't like to lose my shopping cart due to accidentally closing a tab in a shop I haven't used before), but I certainly would like to be able to ban sites that use cookies in a user-hostile manner from storing cookies permanently.

It would be even better if Mozilla used the voluntary telemetry + manual user feedback to decide which sites benefit from cookies (e.g. web shops), and applied this to everything else (e.g. news sites).

> I don't want to delete cookies by default

IIUC you want a container to be isolated to a single website, you have to create a container for that website, open the site in that container, and then click the multi-account container tab in the toolbar, and select to option to open this website in that container by default. From then on whenever you open the website, it gets opened in that container by default.

From what I understand, every container has their own cookies and local storage, so it's like using a separate browser for every website.

I just updated and immediately noticed two things that bothered me:

1. It added a new search engine (amazon.com.au) to my list that I'd culled down to just DuckDuckGo

2. It started recommending extensions for websites I visited

Maybe I'm overreacting a bit but why does it feel like every time I update anything recently I then have to spend 5 minutes going through the preferences to make sure it hasn't changed anything. It makes me hesitant to update any app when I feel like the developers are working against my interests.

I was annoyed about the recommending plugins thing as well. It's quick to turn off if that's any help.

No one mentioned WebP support yet! This is great news, Firefox will now feel faster on all the sites that use WebP. Safari is the last major browser to hold out.

What's so great about webp? I haven't heard about it. Is it one of those jpeg competitors? Because I did see a blog post recently that said none of the the things people are coming up with are more than marginally better despite jpeg's age.

I suspect those blog posts focused on photos.

However, WebP works well for a much larger range of image types[0]. It has lossless options with better compression than PNG (because PNG is incredibly simple compared to more recent image formats, in my opinion in a kind of beautiful way). And when compressing graphically simple illustrations with lossy settings, it can get much better results than JPG, both compression and quality wise.

Having said that, I still hope FLIF (or some descendant of it) will get some traction eventually[1].

[0] https://www.andrewmunsell.com/blog/png-vs-webp/

[1] http://flif.info/

In my experience, webps are smaller (25% - 50%) compared to their jpeg counterparts with same or similar levels of quality.

I love Firefox and use it all the time, but for the last few years one of my biggest pains is the OSX performance on my Macbook Pro. Is this close to be resolved soon?

Are you using a non-native display resolution? That seems to be the common factor for people who have significant problems.


A scaled resolution is now default on most MacBooks. I haven't noticed battery problems but certainly seen RAM usage at least double that of Chrome and considerably above Safari.

If you are worried about battery performance, I can personally attest that Firefox is better than chrome in terms of battery performance.

Still Safari is the best in terms of battery endurance (its a VERY SIGNIFICANT difference). But then again its Safari, so you have to decide whether you want a capable browser (FF/Chrome) or a battery efficient one.

IMO Safari is a very capable browser. I switched to it long ago and found I didn't miss anything from Chrome. Experimenting with Firefox now because I love the containers feature, I hope we'll see it in Safari some day.

In my personal experience, other than Apple crippling the extension feature by moving them over to the App Store, Safari is a pretty damn capable browser. It's actually been rather interesting to see drastical improvements in it yearly, even the Mojave build's come with a serious performance boost. Firefox overall feels a lot slower than Safari with some exceptions. The only major killer feature I've noticed is the Devtools on FF.

Safari is my main browser when I am not working with frontend stuff and it's perfectly fine.

What do you mean by performance? I’ve found Firefox to be generally faster than Safari and Chrome on my MacBook Pro. I generally find it to be more energy efficient than Chrome, but Safari is still king there.

This is personal experience, no lab tests, etc.

Regarding the other replies, I believe that Firefox is extremely slow in OS X only if you use a scaled resolution.

That seems to be a known issue, but for some reason it's taking a long time to be solved.

Scaled resolution is default on most/all retina MacBooks which is unfortunate.

For me it is only slow if I go outside the default scaled resolution. On my 2015 13" mbp I use the default scaled resolution to get around the performance issues.

Another work around if you want to use a non-default scaled resolution with firefox is to use firefox in non retina mode. I prefer not to use this as it makes the experience blurry.

I do wish they would fix this issue though.

Oh ok, great. For me the default scaled works best visually and I don’t use an external monitor so glad I won’t have this issue.

