Firefox 82 | Hacker News Reader

Firefox 82

mozilla.org

348 pointsagurk230 comments

    For Windows users, Firefox now uses DirectComposition for hardware decoded video, which will improve CPU and GPU usage during video playback, improving battery life.

neato.

    Websites that use flexbox-based layouts load 20% faster than before.

I haven't noticed this yet but it sounds promising.

The rest of it doesn't really interest me. Picture in picture is a great feature but I don't use it because I just throw videos on a different screen to keep working.

Does that mean any website with a flexbox will be 20% faster? Or that any paint time on a flexbox element will be 20% faster?

i suspect its the later

I'm hoping this update helps with Firefox's performance watching Twitch streams. Twitch kinda chugs no matter what browser I use but it's noticeably worse with Firefox. From the googling I've done, I'm not alone in this. I considered switching back to Chrome but didn't feel like moving all my stuff over, so hopefully now I won't have to.

The extension Alternate Player for TwitchTV[0] I've found is pretty good performance wise. It adds an extra second or two of latency I've noticed when watching the same thing as friends, but otherwise it's pretty great.

0:https://addons.mozilla.org/en-GB/firefox/addon/twitch_5/?utm...

The CPU and GPU usage during video playback is pretty big for me, I noticed that Firefox could get pretty intense on resource usage when playing longer videos. Not sure how much impact it will have on battery life but fingers crossed.

The two things that would get me to consider switching back to Firefox:

1. It not longer forces to restart my entire browser for an update, where every new tab I open says "restart before you can use your browser." I am not going to use a browser that closes my various windows full of API documentation once a week. I know I can disable auto-updates, but that is also not what I want: it should update when it is finally closed (like Chrome), but not update and then break my browser until I close it.

2. Individual threads can crash without taking the browser down. This feature, individually, is why most people switched and did not look back. As with the problem above. It has been requested since 2008 (http://forums.mozillazine.org/viewtopic.php?f=8&t=890085), and again in 2015 (https://support.mozilla.org/bm/questions/1092836).

Really, it's the same thing: closing my entire browser solution is an inhibitive inconvenience in my modern workflow, and Chrome never forces this action. Instead, it does everything it can to keep my tabs alive and usable. I will reconsider Firefox when they integrate this (IMO very obvious and critical) requirement. I don't care if it's faster, or uses less battery---if it regularly forces me to close all my tabs, it's unusable.

On 1: Your tabs don't re-open after an update? You literally click a button, the browser closes and re-opens those same tabs.

I have 60 tabs. Re-opening it is a hassle all on its own.

Have you really tried? 60 tabs is peanuts for current Firefox.

> It not longer forces to restart my entire browser for an update, where every new tab I open says "restart before you can use your browser."

Are you on Linux? My understanding is that the distro packages exacerbate this problem, and it is less of an issue if you use the builds direct from Mozilla.

Like dblohm7 said, the issue here is when something besides the currently running instance of Firefox does an update. This can happen with the Linux distro packages or if you open Firefox with multiple profiles at the same time. I know it is a big pain, but hopefully this at least helps you understand what is going on.

If you choose "restore previous session" under general preferences, then you can at least not lose the set of open tabs when you restart.

As for 2, Firefox already does this, and has for about 4 years. Since those comments in 2008 and 2015, Firefox has shipped the Electrolysis project, which makes Firefox multi process. The UI (mostly) runs in a different process, which massively improved UI responsiveness, because janky websites can't interfere with clicking on stuff in the UI. One web site crashing can still cause another web site to crash, if it happens to be in the same process (out of 8) but we're hard at work on the Fission project which will move individual website domains into their own process. Though you can still end up with one instance of Google Docs crashing and taking down another one.

> I know it is a big pain, but hopefully this at least helps you understand what is going on.

Frankly, and don't take this wrong way, I don't care. I'm working on other problems, and don't want to worry about what my browser is doing. I just want it to work.

> If you choose "restore previous session" under general preferences, then you can at least not lose the set of open tabs when you restart.

Have you ever tried restarting it with 60 tabs open? It doesn't work on my machine.

> As for 2, Firefox already does this, and has for about 4 years. Since those comments in 2008 and 2015, Firefox has shipped the Electrolysis project, which makes Firefox multi process. The UI (mostly) runs in a different process, which massively improved UI responsiveness, because janky websites can't interfere with clicking on stuff in the UI. One web site crashing can still cause another web site to crash, if it happens to be in the same process (out of 8) but we're hard at work on the Fission project which will move individual website domains into their own process. Though you can still end up with one instance of Google Docs crashing and taking down another one.

I don't care if it's multithreaded to separate rendering from management for performance purposes. What I care about is keeping my remaining tabs when one crashes.

> we're hard at work on the Fission project which will move individual website domains into their own process

When this occurs, I will once again consider switching back. Though this means if one API tab breaks, they all do---it's still a massive problem.

IME, "restore previous session" works for my 200 tabs and single tab hang won't cause entire browser hang. Restore feature is better than Chrome because it's lazy load rather than immediate.

Even with 4000 tabs open, Firefox restarts pretty fast.

I'm sort of, but not really, surprised so many smart techie people are so adamant about using Chrome when we're well on our way to Blink being the new IE6.

I don’t think people complained that IE6 was the only option. People complained because it was buggy and didn’t follow standards so that, compounded with monopoly, meant that apps only worked in IE6.

Sounds like the current Firefox situation only because 1. Google deliberately ignores Firefox support and 2. Firefox does not support some standard APIs that even Safari supports.

The former is only Google’s bad politics while the latter is just Firefox playing catch up. Neither attributable to the monopoly.

Even if I don’t use Chrome, it’s currently the best browser by far from a user’s standpoint. This was not the case for IE6.

What browser was better than IE6? From what I remember none.

It's kinda like chrome all over. IE6 was faster than other browsers and supported the APIs site developers were using. From the user's standpoint it was the best browser by far.

IE7 came out with tabs in late 2006 while Firefox had it at least since 2002. Imagine having no tabs in a browser.

