Firefox 52 released | Hacker News Reader

Firefox 52 released

mozilla.org

489 pointsgaul360 comments

Firefox is the first major browser to support CSS Grids out of the box.

https://developers.google.com/web/updates/2017/01/css-grid is a simple introduction to it.

http://gridbyexample.com/ is probably the best reference site.

https://tympanus.net/codrops/css_reference/grid/ is another very nice single-page reference.

Some technical examples from Igalia, who have been implementing Grid in Blink and Webkit: http://igalia.github.io/css-grid-layout/

Some lovely creative examples from Mozilla's Jen Simmons:

http://labs.jensimmons.com/

http://labs.jensimmons.com/2016/

And the Firefox Devtools now also have an excellent tools for learning, visualizing and debugging CSS Grids: https://hacks.mozilla.org/2016/12/css-grid-and-grid-highligh...

Well done, new feature coming along with Devtools for it.

And we expect Chrome 57 to ship Grid (which should ship around March 14th), Safari 10.1 (part of macOS 10.12.4 and iOS 10.3, expected March/April), and Edge to probably ship it later this year.

Sorry to be so blunt but I dont care about another flexbox that much. What they should have done is to implement the shadowdom. Instead we have to wait another half year for them to get to parity with Chrome.

What's the difference between grids and flexbox? To the untrained eye, they look like two alternate layout systems where one should suffice.

Flex is only built and meant for handling linier layout, either up and down or left right, and ot does so well with good flexibility and options and developer experience. Or at least compared to the other possible options. It was never intended for making grid layouts, which people are often trying to make it do, because all the other options are so lacking.

Basically grids is a 2D layout system, while flexbox is a 1D constraint system that results in stuff happening on the second dimension. They're complementary and it would be foolish to use one when the other is needed.

Does this mean we're finally coming full circle after all these years back to Tables for layout?

No CSS Grid Layout is far, far more powerful and versatile than tables ever were. Certain significant classes of layout problems that were either a major pain or even outright impossible with current CSS tools are about to become fairly straightforward.

some of us never left

I wrote up an explanation here: https://rachelandrew.co.uk/archives/2016/03/30/should-i-use-...

Very nice. Thanks.

Firefox is one of those projects that just keeps on giving, and it's all thanks to the hard work put in by all the contributors. It's easy to forget that sometimes. We use Firefox headlessly to render videos at https://www.musicvideodispenser.com , a task that certainly wasn't the initial intended use-case for that browser, but it still works, and we have the Firefox team to thank for that!

Can you elaborate on the "using Firefox headlessly" ?

I recently had to deal with WebRTC on a headless machine and managed to hack up something with some Python scripting and GStreamer. I didn't know it was possible to run Firefox headlessly and now you got me wondering whether that would have been preferable to my stitched up solution.

I also tried to hack together WebRTC headlessly a while ago actually! I looked at GStreamer and Kurento, but ended up using Janus instead. Anyway, turns out you can use Xvfb, essentially a "fake screen", to headlessly run GUI programs: https://en.wikipedia.org/wiki/Xvfb , which is what we're doing with Firefox.

Thanks, I will look into it. I started off with Janus too but I had some odd bugs and segfaults. Since time was short I decided to give Kurento a try and since it worked, I stuck with that. There are still some warts but it does the job so far.

Xvfb looks to be pretty straightforward about what it does; now I just need to figure out how to control firefox after it was started.

Firefox is basically the first port of call for running web tests on buildservers. Run Firefox headlessly in xvfb, then throw Selenium or similar at it. It's a pretty mature setup.

I would think PhantomJS, being a headless, scriptable browser (plus built on WebKit so closer rendering to the most popular browser), would be the "first port of call."

Why use a fake browser when you can use a real one? It's also harder to inspect the actual rendering of an invisible browser.

Near the top of a google search for 'phantomjs vs selenium', we have folks on HN complaining about phantomjs https://news.ycombinator.com/item?id=8418071

I have used phantomjs (with casperjs) for scrapping a website, it was complete nightmare.

1. It use javascript to script so writting sequential code wasn't possible so the code ended up using continuation passing style.

2. The very first example snippet on the front page don't work in the repl (https://github.com/ariya/phantomjs/issues/11180). So the repl is useless.

3. When an error occur in your script (not the one on the webpage) there is no message anywhere no matter what you do.

My mistake was to continue with phantomjs the hype around made me think others options would have been worst.

I have used selenium+firefox+php to do a quick test it was wayyyy better, just for n°1 using sequential php remove a lot headaches.

Even better, selenium has preconfigured containers for Firefox and Chrome and that use Xvfb. At the start of a web automation project I worked on, I spent a lot of time trying to get various versions of browser, selenium and pyvirtualbuffer installed and running with mixed results. These containers are well tested and work very well.

On a unix system you can run any browser headlessly by using a headless X11 display such as Xvfb.

Instead of having a real screen, Xvfb uses an in-memory framebuffer for screen output that you can screenshot etc if needed.

It's as easy as:

  Xvfb :123 &
  DISPLAY=:123 firefox
  ### firefox is now running and outputting
  ### to the framebuffer display on :123, and
  ### can be driven by WebDriver etc.

For practical use you want the xvfb-run wrapper script, and use a saner mode than the default 256-color one:

  xvfb-run --auto-servernum --server-args="-screen 0, 1024x868x24" firefox

You may be interested in the work currently being done to add a native headless mode and browser control API into Firefox.

https://bugzilla.mozilla.org/show_bug.cgi?id=1338004

I did a lot of work on Firefox's NPAPI implementation over the past decade. Improving that code was rewarding in that it had very tangible benefits for Firefox users but it was pretty clear that improving Firefox's code was only going to get us so far, and not far enough. The system is a mess and I couldn't be happier to see Firefox dropping support. I hope I can be the one to rip out the code altogether some day.

What I like most about firefox is the mobile version supports addons like ublock origin. In mobile Chrome the support for add blockers is not really well supported.

uBlock is the sole reason I use Firefox over Chrome on Android.

Not only does FF Mobile support uBlock, I find it to be an excellent day-to-day mobile browser in its own right actually. Compatibility and performance gap vs mobile Chrome is much smaller than on the desktop for the sites I visit (or is even non-existent).

There's also the part where youtube videos keep playing if you move to another tab or app

That's how I listen to music and talks on the go :) Ublock origin and that.

https://news.ycombinator.com/item?id=13746830#13751653

You can use newpipe (find it on f-droid) as an alternative to the youtube app. It has some other benefits too. It seems to block ads as a side-effect. You can also fix the stream quality. It has a few limitations too, but it's better than playing them in firefox, and even the youtube app for my use cases.

Just tried it out on my phone (android), and there's no notification tray widget to pause the audio?

Is this just not a default setting? An add-on I can get? Or just not something you can do?

So worst case scenario is:

    1. Wake the screen.
    2. Unlock phone.
    3. reopen Firefox, because I've checked texts/email/social feed whatever after starting the music.
    4. find the tab playing music.
    5. Navigate to whatever in that tab controls the audio .
    6. pause it.

vs. what I've come to expect from other apps (inc. Chrome, despite it pausing audio even when you swap tabs!) of getting a small widget in the notification drawer with at the very least a play/pause button I can hit after waking the screen.

I wonder how accidental that feature is. It's certainly not something that Firefox can advertise, since other apps which did that have been kicked out off the Play Store.

Also, I also found that feature to be not very reliable since Android 6, which quite aggressively puts background apps to sleep (at least on my phone with "only" 2GB of RAM). I find that any extensive use outside Firefox will make the background playing stop rather quickly.

Sounds amazing.

It annoys me every day that I can't put on a Youtube video and then continue listening to it while I do something else with my iPhone afaict.

If I didn't want that, I would pause it first.