With the handoff working too I’m downloading tonight to give FF a fair shake.

I wonder if this is due to not using the native compositor which they are looking at fixing this year.

See "Adding support for native OS compositors to WebRender" at https://pcwalton.github.io/2018/12/07/plans-for-2019.html

Had this issue... until this version! I'm using MBP 15 inch with Firefox 65 (Beta channel), it doesn't use as much CPU as it did before, and it doesn't suck up all battery. This is on scaled resolution of 1920x1200.

It’s strange - I could have sworn that, for a while around where “quantum” was released, OS X performance was really quite good. I ditched it for chrome for a bit, and it even held up with my daily 20+ tabs.

It feels slow because UI is not native. Chrome, Safari, Opera have native UIs on macOS. I mean they are rendered by macOS instead of browser engine. You get 60 fps for every action which doesn't happen with Firefox. Maybe they should use a process just to render the UI.

It's definitely much improved in the last few versions (64/65). Probably still quite power hungry, but it's not making my fans spin up anymore...

Try Firefox Nightly. Runs super smooth on my Mac, and I think WebRender is what solves this for many people.

I feel like I've been waiting a long time for autoplay blocking - does this keep being pushed back or have I missed a feature cancellation/config setting/extension announcement?

We implemented 2 versions ago, however we wanted to do some user testing to make sure it wasnt going to break too many websites and annoy too many users (We ended up changing the behaviour based on those tests).

It will ship and be enabled by default in Firefox 66, it will not have a whitelist of allowed websites but autoplay on all websites will be disabled by default (configurable obviously), if autoplay is blocked there will be a small notification in the url bar to tell you and give you the chance to enable it for that site only if you so wish

As a quick side note, I just want to throw out how grateful I am both as a developer and as a user that Firefox handled autoplay blocking in a sane way.

- An actual notification the user can see and control, which means as a developer I can give people an easy way to re-enable audio if something breaks.

- Actual user controls instead of an opaque algorithm I can't override that randomly decides for me on the fly whether or not I want to hear something played.

- No default whitelist, which means when I turn it off, it actually turns off everywhere, and I don't have to spend a week wondering if the next link I click on is just going to start autoplaying anyway because the site is popular.

It still boggles my mind that Chromium's policy is to just turn autoplay back on if you navigate to a new page on the same domain. It makes the entire feature worthless, it's like they assume the only way users are ever going to visit websites is from a search page or social links.

I don't think Firefox's policy is perfect -- user gestures are interpreted so broadly that they're incredibly easy to abuse[0], and I fully expect websites to start widely abusing them in the future. And the UI for adding exceptions is still really cumbersome -- as a developer I wish there was some way to whitelist domains from the notification bar without going into settings.

But all of this is just quibbles, the implementation is fine. It's just so much more straightforward and logical than what Chromium is doing. I'm at the point where I'm just really happy that Firefox is still rolling out features with predictable behaviors that feel, for lack of a better word, sensible.

I feel like increasingly where features like this are concerned, Firefox is the adult in the room, and for whatever it's worth, I really appreciate y'all being there. Seriously, I hope that everyone on the Firefox team feels good about themselves when they go home at night. I think you're doing good work.

[0]: https://danshumway.com/blog/chrome-autoplay/demo/

Cheers, that very nice to hear

    > And the UI for adding exceptions is still really 
    > cumbersome -- as a developer I wish there was some 
    > way to whitelist domains from the notification bar 
    > without going into settings.
That was implemented last week :) https://bugzilla.mozilla.org/show_bug.cgi?id=1517526 you should see it in nightly, and will be part of the release.

I was hoping to see this pushed out by default because I agree that it's the best 'new' feature I've found in the browser in some time. You can change it in about:config, searching for 'media.autoplay'. I've changed the following:

1) Set 'ask permission' to true

2) Set 'block webaudio' to true

3) Set default to 2, which is 'ask per domain'. As I understand, the default of 0 is 'autoplay', 1 is 'block' and 2 is 'ask' but your favourite search engine of choice can provide clarification on this and the other settings. 2 will cause a pop-up beside the URL bar, like with password save popups, letting you accept or block and with a checkbox to set it permanent.

Why don't they write it as Block, Ask, Allow? Much easier this way.