IE6’ advantage was the ActiveX features, which were not documented nor a standard, unlike Chrome’s features.

Let's not forget that a BIG IE6 issue was that people used IE6 for YEARS. It was tied to the OS.

A big feature of how chrome, firefox, and edge work today is that everyone updates to the latest version usually within days of their releases.

IE6 was still in use when IE7, IE8, and IE9 had been deployed for years with better rendering, CSS, and javascript support. Primarily because a lot of websites and internal applications had stupid things like "activex" plugins required for proper usage.

To say that chrome is the new IE6 really misses out on how horrible the internet was during the heyday of IE6. It's fairly hard to come up with a website that works well in Chrome and not so well in Firefox. Unless you are pulling out brand new CSS or JS features, you are very likely to get relatively consistent rendering between the two.

> Firefox does not support some standard APIs that even Safari supports.

Such as?

> I don’t think people complained that IE6 was the only option.

Some people did. I complained -- at the time -- that the monoculture was harmful, and I used Firefox as my primary browser for that reason (but there were lots of sites that would only work on IE). I loved Chrome when it first showed up: a new entrant increasing the diversity and bring lots of new technical and UI innovations. But once Chrome reached the point where it dominated the browser ecosystem I returned to Firefox because I really DO think that diversity is important.

Specifically, I believe that when the largest player in the market has no more than ~40-50% market share AND there are at least 4-5 different widely-used tools, that producers of content (as well as the tool providers) will feel compelled to focus on delivering a compatible experience. But when the largest player in the market has 75% of all users, it's advantageous for content producers to focus on delivering for that platform AND it starts to be advantageous for the producer of that tool to become incompatible with other tools.

To keep the ecosystem healthy, I want to see more diversity than we have today.

Exactly. Thank you.

You missed: 3. Cases where Google doesn't follow standards.

There are plenty. People are just coding to Chrome instead of standards, just like they coded to IE instead of standards.

> This was not the case for IE6.

It sure was. For years. IE6 was a _much_ better browser than the contemporary Netscape versions, for example.

> Cases where Google doesn't follow standards.

Like what? If there’s a standard, Chrome follows it. If there’s no standard, they propose one. They used to have a lot of custom features with prefixes, but that rarely happens anymore.

> If there’s a standard, Chrome follows it.

That happens to not be true all the time. Some examples off the top of my head:

* https://bugs.chromium.org/p/chromium/issues/detail?id=102719... points to various places where Chrome's security checks don't match the spec and therefore can allow things that other browsers and the spec don't.

* https://bugs.chromium.org/p/chromium/issues/detail?id=878491 -- non-spec implemetation of document.close (shared with WebKit)

* https://bugs.chromium.org/p/chromium/issues/detail?id=100027... -- broken .complete attribute behavior, broken since forking from WebKit.

Those were all found due to sites coding to Chrome instead of the standard and therefore not working correctly in Firefox.

If you want something more long-standing, https://bugs.chromium.org/p/chromium/issues/detail?id=286670 is a clear bug in selector-matching optimization that leads to incorrect behavior and just turned 7 years old.

Or you could look at https://bugs.chromium.org/p/chromium/issues/detail?id=340414

Or https://bugs.chromium.org/p/chromium/issues/detail?id=352144

There have also been multiple cases in which everyone just gave up, and changed the standard and all other browsers to match Chrome's behavior in cases where it was causing serious compat problems for everyone else.

It lacks all the ... quirks ... that made IE6 such a joy.

Safari on iOS is already the new IE6

As a web developer who suffered through IE6's pain-points from its heyday until 2014, I've said this about Safari on iOS for years now.

I had a huge retail client who's customer base was largely on Win 2000/XP equipment, so they insisted on it until usage dropped off enough. Sorry I can't be more specific.

Luckily, it's not actually quite that bad yet, but I've built PWAs and some of the bugs are truly awful.

I think it was iOS 11 where the camera didn't even work in PWAs for months- we had to get people to use the website through the actual Safari browser on their phone. It wasn't actually fixed until 11.3. It was brutal.

Those who say that didn't work with IE6...

Chrome is a good browser. When it works better than Firefox, that's because it really works better, not because it has a weird nonstandard quirk that everyone designed for because it is the dominant browser. Chrome/Blink has some quirks but it is nothing compared to IE6.

The problem with Chrome/Blink is different: is is essentially a way for Google to push whatever standard they want. Emphasis on "standard". Microsoft wanted to control the web on the client side. They made IE, and therefore Windows a near-requirement to access the web, the whole embrace, extend and extinguish thing. What Google want is the opposite, they want every machine to use Google services, and the meaning of Chrome is "here is how a web browser should be, please copy me".

Unlike Microsoft with IE, Google wants Firefox to implement every feature Chrome supports, and for that reason, these new features are much less quirky. The problem is: what Google wants Firefox to implement may not be in the user best interest.

Excuse me but I've been burnt by IE6 and have the scars to prove it.

Monopolies and monocultures are bad news anywhere, like in my city of Winnipeg where we have one of the largest elm canopies in the world which is expected to disappear in the next 10 years due to disease and infection.

Yes, during IE6's heyday, we designed for its quirks. When we moved on, we built standards-based code and then broke it to run in the lingering IE6, 7, 8...

I'm not really sure exactly what you're getting it. None of what you say is false, but if Google gets a certain share of the market, they won't care at all if Firefox goes along with it or not.

That's what I'm afraid of.

Happy Firefox user. I love the multi-account containers and the picture-in-picture. That, and you aren't "logged-in" at the browser level.

Then I think you'll also love the Container Default Pages extension [0]. I lets you assign a default website for each container. So for example if you have a container just for YouTube, you can set YouTube to open automatically when you open a new tab in that container. It's very convenient. Disclosure: I created it.

[0] https://addons.mozilla.org/en-US/firefox/addon/container-def...

The extension already supports that, albeit in a semi-convoluted way: you have to visit the site to set a default container for it. If yours let's you edit it by hand, that is a pretty big quality of life improvement.

I think what you're referring to is the "Always Open This Site in .." option. That one assigns a website to automatically open in container. This on the other hand, assign a container to a default page. It's similar to how you have home page for your whole browser, but now for each container.