Not supported at all, Google don't want to kill ad revenue. There are work-arounds that are a hassle but FF + UBlock or Ghostery is awesome on mobile where ads are just wasted bandwidth

Ghostery is not awesome, ghostery is profiling people who block ads to provide data to advertisers so they can find ways around ad blocking.

Using ghostery to block ads and tracking is shooting oneself and everybody else in the foot (feet?).

Is this for Android only or iOS also? I can't find an option for addons in iOS Firefox.

Only for Android; Firefox for iOS isn't really Firefox.

Only Android or Desktop, because iOS does not allow other rendering engine like Gecko, only its own webkit.

Firefox for iOS is a lie. It's a non-Firefox browser just set up to sync FF settings and bookmarks and kinda look like Firefox.

That's the case for all non-safari browsers

> Firefox for iOS is a lie.

That's a bit harsh.

> It's a non-Firefox browser just set up to sync FF settings and bookmarks and kinda look like Firefox.

It's a WebView-wrapper like all other iOS "browsers" because that's everything Apple will allow. To be fair though, synced data, history, passwords and settings is probably one of the most important things for a mobile user, where entering things manually is a pain in the ass.

Basically NOT having this would risk losing users on the desktop (to something which does sync). And right now, Firefox can't afford to do that.

And still the usability of FF on iOS is so much worse then the one of safari it stands no chance and that is not all Apple's fault :(

What do you mean it's not Apple's fault??? Apple is 100% at fault. They are the ones who won't let actual Firefox be released for iOS in normal channels.

I'm not too familiar with how heavily locked down iOS is: does this "WebView only" policy only apply to apps released via the AppStore? Meaning, can I build locally and sideload install an app that uses my own engine? And if that existed, could I open source it for others to use?

Yes and yes, though that would require your users to pay $100 per year, install Xcode, and compile your app.

> and compile your app

Which in turn would mean that lots of people would have to shell out several hundred dollars more for a Mac just to build it.

It's not exactly a stellar deployment-model when you try to push it at scale.

These days you can sideload apps with a free developer account (with certain restrictions), but it's meant for development - AFAIK using it as a general-purpose app distribution mechanism is a violation of Apple's ToS. But people do it anyway, and you could too.

Separately from the ban on third-party browsing engines, iOS has technical restrictions preventing you from mapping unsigned code as executable, so JITs don't work. There may be some ways around that if you have a developer signature, but its messy at minimum.

Sadly this is Android only.

We can only blame Apple for that though, and not Mozilla for a lack of trying.

I was extremely surprised and disappointed (to say the least) when I moved to iOS and found this missing.

The only thing missing for me in Firefox on Android is text reflow. I just can't use a mobile browser without this feature.

There are some addons which try to emulate text reflow on tap, but it's pain to use at best.

Can't you use reader mode?

Text mode doesn't work on some sites because it can't identify the content. Furthermore it turns the page in a kind of text only ebook whilst text reflow works within the original page.

Reader mode is a workaround, text reflow is the solution. It a reason to use Opera on Android. I wonder if it's so hard to implement that nobody else coded it.

Firefox on Android used to have text reflow, then they removed it. Then the crappy extensions were made. I used to use text reflow on Conkeror (Firefox based) on desktop, but a year or two after they disabled it on Android they removed it from Firefox. I agree, I also use Opera on Android, for this feature.

Thanks. I googled for that and found this bug report https://bugzilla.mozilla.org/show_bug.cgi?id=710298

Unfortunately the about:config setting for text reflow in one of the last comnents is not in Firefox anymore.

Ironically that page is an example of a page where reader mode doesn't work.

Reader mode seems to break on web pages with an ampersand in the URL (on my phone, at least).

Yandex browser can use chrome extensios on mobile

There was this other hn post on latimes.com which wouldn't open with adaway. I have firefox with ublock on the phone, but I rarely use it. So I decided to try it to check if it circumvents it. It doesn't, but there is anti-adblock-killer for those cases. That needs tapermonkey, but tapermonkey doesn't work on mobile. So there's this other workaround based on USI (Script-Manager) for that. After all that, I got anti-ad-block killer to work, but latimes still wouldn't open. I found this open issue https://github.com/reek/anti-adblock-killer/issues/2513, and the proposed workaround finally worked. Just goes to show how difficult it is to block ads on mobile.

> Added support for WebAssembly, an emerging standard that brings near-native performance to Web-based games, apps, and software libraries without the use of plugins.

I've been watching this from the sidelines... what's the best way to dive into WebAssembly? Or is it just waiting for tooling to catchup to produce it for WebAssembly-enabled browsers to execute?

The WASM Explorer is a good place to start. I wrote up a little blurb about it: https://hacks.mozilla.org/2017/03/previewing-the-webassembly...

Lin Clark's Cartoon Intro to WebAssembly is also great: https://hacks.mozilla.org/2017/02/a-cartoon-intro-to-webasse...

Rasumus Andersson's WASM Intro dives deeper into the binary format and wasm semantics: https://rsms.me/wasm-intro

Lastly, the WASM Community Group's page at http://webassembly.org is quite good.

The tooling story is currently the same as it was for asm.js: Use the Emscripten fork of LLVM. Upstream support is being worked on in LLVM, and Rust has a cross-compiler in the works.

Given that WASM is a stack-based language, do you think it would be a good target for a modern concatenative language? Something like, say, Kitten?

http://kittenlang.org/

This video is a great start to it

https://youtu.be/3XrGjSnPHGY

> Enhanced Sync to allow users to send and open tabs from one device to another.

If you haven't used this, try it out, it's a fantastic feature. I use it all the time to send a page I'm reading on my phone to my laptop, and vice versa.

Sounds like a part of what pushbullet does. That's a welcome function.

It's also been in Chrome for years.

In the form of an official Chrome extension that was discontinued just over two years ago.

How is it different from the native functionality in Chrome that lets you see (and open) any tabs you have open in Chrome across any device?

The difference is that you can push a tab from one device to another. So if you're reading something on your phone, and push it to your laptop, it will automatically open in a new tab there once it syncs. It's good for when you're on the go, and want to remember to read something later on once you get back to your computer.

By the way, Firefox also has the feature you mention (view tabs that are open on devices).

Ah, I guess the benefit of "pushing" a tab shows in stateful applications or something, so it keeps the same "state" on the new device? Or, does it keep the same distance scrolled down the page so you can pick up where you were?

I'm confused because I open desktop session links from my mobile all the time and I'm curious if there's features I'm missing out on by pushing from one device to another, instead of syncing.

For example, if I open Chrome, I see the tabs open across all my other computers[1] and I usually just resume relevant tabs from there. It keeps me logged in, but sometimes reverts things like filters or sorting (on-page JS) unless it is part of the URL. It doesn't scroll me where I was in a page either, which would be nice.

[1] http://i.imgur.com/MgPCMy2.png

But doesn't Chrome send your data through Google?

Yes, though you can encrypt it locally before it does so.

But do you do? (Honest question; I mean: is there a reliable and easy way to enforce encryption happens?)

Yes, it's a simple option in the settings: https://i.imgur.com/0np3ItM.png

Oh okay, didn't know Chrome provided this feature natively, I thought you were referring to some kind of crazy DIY encrypt-things-behind-Chrome's-back monitoring script setup.

Thanks, and thanks for the precise screenshot :)

It's not on by default though, meaning most users will be passing data their browsing data through Google unencrypted. Not only tabs but cookies, full browser history...

I'm always surprised that no one seems to think this a big deal. People will install and recommend tracking-blocking extensions while allowing Google to hoover up all this data without a second thought.

By default Google encrypts it. Its clearly in that picture

No, danielsamuels has manually changed that setting. The default is to only encrypt saved passwords: https://www.spinda.net/files/chromesync.png

On desktop, does it still require installation of this addon to get the UI bits? https://addons.mozilla.org/en-US/firefox/addon/send-tab-to-d...