Apparently, setting both “ask-permission” and “enabled.user-gestures-needed” to true then enables a setting in the privacy area of settings to have a dropdown box with these options. I’m guessing it maps to the 0/1/2 setting I mentioned earlier


Thanks! I also found this page with a bit more info on the settings:


I changed all the media.autoplay settings from the default except block-event.enabled, with default set to 1. With that it seems to do what I want. I'm not sure if the strange media.block-autoplay-until-in-foreground behavior described in the comment is still happening but I set that to false also.

I’m sure pull requests or bug reports are welcome from Mozilla if you feel strongly about it.

Thank you for this, this makes me happy. I always forget to occasionally explore the about:config after a few updates.

Thank you!

Switched to Firefox from Chrome a month ago and my only regret was not being able to control autoplay audio.

This. I switched to FF 20 years ago because of the popup killer. Autoplay is the new popup.

Ok sorry it wasn't 20 years but something like like 15-12 but anyway it was a long time ago folks !

If you want to try it out today you can use Firefox Developer Edition [0]! I use it as my daily-driver and rarely encounter any kind of issue.

This is the documentation for the feature: Allow or block media autoplay in Firefox [1].

[0] https://www.mozilla.org/en-US/firefox/developer/

[1] https://support.mozilla.org/en-US/kb/block-autoplay

I've found this plugin works well. By default nothing autoplays, but I can whitelist websites.


It's still hidden under about:config -- I've set media.autoplay.default to 2 and it'll prompt me if I want allow the playback.

I see it's available on Firefox Mobile (Android) under Settings → Advanced → Media → Allow autoplay.

Is there any news on Wayland support? I had read here previously that Firefox 65 was when support was going to be available by default[0] but there are no references in the release notes.

[0] https://news.ycombinator.com/item?id=18657960

It's opt-in at runtime. https://glandium.org/blog/?p=3899 https://news.ycombinator.com/item?id=18465506

And because that can cause problems when spawning third party applications from Firefox, there's another environment variable to opt-in in last nightly: https://bugzilla.mozilla.org/show_bug.cgi?id=1522780

They finally implemented globalThis [0]! It's not a huge or complex feature, but it's been something I've been wanting for a long time. Now there's a portable way to access the global object. I would've preferred `global`, but that broke web compatibility which made it a non-starter.

I just checked and was a bit sad to find that the latest version of node still doesn't support it. This is a bit surprising as globalThis is already supported in Chrome.

[0] https://github.com/tc39/proposal-global

I would love to switch to Firefox but it still lacks -

* Per-site JavaScript controls (which Chrome has)

* Pinch to zoom (on macOS)

* Customisable keyboard shortcuts for Add-ons (which Chrome has)

* No elastic scroll (bounce effect) (which Chrome has)

I stick with Chrome as it works how I want it to although I find myself using Safari now and then but Safari's extensions are pretty crap in comparison to Chromes.

Customizable keyboard shortcuts for extensions just landed in nightly last week. Expect it soon.

Well god damn! I had no idea. Thank you for replying :)

Smooth pinch to zoom is in development but until it's done I've made an add-on for it https://addons.mozilla.org/en-GB/firefox/addon/multi-touch-z...

>Per-site JavaScript controls

It has, but lacks UI.

But extensions (uBO/uMatrix) do that (and more) way better.

> * Pinch to zoom (on macOS)

And Windows!

I wish Mozilla started to compete with Google on ChromeOS. Just make it the browser and leverage containers to keep things locked down. There’s a total business here and Mozilla could make it happen, without the entanglements of the ads business. Real missed opportunity to help promote open standards at the services level as well.

They had FirefoxOS remember? Doubt think they'll want to do it again soon.

I think the intended market for FirefoxOS was pretty different than ChromeOS. Wasn't FirefoxOS for low-powered phones? ChromeOS is designed around a "streamlined" laptop system.

Yeah, FFOS essentially had the SteamOS problem - targeting a market that didn't have any 'wedge' users in their existing demographics. For Firefox OS and Steam OS to take off you needed to buy new hardware and the people most on board with this already had hardware and didn't have a use case for buying more (low end phone for Firefox, PC-as-a-console for SteamOS). In either case there was a small pool of first movers and then almost no early adopters. This was pretty stark in the steam case as Linux usage, sales and games rose considerably but hardware sales were pretty thin.