Ah. I understand. Thanks for clarifying.

I happily switched to Firefox Quantum when it first came out, but have periodically switched back to Chrome because Chrome feels faster and snappier. Not sure why that might be the case. I'm using the latest MacOS X with the same set of extensions/add-ons in both browsers.

Best viewed in chrome isn't pasted on most sites, but might as well be.

Sadly, it is. The IE7 era, perpetrated by Google.

When I saw a dev I mentored put that on a site I realized I had failed in my attempts to fully explain the dark days of the browser wars. Chrome truly is IEv2.

OVH client interface goes into a login loop on anything other than Chrome. There are a few sites like this which are the only reason I have Chrome installed.

I've been able to solve that on a few sites by turning off the "enhanced tracking protection" next to the address bar (haven't tried OVH specifically).

For OVH I tried a lot of things, disabling extensions (one by one, and all of them), using a blank (firefox) profile, the usual cache/cookie invalidation, disabling various privacy protection enhancements (one by one and all of them), etc. Chrome worked.

For other sites yes, often just cache/cookie invalidation is enough.

Exactly. Chrome just feels faster overall.

Chrome is so popular that developers optimize for it. So for example, even though Firefox has a much faster CSS engine than Chrome, most developers don't use the kind of heavy CSS shenanigans that would make Chrome seem any slower than Firefox because if they did the site would be slower to the majority of users.

It's a chicken-egg problem, basically.

I really doubt that most developers deliberately optimise for Chrome, or do you mean subconsciously? I'm a web developer and although most others I know use Chrome (I primarily use Firefox even though I'm not entirely satisfied with it) I've never heard of anyone deliberately optimising specifically for performance in Chrome.

It's not intentional, but The Market doesn't care if it is: They optimize for Chrome by virtue of optimizing in their choice of browser, usually Chrome. High usage of Chrome causes Chrome to be even more preferable.

Devs optimize for Chrome by testing on Chrome. I'm starting to see sites that don't even work on Firefox.

The worst is when sites serve a crap version of their page to any non-chrome useragent. Many sites do this, Google being the worst offender (nearly every Google app)

I fake my user agent to recent Chrome to avoid these shenanigans. But you will get Capchas all the time because sites think you're a bot

FF performance is very poor compared to Safari and Chrome (on Mac).

Have you enabled WebRender? It's not yet enabled by default on Mac and it does improve performance significantly.

Pulling out a tab into a new window is really snappier in chrome than in firefox. Every time I do this in firefox, I ask myself why this is such a burden.

Browsers feeling snappier is sort of like cell phones being thinner, at least for me. There is a threshold below which it just doesn't matter to me, and browsers and phones have been there for a few years.

Nothing beats Safari on the Mac, but I’m rooting for Firefox for diversity and innovation on the devtools.

Anecdotally I find Safari > Chrome > Firefox on my 2015 MBP. I've been using Firefox, despite its comparatively poor performance, for years now mainly to try to support some competition to Chrome. Admittedly I'm on my way to switching to something else though. The Mozilla Foundation is a bloated, politically preachy catastrophe which clearly does not prioritize Firefox highly anymore and so it feels like the whole thing is, in essence, on borrowed time at this point.

> Nothing beats Safari on the Mac

As a Safari user on a MacBook Pro, I disagree. Safari sucks ass. I use it only because it offers the best battery longevity, but that's it.

The devtools are horrendous, the text selection is extremely bad (example: https://i.imgur.com/qC7amU1.png), extension support is awful (I only use AdGuard and LastPass and both are bad, I needed to install apps just to use them, and they frequently don't autostart).

It's also the only remaining browser that sometimes has plain wrong rendering of elements.

Sort of true. Switched to Bitwarden from Lastpass because of poor support for Safari by Lastpass.

I'd have phrased it as poor support for LastPass (and every extension) by Safari.

What's the extension/addon/plugin situation? The last time I tried, I was annoyed with it and moved back to Firefox.

It will supposedly get better with Safari 14 thanks to the support for Web extensions, but I haven't seen any improvement yet

Except when you want a browser that actually does what you tell it to. For example (something that I ran into the this morning): try taking screenshots of a video from a streaming service (in this case Hulu). I don't need a lecture of why Apple is broken here; two screenshots is 100% fair use, and the legal ramifications are on me to manage as a responsible adult.

- Hulu app on iPad: black rectangle

- Hulu site on Safari on macOS: black rectangle

- Hulu site on Firefox on macOS: works

Now, I prefer Apple's products because they've been historically been more accessible / high performance. But Apple's recent user-hostile behavior is slowly pushing me to switch.

>>DevTools now shows server side events in the Network panel. This allows a server to send new data to a web page at any time allowing developers to see events they previously couldn't and help with lower-level troubleshooting.

Would someone please shed more light on this? I work on BMC’s Remedy ecosystem and server side logging something that needs to be turned on separately.

Does FF 82 help me capture it just like that or I need to do some configuration changes on the server?

Pardon my simplistic question.

That looks to be a typo in the release notes. "Server sent events" is a specific server-> client HTTP communication mechanism.

The Network DevTools already showed standard AJAX requests, and support for showing messages sent over websockets was recently added.

This adds support for another mechanism that was not previously displayed:

https://developer.mozilla.org/en-US/docs/Web/API/Server-sent...

That's good news. That was one of the things for which I had to switch to Chrome.

I too was interested by this. If you follow the links it's all documented:

https://developer.mozilla.org/en-US/docs/Web/API/Server-sent...

> Traditionally, a web page has to send a request to the server to receive new data; that is, the page requests data from the server. With server-sent events, it's possible for a server to send new data to a web page at any time, by pushing messages to the web page. These incoming messages can be treated as Events + data inside the web page.

Push-based server->client messaging. Exciting stuff.

How is this different from SSE or WebSockets?

Isn't SSE the abbreviation for Server Sent Events?

This is SSE. Here's a good answer on SO comparing the two: https://stackoverflow.com/a/5326159/1714997

Updated Firefox and the first thing I'm greeted with is an advert for their VPN. Not too happy about that.