EDIT asking because it's been this way for a long time, see https://wiki.mozilla.org/Services/PushTabToDevice#Desktop_Im... which says "Currently, push-tab-to-device on desktop must be used through an addon: [send-tab-to-device]"

No, I don't have any extensions like that installed, and the feature just showed up for me.

Indeed, we added a "Send Page/Tab/Link to device" menu item to the contextual menu.

Is it new to 52? I hadn't seen a menu until today and really wanted to have that feature.

Yes this UI is new, there was an unofficial add-on [0] before that did something similar.

During my internship this summer we also added Push capabilities to decrease the time it takes to send a tab to another device (it used to be in the order of minutes).

[0] https://addons.mozilla.org/En-us/firefox/addon/send-tab-to-d...

It appears you don't have to have tab sync on for it to work. Mine does take a very long time to be received. Definitely a great feature, I don't want full tab sync just this.

I don't have that addon installed and am able to send links to my other devices.

Is it just me or was this already available in Firefox at some point in history?

I seem to remember I used to be able to send an open tab on my Android phone to my desktop Firefox, but that option later disappeared with some update.

Edit: It sounds a lot like the "send to device" add-on mentioned in another comment. Was this bundled with Firefox for Android at some point?

This feature appears and disappears randomly and it has me demented. Much of the time it works, and I am definitely signed in to FF Sync on desktop and mobile at all times. But that Firefox icon just vanishes for weeks at a time from my mobile browser. I even tried sending a tab from desktop to mobile recently (to troubleshoot) and it arrived over 24 hours later as a notification. Still no way to send tabs back to the desktop at the moment though, and all I can do is hope it reappears some day.

I see it in the menu on top right, and on android share dialog. Always worked in general, though with frustratingly long delays at times.

Starting from this release, Firefox now requires PulseAudio for sound on Linux [1]. ALSA can still be enabled at build time for now, but is not supported.

[1] https://bugzilla.mozilla.org/show_bug.cgi?id=1247056

How does this fit in with the current state of sound on Linux?

I'd guess that this would impact relatively few people if they're willing to require it.

I do most of the audio work on the Wine project.

I personally think it's a bad move to drop official support for ALSA. Not everyone can or wants to use PulseAudio, for various good and bad reasons. Supporting multiple backends also keeps you "honest," so to speak, preventing you from relying on certain behaviors or features that only some platforms support.

But, I can understand the practical arguments behind dropping support for ALSA. The ALSA API really sucks to work with. There's a fair number of bugs and device quirks and whatnot that you have to take care of as an application developer. Standardizing around PulseAudio means only PulseAudio has to care about those quirks, instead of every single application. In practice, the vast majority of users use PulseAudio; there's a small (but vocal) minority that don't.

So really there's no clear winner on whether to provide support for ALSA.

Like many Poettering products, PulseAudio polarizes. There's one minority of users with either quirky setups or just bad luck for whom it just doesn't work, so they have to stick with ALSA. One minority of users (including me) who really loves the features it puts on the table (e.g. per-app volume control, effortless network transparency).

And of course, there's a large majority of users who couldn't care less, and who will just choose whatever their distribution installs (which is why I think the most Linux users not using PulseAudio are on those distros which don't install any sound stuff by default, e.g. Arch).

I'm definitely in the camp that takes whatever the distribution gives me. Of course, the choice of distribution is heavily driven by what I want from it (e.g., KDE desktop, etc). I've honestly never given any thought to what audio subsystem it uses.

Fortunately, OpenSuSE Leap appears to be a PulseAudio distribution.

"PulseAudio solves all your audio problems on Linux, except the ones you have."

> Like many Poettering products

I didn't know pulseaudio was a Poettering product, but I know very well that pulseaudio has been a nightmare for me over the years on most setups with a variety of linux boxes and distros. And the fix when an audio problem arise is simply to uninstall pulseaudio and use alsa.

Now that I know pulseaudio is Poettering work, this makes a lot of sense.

Pulseaudio, Avahi and Systemd are his. Seems to be directly or indirectly inspired by Apple.

I have seen claims that Network Manager is his as well, but so far i have found no clear link.

What if one just wants basic audio as provided by ALSA

Actually quite a few people don't run PulseAudio, as you can see from comments in that bug and the amount of votes the Arch Linux task [1] gathered in a short time. This is definitely going to be a problem.

[1] https://bugs.archlinux.org/task/53195

Well when firefox 54 will be out, we without a working pulseaudio will have no choice but to switch to a different browser (pale moon ? otter ? vivaldi?).

I suppose mozilla does not care enough about losing these fringe users to actually do something about it.

At this point i wonder if they care for anything but OSX, and are either letter Gnome dictate, or at the very least taking strong cues from that camp, about the future of the Linux desktop...

Like i needed another reason to stay on 45esr and pondering moving to Palemoon permanently.

Frankly it seems like the Firefox devs could just hand the Linux branch off to the Gnome devs at this point.

I just added Firefox 52 to Browserling. You can try this new version at this URL without installing it:

https://www.browserling.com/firefox/52/news.ycombinator.com

If the demand is too high then you'll have to wait in a queue for a while to try it. I'm adding more servers right now to let more people try it without waiting.

Out of curiosity, what's the difference between browserling and a service like browserstack?

Biggest difference is that Browserling is powered by alien technology, unlike Browserstack that is powered by puny human technology.

Nice, Can your service be used to test Chrome extensions (like extension Popup)?

Yes, you can install Chrome extensions and test them but when you close Browserling, your extensions will be gone.

Congrats Firefox team! I've been using the latest Nightly releases (for container tabs!) as my daily driver and it has been a rock-solid experience.

Thanks for the hard work

There's now also a Firefox Testpilot experiment for container tabs: https://testpilot.firefox.com/experiments/containers

Thank you! I've been eagerly awaiting containers for a while, but didn't want to switch to nightly

"online trackers can’t connect your browsing in one container to another"

Yeah sure, because we all know that deleting cookies makes you totally anonymous on the internet. /s

This looks great! It would make for a natural fit with tab groups, unfortunately Mozilla has removed that feature, and the extension will not be updated [0] to support the new extension API at the end of this year.

[0] http://fasezero.com/lastnotice.html

> Removed support for Netscape Plugin API (NPAPI) plugins other than Flash. Silverlight, Java, Acrobat and the like are no longer supported.

Good bye, you will be missed (not).

I will miss them...

They will return as WebAssembly, just wait.

And we'll be better off because of it.

Until 20 years in the future when Firefox 66 (if Mozilla is still around after murdering the only reason people use their browser) when the new hotness is whatever WASM's successor is, after it's widely deployed and in use everywhere, people making snarky comments about how they're glad WASM is dead, and we get to piss away another tens of millions of man hours and annoy tens of millions of users by changing things for the sake of change rather than objective technical benefit.

Yes, I'm slightly bitter. I'm starting to think the Suckless guys have the right idea...

I have my doubts, at least I could disable plugins when I wanted to.

Now I have to disable dns entries.

With this goes the support for DivX player and VLC plugin for Firefox. I'd surely miss the ability to play any format in my browser until PPAPI replacements come up.

Firefox is not implementing PPAPI. Additional audio or video codecs can be added to a browser but that doesn't require a plug-in architecture. Firefox's UI makes it a little confusing because things like the OpenH264 video codec from Cisco is listed under Plugins in about:addons and the description starts with "This plugin is automatically installed by Mozilla..." Maybe they'll make an addons section called Codecs for listing things like that and the junk like the Widevine DRM.

They were the only way to get hardware accelerated video on Linux.

Even today, Firefox decodes all video (even h264) on CPU.

Which causes, with 4K 60fps video, 50% CPU load on an entire core for me.

I think this is only true on Linux. On my Windows and macOS installs, hardware video decoding works fine.

Definitely. Mozilla removed hardware decoding on Linux in Firefox 3.6, and never added it back.

The bugs on bugzilla are all marked "WONTFIX", and Linux seems to be an afterthought.