Schools. Schools would be pretty motivated to drop Google because Google is an ads company.

Schools love chrome books and gsuite, but struggle with trusting an add company.

If they refactored Firefox OS to be a ChromeBook competitor, sure. They heavily marketed their phones to the developing world markets.

I may be wrong, but I have a suspicion that Chromebooks only got a foot in the door in the education sector because they had such a large marketing machine behind it. I wouldn't mind betting money that Firefox wouldn't be able to get an equivalent foothold (even comparable to their browser market share right now with like-for-like hardware and form factors) just because their exposure for any given initiative is a fraction of what Google can summon.

I'm kind of surprised no one has put something like this together as a hobby project. Just make a custom version of Debian which autoboots Firefox...

Yeah, I've been saying that for years, but I think it's too late for that. Now that ChromeOS supports Android and Linux apps, I don't see how Mozilla could seriously compete against that. Chrome OS is also heavily tied to Google's services, which are quite good.

Ecosystem is everything.

Huge props to the Mozilla crew. Firefox has been an awesome browser and I'm thankful for their product.

My only major gripe is video sort of sucks in it, especially when I'm on my laptop.

I'm glad we're starting to see AV1 rolling out in places. Full support can't come soon enough. Last year at NAB everyone was saying it's about 2 years out from solid support. 2020 should be a good year. AV1 looks pretty promising.

I'm very happy with the wayland performance in this release. XWayland looks really bad with sway on a HiDPI display, and previous to this release, firefox wayland was too slow to be usable.

That's no longer true. I'm giddy with how clear everything looks and how fast it is. This is a huge deal for me.

At the risk of a downvote storm, a few months ago I tried to switch from Chrome to Firefox until I had a call of two hours via Google Hangouts and Firefox started to degrade the voice until a point of no return. I don't know if this is common in very demanding "web apps".

For whatever reason (I hope not nefarious) Google apps perform horribly on FF.

I actually do suspect it's intentional, but have no way to really prove it.

Once Google finishes killing off Hangouts, maybe you can give it another try. Google has no incentive to fix performance issues for competitor browsers in a product that is end-of-life.

I keep an instance of Chrome running just for Google apps and use Firefox was everything else.

I've noticed the same sort of "degradation" when having large PDFs open as well. I have to refresh the "page" (the PDF) to be able to search for something and find that something that I know is in the PDF. Small F5 issue once you know about it cause it remembers where you are in the PDF.

My guess is RAM management/garbage collection going on in the background that needs to whitelist a few things or have some additional fancy stuff going on to not allow for memory leaks but allow for 2 hour phone calls and searchable PDFs after they've been open for some time (and possibly with a 10+ tabs open).

Definitely file a bug. I would imagine both these issues are known about and have a similar root cause.

I love Firefox otherwise.

Blame Google, not Mozilla. They're doing it on purpose and it's the main reason why more people should make the switch now before it's too late and nothing else works with Google's stuff but Chrome:



Google intends to kill Hangouts by the end of the year anyway:


I don't have any proof for this, but I think everything I've seen about Google's Invoice was also meant to kill the open email standards by convincing everyone that the proprietary "AI-enhanced" features of Invoice were worth it over interoperability with other email providers.

Thank goodness that app failed. They're still trying to do it through Gmail now, but it's going to be a much slower process and hopefully people will have enough time to catch on to them before it's too late.

Google is becoming a monopoly in the classical "evil company" sense - a monopoly that's no different than any other monopoly in the past, and that will try to exploit the users and kill competition in the same way others have done it before. More people should start to seriously consider this before going more "all-in" with Google than they already have.

> I don't have any proof for this

That sums it up nicely.

it's common on google apps ;)

Is there any progress with GPU accelerated video decoding / encoding in the Linux version? Now that WebRender is in place it should be possible?

Without it, any WebRTC video conferencing applications are very CPU heavy, which cripples them on laptops especially.

I don't think WebRender has anything to do with accelerated video decoding.

Here's the bug btw: https://bugzilla.mozilla.org/show_bug.cgi?id=1210727

WebRender seems to be a prerequisite for it: https://bugzilla.mozilla.org/show_bug.cgi?id=563206

> The GPU->GPU copy is only usable if you also have hardware accelerated composition.