Would you pay a small fee, say 20 Euro, a year for a Firefox version which guarantees no-ads or pocket? If enough people feel like that they might consider it.

Meanwhile something has got to pay for the bills. I don't think advertising for their own VPN service is too bad.

I would have but since they fired the devtools and mdn team, it's only a matter of time until firefox becomes totally irrelevant.

Can't have a dev team using up wages when that money could go on bonuses to an already hugely overpaid executive now can we!

> Would you pay a small fee, say 20 Euro, a year for a Firefox version which guarantees no-ads or pocket?

i would have, even $50. but not after they've laid off most of their servo engineers to focus the company on more idiotic moneygrabs.

there was never a way to pay for firefox. only some black hole foundation contrib that probably went to their many failed, poorly-conceived initiatives.

Those "money grabs", by definition, make more money than they consume.

Money grabs by definition never fail? No, money grabs aspire to make money but may not succeed at it.

I'd honestly happily pay for an FF license (or support license or something) if it was guaranteed the money was going directly to FF development instead of other Mozilla initiatives.

I'd happily pay at least £5 a month for Firefox, if I could.

If I were richer, I'd pay more, and I totally understand the millions of people who don't want to pay a penny.

From having read lots of posts over the years - here and elsewhere - I get the sense it would at least be worth their while opening up some kind of donate/subscription function, for people who feel they want to help.

Why? It's 'their' product and fits within Mozilla's ambitions for the product, namely making the user central to the online privacy debate.

If I paid for Firefox, I'd perhaps be a little irked, but given I've never given Firefox a penny, I'm happy for them to try and get me to spend money on their things (even though as a Euro, I can't yet - I think the VPN's US only?).

If anyone from Firefox is reading this - please hurry up and allow me to pay for your product. I'd dearly like to!

- ed: typos

> Why?

Because people want to use their browser to browse. They don't want to be advertised to while they're trying to work or otherwise get on with their hectic lives.

They also don't want to hear about silly TV shows while they're browsing.

All things Mozilla needs to learn.

But a VPN is arguably an help for a better browsing, so it's in line with what the user wants already.

> But a VPN is arguably an help for a better browsing, so it's in line with what the user wants already.

Lol and Coca-Cola is delicious and refreshing so it's arguably in line to put it on the side of a bus when it goes past people who are thirsty.

I don't want your product shoved in my face, no matter how awesome you think I should find it.

Ok, I've updated to see for myself. The 'advert' is in the start up page we see upon every update. It's as much of an imposition upon my browser experience as a welcome page in any application, and because my hand is already hovering over Ctrl+W, it's gone before I even got a chance to care.

I just don't get the irritation here.

My web browser is my single-most used application and I've never paid Firefox a bean for it. If they want to promote an extension to their service that they presumably feel is relevant to a percentage of their userbase, then they should go for it - even if the likes of you or I don't really care for it.

The Mr. Robot promotion was a screw up, no denying it.

> They also don't want to hear about silly TV shows while they're browsing.

Considering that was nearly 3 years ago, I think they learned.

Yet here they are again... pushing more bullshit that interests them as highly-paid tech elites in Silicon Valley and assuming it'll interest me and my grandmother as well.

Mozilla: not everyone watches the same TV shows as you... not everyone wants a VPN like you. Is it a diversity problem? Do they need to hire people outside their bubble?

> Why?

Because ads suck.

It isn’t their product at all. It’s just a white label service they’re reselling - at significant markup.

I pay $4.99 a month. It works fine, and if that is a significant markup, I'll never notice it. Indeed I am glad to support Firefox.

The partner they currently use to do this is actually €5(~$5.91)/mo as someone else mentioned so there's no markup but actually a cut in price if any.

Hence my using single quotes around 'their'.

Just to clarify here, the Mozilla VPN is $4.99/mo, whereas Mullvad (the partner it currently uses) is €5(~$5.91)/mo.

There is no markup, it's actually about 15% cheaper.

> I'm happy for them to try and get me to spend money on their things (even though as a Euro, I can't yet - I think the VPN's US only?).

That's one of the ridiculous parts of this: They're just reselling Swedish VPN provider Mullvad's [1] product. That product (which also supports Wireguard, yay!) is available worldwide for €5/month. You can even pay by "anonymized identifier string and cash in an envelope" if you are so inclined.

Come on, Mozilla – just keep on making the world's best browser (and the world's last free browser). Stop paying your CEO millions. Stop making bookmarking services. Stop reselling VPN services. Just act as a shovel for money to go into the pockets of talented browser developers!

[1] https://mullvad.net/en/

Ah, thanks for making me dig around - apparently Mozilla's offering does cover a few more countries than just the States, as it did prior - mine own included.

So, I can finally give Firefox 50p a month! (Or whatever their share of the charge is - not a heck of a lot, I imagine).

> Stop reselling VPN services.

People hate Mozilla for taking money from Google. People also seem to hate Mozilla for trying to earn money from somebody other than Google.

How do you recommend they pay for browser+engine development?

> How do you recommend they pay for browser+engine development?

Current funding levels would go a lot further if the organization cut everything that isn't a mechanism to support the browser developers. Why does Mozilla need a CEO making millions? Why do they need a bay area HQ?

Said differently: I have no problem with how Mozilla raises money (although I do have some doubts about how much they'll make by reselling a €5 VPN service for $5, even taking into account discounts), I have a problem with how Mozilla spends money.

> I have no problem with how Mozilla raises money

So you just need this sentence: "Come on, Mozilla – just keep on making the world's best browser (and the world's last free browser). Stop paying your CEO millions."

You don't need: "Stop making bookmarking services. Stop reselling VPN services."

I would be fine with the notice if it supported the machine that I am on. Come on Mozilla, you know what OS I'm running, don't advertise "coming soon" to me.

It's not a great default, but you can disable these by clicking the gear icon in the top-right of the new tab page and turning off "Snippets".