To be fair, I've looked through the Bugzilla threads about this (here are some for reference: [0][1]) and it seems to be a very difficult thing to implement. It's a bit beyond my reach, but my understanding is the tricky part is compositing the frames decoded by ffmpeg onto the document, and getting that all to work with all possible CSS transformations and such. So it's not just a matter of the Firefox devs neglecting Linux.

[0]: https://bugzilla.mozilla.org/show_bug.cgi?id=563206

[1]: https://bugzilla.mozilla.org/show_bug.cgi?id=1210729

Except that Chrome manages to do this, as do all other linux browsers.

And Mozilla, which explicitly writes

    Out of Scope
     - Linux testing

On their QA pages in the wiki regarding this, well, doesn’t inspire confidence. I'm still using Firefox, still supporting Mozilla, but I'm disappointed nonetheless.

> Which causes, with 4K 60fps video, 50% CPU load on an entire core for me.

Is that a punch line? Because it sounds like one. What is that, like, 6.25% of your total CPU power?

I say this because many people have machines with CPUs that can't decode 4K 60fps video in realtime at all. So your stat sounds like a great deal to me.

Having said that, hardware video decoding is important for this very reason. It's a shame that the situation isn't better on Linux. It seems like, no matter what site or browser--Chrome or Firefox--VLC always plays the same video with far less CPU usage.

It is problematic especially for newer codecs like h265 which uses approximately 3x the CPU cycles.

Which pushes you beyond what a single core is capable off.

You can get hardware accelerated video in Firefox on Linux with Flash? I have tried that by adding &nohtml=1 to YouTube videos, but it does not work. The player changes, so I guess it's the flash player. But my laptop still struggles to decode the videos smoothly.

It depends. For me, flash actually used VDPAU and VAAPI, so I got hardware-accelerated decoding, and it was nice.

The mpv plugin worked as well for HTML5 video with hardware acceleration.

Now, well, not anymore.

> Java

Another internal tool to rewrite from scratch, yay.

> Acrobat and the like are no longer supported.

Is the build in PDF viewer still burning through battery power like the electric heater it turned my laptop into the last time I used it? Been some time so that would be interesting to know.

> > Java

> Another internal tool to rewrite from scratch, yay.

If you care about security in the slightest, you should probably already have done that.

> > Acrobat and the like are no longer supported.

> Is the build in PDF viewer still burning through battery power like the electric heater it turned my laptop into the last time I used it? Been some time so that would be interesting to know.

You can still use an external reader of your choosing; for example I'm using the marvellous Okular. You just can't use an NPAPI-based PDF reader plugin.

> If you care about security in the slightest, you should probably already have done that.

Signed internal only use tool. Not running untrusted applets anywhere. Also no time to rewrite bug free working code.

> You can still use an external reader of your choosing;

So the answer is that it is still that bad? I guess I can just browse the web using wget and less if external tools are the next big thing. /s

If you care about security in the slightest, you should probably already have done that.

These plugins have been click-to-play in the major browsers for a long, long time.

Obviously they still provide some attack surface, but the biggest threats today don't come from plugins, they come from the main browser functionality, thanks in no small part to all the things browsers now have to provide so we can still do useful things that we used to do with plugins.

> > > Java

> > Another internal tool to rewrite from scratch, yay.

> If you care about security in the slightest, you should probably already have done that.

What has one to do with the other

Haven't done anything with Java applets in years, but can't you somehow re-use most of the code and attach a new UI (yeah, still a lot of work probably)?

I haven't looked into the space recently; how would the UI be attached? The only thing I can think of would be <canvas> (2D or WebGL), which both seem like a ton of work that the vendors don't have yet. Or do a full rewrite to output HTML of some sort. I thought Java applets were popular partly because the widgets are built-in (not always, of course, and the old widgets look pretty horrible now).

I'm more thinking of something like Vaadin/GWT where you compose the GUI in Java code, in a similar style as e.g. JavaFX, and the framework takes care of displaying it in a web front end.

Or, if the application is not strongly coupled to other server-side systems, make it standalone with a Swing, JavaFX or whatever front end.

As recently as 3 years ago a customer of mine wanted a Java applet for drag-and-drop/file upload from Outlook. Part of the problem was that Outlook doesn't expose serialized object data to DnD; it was necessary to call into a special API to get at the attachment data (but MS invented DnD so it can't be called non-standard I guess).

However, another difficulty with using browser file upload controls was that, for security reasons, programatically (JavaScript) putting a file to upload is forbidden (readonly form attribute) by the respective browser API spec, so in the end the GUI part had to be coded in the Java/applet, too, when the HTML GUI would have been preferable.

So my point is that Java applets in recent years weren't used for GUIs so much as they were used for piercing the browser sandbox, and I don't see a solution for this problem now that applets have gone.

I needed the applet for the unrestricted filesystem access it offered when the companies key was trusted by the user. Now I could write a swing application that interacts with the server, making it less integrated and convenient to use (and depending on your opinion less pleasent to look at).

Crap. The city council in my town uses Silverlight to stream their meetings.

I know, I know.

Still, it means I'll be missing that silverligh support.

Do they support Chrome without Silverlight?

Nope.

Firefox ESR will still support it in their release of v52.

Well, hopefully some web guy at least got a nice lunch from the MS rep out of that deal.

> Good bye, you will be missed (not).

I'd like to say they won't be missed, but I've got a state Medicaid portal that uses some annoying Java applet to handle authentication I have no choice but to deal with (I've actually got a little flask web service that uses Selenium and Firefox in Xvfb just to log into the stupid thing and return the authentication cookies, since Chrome ditched NPAPI a while ago and IE isn't a sane option).

I won't be sad when applets finally go away, but that day is not today for me and it's going to really screw things up :/

Does Java Web Start support applets? I haven't had to use it in years but I seem to recall that it does.

I believe that is Oracle's official recommendation for dealing with NPAPI deprecation.

Several banks here in Brazil require a "security" plugin (which acts more like malware) for one to be allowed to use online banking. I wonder how the end of NPAPI will affect this.

If you're really Brazilian/in Brazil, you don't really wonder. You just know the answer: all those bank sites will tell you to use Chrome instead.

Bradesco's fortunately works on any browser.

Is the Brazilian "security" plugin an NPAPI plugin or a Chrome extension? Chrome doesn't support NPAPI plugins, so these banks either have non-plugin solution for Chrome or they don't support Chrome.

Many Chinese ecommerce and government sites require their own "security" NPAPI plugins. These don't work in Chrome, but the large majority of users in China run hybrid browsers that use both Blink and Trident (IE) engines so these sites that require NPAPI plugins can be seamlessly loaded in a Trident tab. The user doesn't know the difference.

I know Chrome stopped supporting NPAPI plugins long before Firefox, so simply "use Chrome" won't be the answer.

I doubt they will give up these "security" plugins, so the answer will most probably be an even more invasive system process. It seems that some banks are migrating to Warsaw, which is known for among other things causing problems with IPv6 (http://ipv6.br/post/bug-em-plugin-de-seguranca-de-bancos-blo...).

I don't understand why they removed support for Silverlight and Java, but kept Flash.

Are Java applets and Silverlight apps demonstrably less secure than Flash?

Just as a naive metric, if you look at the CVE database, there have been a total of three CVEs for Silverlight last year; whereas for Flash... well, I stopped counting.

I don't know of any bank or government agency that uses Flash, but I certainly know that the Hong Kong government uses Java on some of its websites.

It feels like Mozilla is keeping Flash around just because it's more popular with browser gamers and video watchers. I don't think that's doing the right thing.

> It feels like Mozilla is keeping Flash around just because it's more popular with browser gamers and video watchers.

That's exactly right. The amount of Flash content dwarfs the amount of other plugin content.

Banning other NPAPI plugins doesn't avoid the Flash security problems but it does avoid the other security problems. It's a step forward, but it's not the final step.