I thought WebRender is implementing such compositor.

It is, but there was also another one in Firefox (you could enable it on Linux by setting layers.acceleration.force-enabled to true).

I suppose they didn't want to actively work on such features for something that's going to be deprecated. Since WebRender is their way forward, they can as well implement video acceleration together with it.

>Firefox will now warn you when closing a window (regardless of whether you have automatic session restore enabled for restart).

To whoever is responsible for pushing this change through: Thank you.

Yes, it is much appreciated.

Can't tell you how many times my finger has slipped and hit Ctrl+Q instead of Ctrl+W.

Oh well! I didn't need all that work...

Did they remove ctrl + shift + t to reopen a closed tab?

C-q closes the window, not the tab.

No, C-shift-w closes the window. C-q Quits the application, meaning it would close all of your Windows. And if the entire application closes, a hotkey to bring the window back won't work. That's why it's so frustrating when I click it by accident.


There's also a shortcut for re-opening the last closed window(s). Though I don't recall if it works in a Private Browsing window.

So you can't get it back? I usually use Chrome, I never lose my session no matter what I slip up to do.

Well, if I were, say, typing up a forum post, that would be lost. More often, however, I simply lose all my tabs. Which are not impossible to restore, given Firefox's "Restore Previous Session" button. But since I have the Cookie AutoDelete extension[0], all my sites get logged out. So I have to re-authenticate with my password manager, log back into all the sites, half of which require me to pull out my 2FA token...

It only takes about a minute total to get back up and running on everything. It's just a nuisance when it happens.


Thanks for the downvote, whomever. I always consider switching to Firefox, as I think Chrome needs competition, but Firefox has just never been up to par in usability (especially developer usability) whenever I've tried to switch. Things like "losing your work" for no good reason (don't Firefox have "Re-open closed Window"?) is just another thing I would add to the list of reasons not to switch, so feel free to clarify.

To quote from the initial comment:

> (regardless of whether you have automatic session restore enabled for restart).

It also already has an optional check if you want to close a window ("you're closing a window with 10 tabs. sure?"), and a list of recently closed windows to restore.

I interpreted the change as if it would always show the quit warning regardless of any settings, since the warning popup is not exactly a new feature. Some more context helped clarify.

Note that it still won't warn you if you have the "Warn you when closing multiple tabs" options unchecked.

Source: https://bugzilla.mozilla.org/show_bug.cgi?id=502908#c68

> We don't intend to add a third, separate pref/warning for people who do want to have a quit warning, don't turn on session restore, and don't want warnings when closing windows / multiple tabs (ie making browser.tabs.warnOnClose = false + no automatic session restore + browser.warnOnQuit = true show a dialog is not something we're interested in doing.

FWIW I don't get any warning when closing multiple tabs (through tree style tabs), have session restore turned on, and do get a warning when closing the window.

It's because you have session restore enabled. In my case, I don't want session restore or a warning when closing multiple tabs. I only want a warning when closing the browser. But that is not what they fixed here.

Will this halt shutdown as well? If so, I hope this is configurable...

For context: Firefox 64 removed `browser.showQuitWarning`, which would show a dialog asking "cancel"/"save session"/"discard session" when you quit.

Hopefully this is easy to disable

It is, there is an option for it in the Preferences pane right under "Restore previous session".

Chrome's "hold cmd-Q to quit" popup is just about perfect in that regard.

I hate that thing! Is there a way to switch it off? (I get it, it can be useful. It's my personal preference, and showing it is a sane default. Less intrusive than a (modal) pop-up for sure. But I like software which I can set up to my liking.)

It's in the 'Chrome' menu. 'Warn before quitting'.

Ah, thanks. I use firefox as my daily driver, but go to chrome for some web pages anyway.

I'm hearing about this feature for the first time - and I frequent HN and other tech sites!

For pop ups, automatic or it doesn't exist.

I really wish this was standard across all apps on macOS. It's beautiful UX, and easily discoverable.

It's not accessible. Applications that are meant to be used by the general public should avoid making assumptions about how the user presses keys. Specifically some users with motor impairment may find this kind of "long-press UX" to be troublesome.


Applications are open for YC Summer 2019

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