After Google used browsers to gain monopoly and cutting off their income, there's only so many monetization strategies you can do to stay somewhat competitive with an advertising company providing browser for free while simultaneously trying to block you from using any other browser.

Won't argue that Mozilla/Firefox has been comprehensively out-manouvered by Google/Chrome.

I'm still puzzled why Moz haven't tried a paid-for Firefox option. They've gone to the bother of creating entire new paid-for products - with all the budget and product/market fit uncertainty that involves. Surely a paid-for FF option would have been a less expensive experiment?

Perhaps the terms of the Google "donation" prevent/constrain that. I wouldn't be surprised if so. Which does lead to an interesting - if probably fanciful - scenario from the anti-trust investigation announced today...

...charactertise the browser as a "utility" and require G to (a) transfer it into Mozilla's stewardship and (b) force them to license from there.

(I know: unrealistic, unworkable, lots of other "just no"s. But a fun idea nonetheless).

>Won't argue that Mozilla/Firefox has been comprehensively out-manouvered by Google/Chrome.

And sometimes without being at fault really. I remember a key moment when firefox was bleeding users was when google decided that H264 would be the standard after switching for a while and mozilla having to run after it trying to get implementations going and then still having a laggy af experience on youtube and the like because the implementation google went with was proprietary.

>I'm still puzzled why Moz haven't tried a paid-for Firefox option.

and what functionality would the paid version of firefox have? Chances are it would cause an outrage and user-loss due to cries that functionality would be premium only tbh. It would also probably make not much of a dent in the budget.

> and what functionality would the paid version of firefox have?

Maybe it's what it _doesn't_ have. No ads for Pocket/VPN/whatever. Maybe it comes with DDG as the default search engine. Or maybe it's pre-configured with a set of tested privacy-focused extensions.

Pro-privacy advocates might be in the minority, but it's a growing and generally vociferous minority. Mozilla's VPN offering suggests they see privacy as a market. FF is still by far their most popular offering; so I'm simply wondering why there's been no effort I'm aware of to explore the intersection of FF, privacy and revenue.

Recognising Google would most likey challenge that.

Just wait until you go to google and see an advert for chrome!

Might be waiting a while. I tend not to go to Google.

God could grant some of you immortality and you'd complain about not being able to instantly transport anywhere.

Firefox is a decent product that you get for FREE, the output of all these groups' work is free and reproducible, but you don't like a once-in-six-weeks underwriting message when a new version comes out.

If someone is so perturbed by something so trivial, it speaks more about the person than the product. The FF team could probably do away with that page and the user will still find something to complain about.

> Restoring a session is 17% quicker, meaning you can more quickly pick up where you left off

This is awesome as I need to use the Total Suspender add-on to keep obnoxious web apps (like anything Google) from cranking up my CPU utilization when I am not looking at them. I am seeing a noticeable improvement.

Several times over the past five years I've tried to switch from Chrome to Firefox and found it to be too slow once all add-ons were included. For a few months I've been using Firefox full-time and I'm finally switched for good. I encourage others to give it a shot.

Same! Containers and Pop-Out video are the killer features too IMHO

yeah, as long are containers aren't synced and available on android it's not really an option for me, it's nice I can isolate certain sites, but if it doesn't work everywhere it's a bit pointless for me.

So having containers and pop out video available nowhere is preferable to them being available on desktop but not mobile?

Progress on Containers is visible here: https://github.com/orgs/mozilla-mobile/projects/51#card-4487...

I just used the pop-out video feature for the first time, after thinking it was going to be useless. It is really awesome and the ability to put the video anywhere, not just in the browser, was one of the first things in a while that made me say wow to a new feature in a web browser.

I love using it as I WFH and use an RDP session to my work computer, open PIP on my local machine and have it hover over my RDP session. Very nice!

Containers are awesome. Still waiting, though, for when I can set bookmarks to use a specific container. I have a few different gmail accounts and it would make accessing them a lot easier (having containers at all is already a huge improvement).

I use pinned tabs for a similar use case (o365). Open the page in the container you want, right -click -> pin tab and it will remain on the left of the tab bar with just the icon remaining.

Yup, that's exactly what I do right now as well.

Click ••• > Always open this in a Container > Pick one. Done.

As I mentioned, it's for multiple different gmail accounts. Same url, different containers.

you might be able to trick around that if you can set up something that redirects. I have multiple bookmarks that go via the same URL via a redirect (it's an SSO login flow), but if I set the initial URL to be tied to a specific container, the entire flow remains in that container. So if you could set gmail.example.org to to redirect to gmail and gmail2.example.org as well, you could tie those urls to a container and use bookmarks. It's a bit roundabout, but it probably works.

No still laggy for me, I only use it for some alternate accounts. Just feels jittery compared to chrome specially when you are using the element inspector.

Same. Firefox gets bogged down debugging one of our more complex websites. Was quite surprised when chrome didn't. Figured all the speghetti meant it was doomed either way.

Yeah, I've found that Firefox is fine (even better than Chrome) on simple websites. But it gets bogged down on complex sites like Facebook or Gmail.

Firefox gets bogged down in Gmail by some sort of tracking pixel or something, and it's noticeably improved by content blocking.

I have the complete opposite experience. Firefox runs smoothly even when I have 150+ tabs open in 4-5 windows. Meanwhile running 20+ tabs in Chromium bogs down my system so hard that the mouse can start lagging.

To be fair it runs very nicely when not excessively tabbing everything, but that's just how I prefer to use the web.

still laggy for me too but the lag factor is now lower than the google-annoys-me factor

I'd consider that if it wasn't slowly eating all the available RAM on macOS.

It doesn’t. Disable your extensions and re-enable them one by one until you find the one which is leaking.

or check about:memory and about:performance

I find that that is usually due to certain sites that use massive amounts of poorly written scripts. You can see it in the Firefox memory report. The new Facebook UI is a good example. If you kill that content process it will free the memory and reload the page.

I'm using it, but I need to restart it several times a week because it stops playing video. For comparison, I only restart Opera/Chrome for updates once in a few weeks.

And I still hate the "many tab handling" in ff where it makes me scroll.