It's not about security. As you said, it's because Flash is more popular, so killing it will disrupt more users' workflows. It's a pragmatic decision…

Personally, I'm hoping we see a WebAssembly build of Flash eventually (once WA is featureful enough to support it), at which point the native plugin can die without breaking (as many) old sites.

> I don't understand why they removed support for Silverlight and Java, but kept Flash.

drm something something...shall I remind spotify web player still uses flash.

Just drop flash already, this thing has been agonizing the slowest death ever.

And still no support for <input type="date"> . I just can't believe Firefox is behind Edge on this one. I'm using Firefox as main browser for years now, and it feels so bad when I always need to use a polyfill when developing a form with date field - to support Firefox. It's embrassing that Chrome supported it since... Chrome 20, in 2012.

Date/time input types are behind a flag starting with v54.

Also Safari still doesn't support it...

I think this is in Nightly and now Firefox Developer Edition (54) behing a pref: https://twitter.com/mozhacks/status/821005161602646016

You might want to switch yourself to the ESR branch if you depend on add-ons that will not update to the new Web Extensions API. The old extensions model is scheduled to be removed in FF 57, later this year. But 52 ESR will be supported until at least mid-2018.

Edit: switched "plugins" to "add-ons".

ESR52 also maintains support for NPAPI plugins for those who depend on that.

Plugins are not affected by the WebExtensions API. Support for NPAPI plugins is removed in this release, support for XUL extensions is a different issue.

It's also scheduled to happen prior to the next ESR, which if the "every seven versions" period remains in use, will be Firefox 59. So switching to 52 ESR might still be a good idea for those reliant on XUL extensions not likely (or not able) to be ported soon.

Based on https://wiki.mozilla.org/Enterprise/Firefox/ExtendedSupport:... it looks like 52 ESR will be the latest ESR when 57 is released. Is that right?

Yup, this is the last ESR before that change is currently expected. Things could change over the next year of course.

Seems i need to consider a full time move to Palemoon, as i am currently on 45esr (their last GTK2 based esr).

Its really "funny" how it seems i can get Qt5 to work on an "old" X11 install, but GTK3 balks at same.

If the FOSS world really wants a foot in the door on the desktop they really need to learn to properly do backwards compatibility. Yes, it is not a glorious job unlike working on the latest shiny framework or language. But it is what has kept the likes of Microsoft in the top spot all this time.

Frankly i fear that once Torvalds steps down as kernel boss, Linux as a whole will fragment into a million variants with diverging APIs.

Past few days I've been making a Befunge JIT that targets WASM: https://github.com/serprex/Befunge/blob/master/funge.js

Ended up finding that Firefox 52 is overly accepting in validation. Dead code is allowed to pop from an empty stack, whereas in Chrome that's not allowed (as per spec). It's fixed already in next versions of Fx

> Past few days I've been making a Befunge JIT that targets WASM

why??????????????

Science isn't about why?, it's about why not!?.

This isn't science, this is art. Specifically, this is like making Pollock's Autumn Rythm out of play-dough.

I see you haven't recognized the reference?

The usual answer to this question is Because we can

The more helpful answer, though, is: Because it's fun (at least to some people)!

sigh I had to switch to Chromium because they've broken + disabled HW acceleration on Linux, and my 2016 netbook is too slow to browse amazon or read news without it. Here is the bug:

https://bugzilla.mozilla.org/show_bug.cgi?id=594876

In the release notes for 52, it says hangouts is broken. I wonder if this means all of webrtc is broken or not.

Regressing "stable" features like these is creating serious problems for end users. I wish they'd focus on keeping the ship afloat instead of continuing to chase the new shiny. I really don't like relying on google (or any other ad/surveillance shop) for my web browser.

Hangouts doesn't use WebRTC, it uses a NPAPI plugin on Firefox and nonstandard APIs on Chrome.

>hangouts is broken

Hangouts is broken because Hangouts uses NPAPI, not WebRTC

Hangouts will be fixed whenever Google actually ships the WebRTC version of Hangouts

Here's Google's official post about the Hangouts issue: https://gsuiteupdates.googleblog.com/2017/02/google-hangouts...

And the HN discussion:

https://news.ycombinator.com/item?id=13763759

Firefox for Android recently stopped showing me the "Send to Firefox" option in the share menu. I previously used it all the time to send tabs to my desktop. Does anyone know what happened or how to get it back?

It's present on mine under the paper plane icon: https://pageshot.net/tXxzW2A7XAhVuAGN/mail.google.com

I'm pretty sure I've seen that paper plane icon before, but it doesn't seem to be there now either.

The one I always used had a Firefox icon.

Same here, I have a Firefox icon in the share menu called "Send to other devices".

If it has been removed, a comment below you quoted the following which would likely explain it?

> Enhanced Sync to allow users to send and open tabs from one device to another.

It could be related, but that comes from the desktop section. It doesn't say anything about Enhanced Synk in the Android section.

Although, if you go back to the 51.0 release notes, it does say "Improved reliability of browser data sync".

https://www.mozilla.org/en-US/firefox/android/51.0/releaseno...

Yea, that looks similar. I'm not in the beta channel, though, I'm on 51.0.3.

I am signed in and have synced recently.

I think it is the same issue. Changing my clock makes the option re-appear, just as described in the issue comments.

I'm not sure about Firefox itself, but in all other apps, I know it's now under "Add to Firefox".

Because the share list is sorted alphabetically, a bunch of apps started doing "Add to..." rather than use their own names.

We're working on a fix, it should re-appear at some point soon (hopefully!).

Thanks!

Update: it's fixed now. Was the same issue as https://bugzilla.mozilla.org/show_bug.cgi?id=1342881

WASM, Grid, deprecation of NPAPI and SHA-1 certs, all in one release. That is huge. A very important release!

Doesn't mention if the major problem I have with Firefox 51 is fixed. It makes a new, unauthenticated request every time I do View Source rather than showing the source of the page I'm looking at, which is an extreme PITA.

If View Source works properly, that'd be worth upgrading for immediately.

That is annoying, but does the Inspector in the Developer Tools show what you're looking for?

but doesn't it show something completely different, ie. DOM after JS ran?

As far as I know the way to get the unmodified source in the Developer tools is to find the html request in the Network tab, click on it, go to the Response tab of the small panel on the right and read the source out of said small panel (it's about 65x11 chars here). I'd much, much rather be able to just hit Ctrl-U and read it in a new tab. If there's an easier way then I'd like to know about it.

For me it suffices, though I too would prefer if "View source" just showed this same content. Maybe an extension could hijack Ctrl-U and display fetched source?

The difference is that the Inspector shows the page exactly as it exists now, including any DOM changes that have been done via JavaScript. I typically use 'View Source' when I want to see the original page content before anything had been altered.

It looks like the the original commenter is talking about was first reported 12 years ago[1], so I wouldn't count on it being resolved any time soon. It just doesn't seem to be something that many web browser users actually care about, so I don't blame the FF team for not assigning much priority to changing the current behavior, even though I'd also like View Source to work the way way Joeboy is asking for.

[1] https://bugzilla.mozilla.org/show_bug.cgi?id=307089

I think it's sporadically refixed and rebroken.

This is the way it should be in every browser, actually. View source (as in the raw non-corrected source) can take up a considerable amount of space since it's only ever needed when someone is viewing the source. Otherwise the "source" gets transformed internally to be compliant with what the browser can understand and render.

Unless you were talking about some other kind of view source?

The problem I have right now is that when I view the source of a page I'm working on, I get the source of the login page instead. I don't think that is justifiable behaviour.

#spaproblems

It's not a spa, these are #websitewithloginproblems.

That is surprising, that would imply that view-source isn't using the same cookies as the regular browsing experience.

Which seems like a bug.

This seems strange. If you view page source on this comments page while logged in to Hacker News, does the source show you as not logged in? When I do that, my source contains the link to my profile (user?id=Forlien) at the top so I know it's giving me the logged-in version.

Actually no, it shows me as logged in (I can see the logout link).

Whereas with the site I'm working on locally, I get the login page. curl with appropriate options copied from firefox gets me the full page.

I don't know why I get different results with the two sites.

You can open the dev tools, copy the request as cURL and run that in another terminal?

Yes that works, although it generates an option that my curl doesn't like (--1.0) so I have to edit that out (or find the option to change it in FF). Better than the strategy I suggested, but still a lot clunkier than Ctrl-U. Grumble.

Considering the potential size of pages and that this would mean the browser would have to keep around all of the unmodified source for all of those, I think it's a decently justified position.

I will say that it would be nice if there was a way for them to recognize when the devtools are open and use that as a signal to keep around the source of the current page. But otherwise it would certainly cause significant memory bloat for each tab open.

Yeah but couldn't they stick some 'yes, please cache the source, I am debugging' option in about:config?

Then it only takes up space when someone needs it.

Does it not work to select all, right click and "view selection source"?

I think that gets the post-js version of the page, although I'm not 100% sure.

Yes I think you're right. Although the post-js version would be the page you're looking at, so it makes sense it'd prioritise that over "what html was sent from the server which may not reflect what you're seeing".

This shouldn't be happening. We have explicit code to force view-source to read from the cache instead of hitting the server...

If you can reproduce this problem in a clean profile, please file a bug, cc "bzbarsky", and I'll take a look. Obviously, steps to reproduce would he helpful too.

If it doesn't happen in a clean profile, then we start looking into what's special about your existing profile.

Please cc me too. (mconley at mozilla dot com)

Ok now I have to do some embarrassing backtracking. In a clean profile I don't get the problem. Sorry for publicly whining about your product, which apparently works fine.

I'm not sure what I've done to make it not work here, but this profile dates from a long time ago so maybe it's just cruft or an extension behaving oddly. I'll switch to using a clean profile. Sorry again and thanks for your attention.

All good. Thanks for checking. :)

This is a nice way to refresh your profile without losing all your history and other stuff: https://support.mozilla.org/t5/Procedures-to-diagnose-and-fi...

Thank you for checking that!

If you still have the profile where it doesn't work, and are willing to do some detective work to figure out what it is that doesn't make it work (e.g. extensions or prefs.js or something else), that would be awesome. You're hardly the only Firefox user with a somewhat old profile, and it's possible that there's something specific going wrong that we should be handling better on our end...

Even better would be to have the possibility to edit said source from cache and see the result as was possible with opera until it kicked out its founder to become a google chrome skin and focus on quarterly profits.

And so the age of WebAssembly begins...

What I ask myself, is this really the end of JavaScript?

WASM reminds me of the NDK on Android. A way to write high performance GC-less code, but nobody started to use it for other things than high performance computations.

Will people really start to port runtime environments to WASM so they can run Ruby, Python and Java in browsers?

Or will this mainly be a place for C/C++, Rust and co to get some niche high end stuff running?

Or a bigger question, will it be a great space to show that Rust is a great language that can scale up and down through the entire stack? There is one language and toolset which can be used to do everything.

Utopia? Yeah. I know people will still choose their favorite languages for task x, y and z, but I can still dream.

Definitely not the end of JavaScript. But it is the end of JavaScript-only.

It's likely we'll see other languages targeting WebAssembly in the next few years. I heard rumors Microsoft is working on something in that area. If I were to guess, I'd expect a TypeScript-to-WASM and C#-to-WASM to follow. Java, Rust, Go, and others will likely have something too.

Additionally, I expect to see some JS-heavy frameworks (React, Angular, etc.) write performance-critical sections with WASM, thus benefiting a great number of web apps.

Rust already has something: https://hackernoon.com/compiling-rust-to-webassembly-guide-4...

Rust on WASM is probably trivial. But C#, Java and Go are GC languages, porting them is probably a much heavier venture.

Where did you hear these rumors? I could see a WebAssembly target for MSVC in the near future (say hello to Office in the browser) but the CoreCLR needs GC so it's currently out of the question.

I heard this directly from a higher up at MS during a conference afterparty. He said he couldn't wait for browser support for Web Assembly to go live so he can show the things they're working on.

And yeah, for C# to WASM, we'd need garbage collection support in WASM. That doesn't exist yet, though it's on the list of future enhancements[0].

[0]: https://hacks.mozilla.org/2017/02/where-is-webassembly-now-a...

IMHO: No, I don't think WASM is the end of Javascript, in the best case both flavours can focus on what they do best (Javascript for light-weight web-frontend scripting and DOM manipulation, WebAssembly for computation modules). And Javascript can focus on being a 'scripting language' again, and not a compile target. I think most serious WebAssembly apps will still use Javascript glue code for the actual HTML/CSS frontend.

I also think there will be many web apps which are 50/50 JS/WebAssembly where WebAssembly is used for libraries that need to do fairly heavy computations, but can do that without having to call into browsers APIs too much (e.g. image processing, physics engines, AI, etc...).

Android NDK is heavily used for games, I wouldn't say that 'nobody started using it'.

I'm not seeing the appeal for running languages on top of WebAssembly that need a heavy runtime environment. The big downloads this requires will kill all the fun. Even C++ will create bloated WASM modules if one isn't careful. Using 'embedded style' programming practices really makes sense when trying to reduce module size for asm.js/WebAssembly.

> Android NDK is heavily used for games, I wouldn't say that 'nobody started using it'.

Which is very much the only use case that Google's supports, and Android APIs that are exposed to it, for anything else welcome to JNI.

True, and isn't that a shame :/

Isn't JNI on Android also "NDK"? NDK provides a compiler toolchain and a build system — if you're building native libraries to JNI into, that's also built by NDK.

Internally: yes. But if all you need is the C-APIs provided by the NDK you don't need to deal with JNI (or Java). Unfortunately the NDK doesn't provide much.

Yes, but I imagine you never had the "pleasure" to use it.

Didn't have any problems either. I mean, writing JNI code by hand is hell, sure. Just… don't do it. javacpp automates all that crap.

https://github.com/myfreeweb/freepass/blob/fad2a1fd592cbe6a9...

The point is calling Framework APIs from C++.

Of course there are workarounds like Swig, Djini, JNICpp, but those are just that, workarounds for the lack of support from the Android team.

iOS and UWP devs support C++ use via Objective-C++ and C++/CX (to be eventually replaced by C++/WinRT), even if C++ isn't the main platform language, instead of leaving the community alone to sort out the FFI issues.

Since Android is anyway a fork of Java, they could either provide C++ Frameworks that wrap Android APIs or provide something like gcj CNI.

I was under the impression that mobile Qt and Xamarin were practically native apps with a thin Java display layer.

Nope. Two remarkably different approaches.

QT has its own widget system on top of low level rendering.

Xamarin wraps native OS widget API's so it's not write once run anywhere but more of a porting solution.

See this: https://www.linkedin.com/pulse/how-pick-right-cross-platform...

Well on Android's case it helps that the Android team is not very keen in allowing the NDK to be anything more than porting native code into Android, high performance games, or just implementing Java native methods.

Even for Android Things they gave up on their plan for C++ Frameworks on Brillo, and adopted the Android Frameworks instead, due to "We incorporated the feedback from Project Brillo to include familiar tools such as Android Studio, the Android Software Development Kit (SDK)".

So to come back to WASM, it depends on how willing the browser vendors are to push it, beyond the initial prototype.

Generally, it seems that the success and failure of a language depends on libraries, and most android libraries are written around Java rather than native.

But if you want, you can write libraries which will allow you to write Android Apps in C, C++, C#, Rust or Go (as people have already done, see Qt and Xamarin for example). The thing is that it's much easier to write a simple CRUD app in Java than in C++.

Because of how Google handles the NDK experience.

Using C++ on iOS or UWP isn't as painful experience as on Android.

No, compiled languages have never ended the existence of Python, Perl and other scripting languages.