There is the “⌵” icon on the right side of the tabs that pulls down a list of open tabs.

I find this more useful than the ever shrinking tabs on Chrome. I have a hard time find a tab in Chrome once the number of tabs gets large.

Compared to Chrome where you just get a row of indistinguishable minute tabs with no visible text? You can configure the min/max tab width in Firefox's preferences to control when it will start scrolling.

It will still start scrolling though. I'm happy with just the favicons, the one extra letter or two don't help really. Never had problem with 100+ tabs when I was using Opera. With firefox it was always a fight to set it up.

Same here, My switch happened over the weekend. Still looking for a convenient way to have profiles that are like chromes user button, but overall I'm happy to escape Google's grip.

I'm assuming you've seen https://addons.mozilla.org/en-GB/firefox/addon/multi-account... ?

It doesn't work in exactly the same way - you don't by default get a different window per-profile. Instead, tabs get a different colour for each.

That Multi-Account Containers extension is awesome. My primary use case for Firefox container tabs is that I have separate Google accounts for personal and work stuff. With Multi-Account Containers, I can have Gmail and YouTube in Personal tabs and Google Calendar and Docs in Work tabs.

You can also use Firefox's about:profiles page can to open and manage different profiles in different windows in one running instance of Firefox.

One of the benefits of Chrome's user profiles this misses is the ability to have separate bookmarks for different profiles (Personal & Work for example)

This is the killer feature I can never drop now that I've started using it.

No, I hadn't - thank you (bow). Chrome is now buried in the backyard. (That's legal right?)

Firefox is unusable on Ubuntu.

They have a years old bug they can't / won't fix that causes stock Firefox to be at 100% CPU constantly, regardless of what you are running.

I finally switched to Brave and amazingly, it manages to run basic websites at minimal CPU. And it isn't like I am running some rinky-dink computer. I have 32GB of RAM and a Ryzen 9 3900x.

EDIT: Single tab of the front page of reddit, no add-ons, nothing loading, Firefox goes between 7% and 21% CPU. Meanwhile I have 20 tabs open in Brave and it is sipping between 0.7% - 3%.

Is it something specific to Ubuntu? I use it on Xubuntu all day, every day, and never see it spike or have problems of any kind. I have a potato.

After looking more, after being prompted for the bug, it may be an issue with Firefox getting stuck in a loop trying to load fonts.

I may work on fixing this locally, but I may just stick with Brave since it works just fine.

https://bugzilla.mozilla.org/show_bug.cgi?id=1495900

I use Firefox on Ubuntu and do not have this issue - is there something more specific that triggers this behavior? Have a link to the bug tracker?

No clue to be honest, I just opened it on a single tab of Reddit and it was using high single to double digits of CPU sitting there.

I looked at the bug again with fresh eyes and it seems like there is a deeper issue that may be my fault.

I am running i3-wm and it seems like Firefox gets stuck in a loop if you don't have all the fonts it expects. I had a previous issue where I was crashing my terminal when a program tried to show emojis, which was a legitimately a fun issue to figure out.

https://bugzilla.mozilla.org/show_bug.cgi?id=1495900

https://bugzilla.mozilla.org/show_bug.cgi?id=1499385 (not sure if this was the one I saw before giving up + installing Brave, but the fact there are _tons_ of "Firefox hitting 100% cpu" is telling)

Aha, that is something to watch out for, and may explain what I've seen on some websites!

I'm using it on Focal and it's fine. Can you link to that bug? Maybe something related to hardware acceleration? I had trouble with that in the past year.

I use Firefox 78 on Ubuntu 19.10 and do not see 100% CPU usage at all. In fact, with over 1000+ tabs all 8 cores are just over 10% (of course there are other processes running as well).

Is there a bug with more details on this?

Curious about this too. I use it on a couple of 18.04 and 20.04 *buntu's, and only occasionally have trouble with it.

I have noticed there are some websites that tend to make it both consume CPU and prevent my machine from screen-blanking, but I find them easily enough with the the diagnostics tool in ff. I don't recall having that problem with Chromium, but I feel like I've given it up for good now.

> Single tab of the front page of reddit

Not surprising, considering the (new) reddit website has the most bloated, sluggish UI I've ever used. And the only change they made recently made it worse as it added autoplaying livestreams.

Yeah, I may have been purposefully putting my finger on the scale a bit.

google.com was bouncing around 1 and 3%. Firefox becomes unusable for me to do webdev on with multiple tabs if each tab is doing 5%+ each.

You've obviously never used Instacart.

Old Reddit Redirect add on is a requirement for me to browse reddit these days. I imagine they'll turn it off some day and hopefully they will Digg their own grave ;)

Sorry but that's clearly something on your end, and you know this.

When a comment like this pop up, I launch this Chromium instance I have installed but never used, and it takes several seconds to launch (from an SSD, so that's no the limiting factor), while Firefox opens almost instantly. I think some operating systems (at least Windows) pre-cache frequently used applications and their data so that they start as quickly as possible. Maybe that's what you notice everytime you try to switch?

I bet you're right. For me on Linux with Firefox as my main driver:

Chromium startup to empty page: 7 seconds. Firefox startup to page populated with tabs: 3 seconds.

It's worse though - since I don't often reboot and I have Firefox running almost all the time the effective start up time for Firefox for me might as well be zero.

Perhaps oddly, I've never perceived Firefox as slow - but I'd be the first to admit that I use it only for rather mundane purposes, so maybe there are use cases where it would be deficient.

I find this too, both on Linux and MacOS Chrome takes longer to load, longer to load pages, and eats huge amounts of CPU and RAM. By contrast (again, my experience), Firefox opens instantly for me and loads pages very fast even with about five addons installed.

If this is on Linux, you probably have the snap or flatpak version of chromium installed.

Every day for the last 10+ years I've run both Firefox and Chrome on each of my screens. I have no idea what you're talking about in regards to a speed difference. The only reason I use Chrome is because it's the the new IE and I need to see and experience what my clients are.

You've got that right. We're quickly heading back to the days of "This website best experienced in IE6." (replace IE6 with Google Chrome). It's not a good thing.