Similar principles will surely apply. For WASM code to interact with DOM or external scripts / JS libs in any meaningful manner, the wasm generator will have to provide in any event additional JS code anyway.

No, but for a different reason.

Right now relying solely on WebAssembly where you can use JS is dead in the water (how many websites still keep up IE support).

The question will be in a few years (especially once WebAssembly DOM support comes in).

Even then, there's so much JS in the wild that only a depracative step (JS is no longer supported) will kill it.

It's like Clojure, Kotlin, Groovy and other JVM languages. Even if they are better by a long shot than Java, the legacy factor will keep JS (and Java) alive for a long time (and the fact that JS doesn't require a compiler or environmental).

If you're using Tree Style Tab and this upgrade causes new tabs to appear after a delay of many seconds... setting browser.tabs.animate to false resolved it for me (macOS 10.12).

Mozilla says 'Enabled multi-process Firefox for Windows users with touch screens', however on my Dell XPS 13 it still says multiprocess windows are 'Disabled by accessibility tools' in about:support, which was the touch screen issue. Any ideas?

Force enable it in about:config, browser.tabs.remote.force-enable i think

Are you running a 32-bit Firefox or 64-bit? I _think_ the touchscreen thing was only fixed for 32-bit for the Firefox 52 release, and the fix for 64-bit will be in Firefox 54.

I'll see about getting the release notes fixed to reflect that.

I'd love to see some webassymbly benchmarks now with 52 stable.

Most important to me from this release: webdriver finally supports self-signed certificates.

Nice. Maybe it's finally time to upgrade to the latest Firefox for testing. We had been pegged at an older version because we ran into some issues with the new (at the time) MarionetteDriver[1], but I imagine the migration might be smoother now.

[1] https://developer.mozilla.org/en-US/docs/Mozilla/QA/Marionet...

Can't wait for servo to be shipped so that everyone can move back again to Firefox.

Despite of my media.gmp-provider.enabled setting being set to false Firefox makes a connection to ciscobinary.openh264.org on port 80 and downloads the h264 decoder binary provided by Cisco.

Does this mean that the library is being downloaded (and executed) using plain HTTP without any authentication? Sounds like a nice target for QUANTUM style MITM attacks.

I'm still rockin Firefox 33 (released 2014) on one of my PCs. Sad to observe websites which fail and throw errors without any grace including big sites like Netflix and others.

Strange errors and broken functionality usually without any mention of my browser being the problem. That's sloppy development. A browser from 2014 isn't "Netscape 3".

Signed up to Google Firebase recently. Tried navigating to the console on my PC with Firefox 33, "there was an error completing your request". That's because I was using a browser from 2014. No mention of that in the error. At first I thought the service was down, then I tried a newer browser.

Backwards compatibility or at least graceful degradation and error handling is accessibility. Your site either has it or it doesn't.

Why are you using such an old browser? Part of the reason sites are built for current browsers is that responsible users should be updating their browsers frequently.

With as many ways as you can have your machine compromised by simply browsing to a website that had some malicious advertising code embedded, I can't imagine why anybody would use such an old browser if they weren't absolutely forced to.

That misses the point. Security is one thing, features are another. Regardless of whether a user should update regularly for security patches, the Internet should support 3-year-old technologies, or gracefully fail otherwise.

For most people it is because the (desired) security updates come bundled with (undesired) UI changes, default search engine changes, increased telemetry, etc.

I work with a visually impaired PhD student, and we had to turn off browser updates because of the hell it was causing him (eg why are my search results different all of a sudden? why are pdfs opening differently?).

This isn't limited to firefox, google voice recently got a UI update that lowered the information density. Where he used to be able to see who he was talking to now there is a big colored shape surrounded by whitespace, etc. If he didn't limit the updates, probably 25-50% of his time would be spent dealing with interface changes.

Nurturing inflexibility to change in a user—especially a student—is setting them up for failure in a world of frequent and sudden change.

Well maybe companies that create this stuff should be more flexible in accommodating the users.

It's not our responsibility to work around someone Else's updates. If something works now, why should it be changed because it's more convenient for Google?

Anecdote: I had a friend in college. Visually impaired - tunnel vision and a degrading retina. Takes him a good few seconds to locate things on familiar UI. New UI is hell. I know because I was by his side all through college.

Your comment makes it look as if you never worked with someone who suffers due to needless design changes.

I work in high school education. I've dealt with inflexibility from staff and students alike. When I started we never ran software updates precisely because people didn't like change.

I changed that policy and moved to tracking upstream as closely as I can manage. While it was very hard at first, users have adapted and become better at adapting to new changes quickly. Five years later, I do not see nearly as many "this changed and I don't know how to do it anymore"-type questions as I used to. The initial break was quite hard, but I'm very glad I did it.

There's a difference between people being just plain stubborn, and people with an actual disability that hampers them.

Indeed. I have worked in assistive technology consulting, and it is very difficult to keep up with all of the "upgrades".

It is clear that we value shiny bullshit more than being accessable and inclusive.

And nobody can even make the case that cutting off more and more people from the web even benefits us.

Technology should serve its users, not the other way around and change-for-changes sake is an extremely bad strategy.

I agree. However, the reality is that software changes constantly. We can accept change and move on with life, or fight change forever.

Not everybody is in a position to 'accept and move on with life', their lives are challenging enough without constant change.

The fact that us techies tend to downplay those challenges does not make them any less real for others.

Why should anyone accept planned obsolescence?

"change-for-changes sake is an extremely bad strategy."

Yup. It's also the go-to strategy for startups and tech companies in general once they have run out of real ideas for their product and one of the quickest ways tech companies' products degrade into shit.

When was it decided that it was the users responsibility to use the latest version of a web browser (regardless of how inconvenient it is for the user)?

Oh right, Google and the cabal that runs the web decided so.

Just because something is "old" doesn't mean it is obsolete. This mentality is very wrong, and needs to be opposed. It isn't helping the users, it is helping the platforms.

Up until a few years ago, it was the responsibility of the developer to gracefully fall back and support as many users as possible. Now, instead, we design under the assumption that everyone has a computer made within the last 12-18 months, with the latest updates, on a high speed, low latency connection.

And that is awful, because it is cutting out users who aren't privileged enough to afford the web.

The new "improved" Google Voice is utter trash. It makes my 2011 Macbook grind to a halt. And for what? So that it cosmetically resembles their new phone app. And most tech reporting websites are gushing about how it looks.

And what is the justification for all of this garbage? Is the web faster? more reliable? More accessible to people with disabilities? NO.

It's not the user's responsibility to update their browser. The updates happen automatically unless a user intentionally disables it.

The point they were making is that users are now burdened with the responsibility of ensuring their computers are compatible with the web, rather than the web ensuring their sites are compatible with users.

I use Firefox 29 for most of my profiles. Because I have a very specific workflow with add-ons, that runs circles around any Chrome user.

It's also more secure, because of my external sandboxing.

I also have my own container mechanism for Firefox that I have been using for years.

And this is what the web is all about. It used to be that the web was accessible to people in remote parts of the world, using old hardware, and anyone could write and add to it.

Now it's quickly becoming a black box that only wealthy people can access.

That's quite the hyperbole.

The performance gains that come with newer browsers (generally) should allow websites to run on older hardware. Not to mention support for new protocols and compression that lead to smaller assets being sent to all devices.

I can see how my comment wasn't clear. The fact that disadvantaged people are being driven off the web has nothing to do with the fact that I have a highly customized browser setup. I probably should have made that more clear.

It seems as if your paradigm shifts it from "only wealthy people can access" to "only people technically competent enough to set up a sandbox can access"

No, I wasn't advocating that anyone use my advanced browser setup. I was just illustrating that there are reasons why someone would want to do so.

I really shouldn't have to do this. Mozilla should have put more effort into add-on compatibility. Add-ons are the only thing that Firefox has going for it.

Curious why are you still using a 3-year old version on your personal computer?