> found it to be too slow once all add-ons were included

That sounds like it's likely to be a performance issue within one or more of the add-ons. Which add-ons are you using? There may be better alternatives that don't cause performance issues.

If Firefox is fast enough for you now with your add-ons, it would be even faster with a more optimal set of add-ons.

It certainly wasn't add-ons. Firefox has a long history of performance issues.

EDIT: I see the firefox brigade is out in full force.

The comment I was responding to specifically said "found it to be too slow once all add-ons were included", so I don't think it's at all unreasonable to talk about the add-ons. That doesn't mean I'm dismissing the possibility that Firefox itself had a performance issue as well.

"The software doesn't perform well with your use case, huh? Have you tried getting a different use case?" :)

CPU doesn't execute <program> fast enough. Have you tried a different <program>?

It would help if you post the details of the machine you are performing the switch.

> For a few months I've been using Firefox full-time and I'm finally switched for good.

What was different for you that made switching permanently to Firefox acceptable this time?

> For a few months I've been using Firefox full-time and I'm finally switched for good. I encourage others to give it a shot.

I agree. I tried to switch several times since 2016. A couple of months ago, Firefox finally performed well enough I could do it. I'm the type of person who has five windows open with twenty tabs each, and it handles that now.

I also added Container Tabs Sidebar, DuckDuckGo Privacy Essentials, Firefox Multi-Account Containers (this thing is amazing for keeping sites separate!), and uBlock Origin. I like it now more than Chrome.

For me the killer addon is Temporary Containers. It creates and destroys containers for each domain or subdomain, but you can include exceptions.

I've been keeping ten to twenty windows open, with 100-800 tabs total, in Firefox, since before Chrome existed. Chrome never has even come close to being able to handle that. I am so baffled by stories like yours.

Same here. I use Firefox with Tree Style Tabs extension with lots of open tabs. I could never manage that in Chrome.

Ever since the Quantum updates Firefox has become faster but it wasn't very slow to begin with. Maybe it was a bit slower on some netbooks bit it hasn't been an issue on my PC.

Workona + The Great Suspender make it manageable for me (72 tabs open currently, probably peaking over 100 in an average workday).

Switched a year go, no issues, never going back to Chrome.

Same here, with my recent OS install I've gone Full Fox and loving it.

Because Firefox disabled all add-ons and only 7 work.

I think you are talking about firefox for android? I have 19 active right now on firefox for macos

Interestingly, this release fixes an undefined behavior bug in Rust code.

"CVE-2020-15254: Undefined behavior in bounded channel of crossbeam rust crate" marked as high.

So now we see undefined behavior rearing its head on Rust too... I wonder how long until it gets to where C++ is.

"Undefined behavior" isn't something that language designers put in because they're dumb, it's a necessary consequence of having many compiler implementations and supported platforms.

Once an alternative Rust compiler comes onto the scene, the language will necessarily fragment too.

Another large reason for undefined behaviour is to allow for optimizations. The optimizer of a particular compiler version is well defined by the compiler's source code and binary. But in order to be able to ever change the optimizer, you need to have undefined behaviour in the language specification.

Undefined behaviours are the accidental outcome when a program fails to be well formed. Nobody who writes compilers goes out there thinking that undefined behavior is the key that let's them optimize code to run faster.

https://en.wikipedia.org/wiki/Undefined_behavior#Benefits

> Documenting an operation as undefined behavior allows compilers to assume that this operation will never happen in a conforming program. This gives the compiler more information about the code and this information can lead to more optimization opportunities.

Yes exactly the requirement that program is "well formed" is what leads to optimization opportunities. Not its "ill formedness". Nobody writes optimizations thinking "this program is ill formed I can enable this and that optimization" but exactly the reverse. "This program is well formed so I can enable some optimization."

I think we are talking past each other. When I said "undefined behaviour" above, I meant the concept of it. The fact that the language spec has undefined behaviour is what enables the compiler to make assumptions. The language spec could for example say that out of bounds indexing in an array is defined behaviour, and defined to raise a trap. But implementing that would slow down a lot of code.

You refer to the ill formed code itself that violates the language specification. That's not what I meant. Yes, the fact that this code becomes garbage is a side effect of the optimizer's assumption that code is always well formed, instead of some arcane idea of punishing people who don't follow an arbitrary specification. I think we both agree.

This is not a new thing, whenever you use the "unsafe" keyword you have to pay attention to undefined behavior. For instance, having two "mut" references at the same time to the same memory has always been undefined behavior in Rust; you cannot construct them without "unsafe", but it's very easy to do with "unsafe" (just cast to a pointer and back). In the same way, "unsafe" allows you to construct references with an arbitrary lifetime, and many other potentially-undefined things; see https://doc.rust-lang.org/reference/behavior-considered-unde... for a list.

I don't think Rust will "get to where C++ is" (whatever that means), because (outside of compiler bugs) you will not have undefined behavior unless you use "unsafe", and most Rust code doesn't use "unsafe"; the pieces that do use "unsafe" can be carefully audited (instead of auditing the whole code).

Undefined behavior on rust isn’t anything new, there’s even a section in the book about it. https://doc.rust-lang.org/beta/reference/behavior-considered... The issue referenced happens in an unsafe code block, see this patch. https://github.com/crossbeam-rs/crossbeam/pull/533/commits/b... There is no promise that Rust protects programmers from making mistakes in code marked as unsafe.

Still no tap or pinch to zoom on Mac, I really want to use FF, but the lack of basic feature parity to Safari/Google on usability issue is astounding. I don't care about picture-in-picture, I just want basic trackpad support.

It has been available for some time now, just not enabled by default. Try setting apz.allow_zooming to true and restarting Firefox.

I have, it's not equivalent to Safari or Chrome, the tap-to-zoom feature is essentially non-functional.

Also, for a behavior that is standard at the OS-level I shouldn't have to literally flip variables after a scare-warning "touch this and break stuff".

allow_zooming is on track to ship in Firefox 83 (November 17): https://bugzilla.mozilla.org/show_bug.cgi?id=1671235

Is this two finger tap to zoom or pinch? Not clear to me.

Pinch to zoom gesture on macOS and Windows. (I don't know if it will ship on Linux at the same time. Linux always has unique graphics issues that take longer to fix.)

That’s too bad, I’ve wanted to switch back to Firefox for years but tap-zoom is a deal breaker. My vision isn’t great and tap-zoom is muscle memory across macOS. It’s just not a macOS application.

The only problem I have with Firefox is due sites optimizing for the Chrome monopoly,and I refuse to let the monopoly force me to give up on internet freedom.

As Chrome slowly takes over all remaining browser market share and more websites are becoming Chrome-only, one feature I wish Firefox had was a way to open a "Chromium tab," which would just be a tab that contains a frame of a Chromium instance so I can use these Chrome-only websites in Firefox without having to open Chrome.

You could use something like https://add0n.com/external-application-button.html that opens the page in an independent instance of Chromium.

Many pages "only support chrome" but work perfectly fine if you switch your user agent to appear like you're using chrome

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

Is there anyone who gets 'Your browser is out of date' type of notification on slack in Firefox, windows 10. I can open slack app on linux with Firefox, but for Windows, slack insists that Firefox is out of date and I should use another browser instead.

Long long time FF/Navigator user at this point mostly by inertia. The biggest issue I have (forever) is FF is a memory pig that eventually becomes sluggish.

Though I do have a lot of tabs in four windows, I've set prefs so it only loads tabs when I click on them. So in reality I have a lot fewer tabs "active" on any given day. Yet I regulary see memory use grow from an initial 600MB to 4 or 5GB. FF seems not to properly release memory back, in particular after new tabs with video are closed. I might live with it were it not for the eventual slowdowns.

This is definitely not normal. Have you checked about:memory

You might be "relieved" to learn that Google Chrome/Chromium is almost universally considered heavier on RAM.

This might partly be due to the "one tab, one process" architecture. Of course, if you are experiencing a memory leak in Firefox, that's another story... But I can usually trace that to a memory leak in some heavy javascript "application" I have open.

Oh yes, Chromium is simply not capable of supporting tab-intensive workflows that Firefox handles without issues. It becomes completely useless after just a few dozen open tabs, from both UX and performance perspectives.

> This might partly be due to the "one tab, one process" architecture.

With site isolation/"Fission" it looks like this isn't going to get any better in Firefox, either, as they'll have to move to "one domain/origin, one process", which could be even worse in terms of memory overhead due to the number of separate processes required.

We fix memory leaks when people report them and we can reproduce them. Usually it is due to some combination of particular addons and particular websites, but of course the obscurity of any particular leak doesn't help you if you are hitting it regularly.

> Restoring a session is 17% quicker, meaning you can more quickly pick up where you left off

I actually wouldn't mind if they throttled this a bit. I keep a tab or two (dozen) open to Hacker News posts I haven't gotten around to reading yet. I dread restarting the browser, as several tabs inevitably end up displaying the "Sorry, we're not able to serve your requests this quickly" page.

Not sure why you're seeing that, a tab shouldn't be reloaded until you click on it the first time in the reloaded session, just retested here to confirm that is the case.

not OP, but probably the tabs are pinned, that causes them to load at startup

I've found the Auto Tab Discard extension to be really helpful. It generally lowers the impact of having lots of tabs open and on startup only the visible tabs load.

Webrender is broken for me in 82 - I get a blank bar across the top of every page.

I've switched back to firefox as my main browser recently and I'm pretty damn happy about it. I'm even more now that I see those perf improvements <3

Huh, this update broke my FF installation on my Windows work computer. First time ever. Had to completely uninstall and reinstall v81.

"DevTools now shows server side events in the Network panel"

Someone nontechnical wrote these notes - it's Server Sent Events.

Or someone just made a typo - thinking "sent" but typing "side" is not too far-fetched (I do these kind of typos all the time when I'm distracted).

Thanks :D

That was confusing me how the browser was seeing server events.

I have seen developers make this exact error quite a few times over the years.

Hopefully they're not paid to write articles, then.

Pretty sure one of the places I’ve seen it crop up was an article on hacks.mozilla.org.

I just took a look at https://hn.algolia.com/?query=%22server-side%20events%22&sor..., it’s mostly (though certainly not entirely) people making this error.

I'm getting "Secure connection failed" errors from a bunch of different sites

Which sites? Does the connection page show have a more detailed certificate error message? Does the site work in Chrome?

Anyone on macOS having trouble with “Find in page” not finding text recently?

Is the address bar still gross and giant?

It's definitely still configurable.

Where do you go to configure this? For awhile it was in about:preferences but then they moved it to the userPref.css file. Is it back in normal config? I would upgrade if I can keep my old address bar

Hamburger Menu -> Customize -> Density -> Compact

I'm happy to report that the text doesn't change size upon selection anymore. I think that some time ago they hit "peak address bar", but now that text size change has gone, it's perfectly usable again.

Websites that use flexbox-based layouts load 20% faster than before; Where is the evidence? The TP6 benchmarcks?

> MediaSession API has been enabled by default which allows web authors to provide custom behaviors for standard media playback interactions, giving them more options than ever.

/me checks what is this... then, from the draft (https://w3c.github.io/mediasession/#security-privacy-conside...):

> The API introduced in this specification has very low impact with regards to security and privacy.

Ok... my concept of privacy, may not be the same as your concept of privacy...

But it look like it has zero impact on privacy. Like I don’t get it. This is a website being able to tell the browser, “hey I’m playing $Song” so it can show out-of-page media control buttons like you get on your lock screen. Where’s the issue?

How does this have an impact on privacy ? It only allows a website to send you additional information about some media, which you can display or not.

It's not sending anything back to the server.

I haven't looked into it much but could it be another point in a browser fingerprint?

how? it looks like all you can do is send some information to the browser, and register some callbacks (if the user presses those keys). I guess you can fingerprint for the existence of the feature, but that's already 100% correlated with the user-agent, so it's a non-issue in my opinion.

Reply on news.ycombinator.com