Couple of reasons: I test my site builds on older browsers - proper full browser versions and OS environments/hardware. Some of my computers I keep up to date, not all.

And I don't particularly respect updates sold as "security" when they often move the furniture around, or remove features, or flog whatever new feature they think we want in our browsers.

Web browsers should be more stable, not frantically updating themselves every 5 minutes. Browsers are windows to online content - the online content is what changes and updates all the time. The browser, as viewer of those changing and varied websites, should be relatively static and stable.

I'm glad that major sites don't support older browsers.

This way users will have to upgrade, and the web can move forward.

One thing is to keep supporting something that costs hundreds of dollars (like an iPhone), but there is very little sense IMHO to spend a lot of effort and—potentially—make everyone's experience worse just because some users won't bother downloading a 50MB free update.

A lot of things happen in 3 years in this field!

Google was smart in making Chrome updates automatic. I doubt the average user cares the slightest anyway.

Stop using vague language like "moving forward".

That is exactly the same justification for "deprecating" America's public transit infrastructure in the 1950s, and replacing it with highways. I don't consider a lot of updates to be upgrades.

And the cost associated with updates isn't 50Mb. The cost is losing essential features that I rely on.

So, you're saying that I should lose the ability to log into my router, all so you can use the latest template bullshit that will be abandoned in a year.

And it's not even that a major feature will be deprecated specifically, it's the threat and uncertainty that the constant update cycle creates.

I deeply miss the good old days, when I could trust that my computer would behave as I expected it to. The fact that things have become so damn unreliable is itself taking away from the UX, and more than offsets any improvements. Google has an awful UX, because of how unreliable it is.

All the financial websites that I use over the last two years have been ruined in the pursuit of UX. Information density has plummeted. I used to be able to copy and paste data on screen in a predictable manor. The browser responded instantly. Now, it's a messy, slow, and far less readable.

I'm thankful that I don't rely on accessible technology.

The direction the web has taken recently is horrible. And I'm convinced that this time will be looked back on with disgust.

I was a web designer from the late 90s to early 2000s. I vividly remember everyone saying how Flash needed to be pushed on users, and how HTML was obsolete. And that unskippable flash intro pages were cutting edge UX. And playing sounds in the background was a great idea. And that all users were 800x600 resolution.

The attitudes have not changed. I think they've actually gotten worse.

Amen

> Stop using vague language like "moving forward".

Sorry, I don't have time to go into the specifics of all the APIs that modern browsers have started supporting since 2014, but they're a lot.

> That is exactly the same justification for "deprecating" America's public transit infrastructure in the 1950s, and replacing it with highways. I don't consider a lot of updates to be upgrades.

Why are you comparing public transportation with web standards? Web standards evolve about 100 times faster than public transportation.

> So, you're saying that I should lose the ability to log into my router, all so you can use the latest template bullshit that will be abandoned in a year.

I don't see why you'd lose it. More modern browsers will still be able to interpret old code (http://info.cern.ch/)! The problem is the other way around: old browsers can't understand new code. Also, although I take it that you don't like updating your software, if something doesn't work if you updated your router firmware or software you might find out that it starts working again. You can also keep an older browser installed if you need to access some specific software, but I don't think it's a good idea to use an obsolete browser as your main browser, if anything for security concerns.

> he direction the web has taken recently is horrible. And I'm convinced that this time will be looked back on with disgust. I was a web designer from the late 90s to early 2000s. I vividly remember everyone saying how Flash needed to be pushed on users, and how HTML was obsolete. And that unskippable flash intro pages were cutting edge UX. And playing sounds in the background was a great idea. And that all users were 800x600 resolution.

Then, I find it hard to understand how you believe that things have gotten worse, since all browsers follow the standards now, most things are open source and build on open technologies.

> The attitudes have not changed. I think they've actually gotten worse.

Judging by your tone and attitude, I actually agree with you on this one.

First of all I'm sorry for my forceful language. I was in a rush and needed to get lunch

> Sorry, I don't have time to go into the specifics of all the APIs that modern browsers have started supporting since 2014, but they're a lot.

> Why are you comparing public transportation with web standards? Web standards evolve about 100 times faster than public transportation.

Web browsers and public transit are both infrastructure that everyone relies on. And we have a responsibility to think carefully before we make big changes that can harm, or even inconvenience people. And I think that too often, we make the wrong choices for the wrong reasons. We take for granted that things like ATM machines, or door knobs will remain pretty much the same. The are infrastructure, too.

And the language used to describe these changes are often misleading. Slashing the number of buses that service a neighborhood is an "optimization". Or like downgrading a proposed light rail line to a bus rapid transit, and then finally to a regular bus line. And then selling it to the public as some sort of improvement.

I strongly dislike when people use the word "progress", to imply some sort of change is an improvement. Donald Trump bulldozing low income housing in order to replace them with gaudy luxury condos would be called "progress". I do not think that much of the changes that we have experienced could be called "progress".

I don't think that updates are "upgrades". And often times this language is used deceptively, to hide value being subtracted, and features being removed.

> I don't see why you'd lose it. More modern browsers will still be able to interpret old code (http://info.cern.ch/)! The problem is the other way around: old browsers can't understand new code. Also, although I take it that you don't like updating your software, if something doesn't work if you updated your router firmware or software you might find out that it starts working again. You can also keep an older browser installed if you need to access some specific software, but I don't think it's a good idea to use an obsolete browser as your main browser, if anything for security concerns.

Um, if I update my software, I might find that features are removed or radically altered. Like when Cisco decided to completely change the operating system on their routers, to one which had far fewer features, and would phone home to them[1]. There are too many examples to list, but, just off the top of my head, most Android updates remove useful features. iOS is so awful, that Apple goes through amazing lengths to force users to update. Do you really think they're doing it out of concern for the user experience? They're doing it because users are less likely to complain if you don't give them a choice.

> Then, I find it hard to understand how you believe that things have gotten worse, since all browsers follow the standards now, most things are open source and build on open technologies.

Yes, I am happy about that. However, websites themselves are as awful as ever, in new ways. We've taken everything bad about Flash, and baked it into html5. Just as many pigheaded assumptions are made about the end user's computer setup as in the late 90s. We're actually regressing in a lot of ways. I've actually seen popups when I try to load a webpage, telling me that I need to have higher resolution. That is geocities crap.

> Judging by your tone and attitude, I actually agree with you on this one.

I guess we have different visions of what the web should be. And your vision is winning.

[1] https://tech.slashdot.org/story/12/06/29/1425210/cisco-pushi...

Apology accepted :-)

It's hard to have a discussion like this, though. You're comparing software to door nobs, trains, Trump bulldozing houses...

We probably just do have different views. That's totally fine, too.

I remember having these conversations with web designers so many times about Flash. I can't tell you how many times I was told that HTML4 was "obsolete", and why can't we just do everything in Flash already???

So what if Flash's bitmapped text isn't parsable, why would anyone want to parse it? And so what if it isn't scalable, everyone should be at the same resolution as what we designed it on! It's the users fault for not supporting our ActiveX controls! What kind of a weirdo doesn't use Internet Explorer??? Users like it when we play soothing background music for them. What? Who uses tabs anyway? Users like informative popups that offer them helpful offers from our advertising partners! Hah, Mozilla is obsolete. LOOK, this is just the way things are, OK? They can't be changed!

And the starry-eyed optimism that /this/ technological ecosystem was great, and utter certainty that this was the correct path. And how I was some sort of weirdo for saying that things could be done way better.

However, unlike 15 years ago, there is no Mozilla Firefox. The Mozilla foundation has been corrupted. If things had been the way they are back then, culturally, then things never would have changed. Ubuntu would have been obsessed with emulating Windows XP perfectly, and Firefox would have been trying to support the latest Microsoft standards.

The direction the web has taken recently is horrible. And I'm convinced that this time will be looked back on with disgust.

Fear not, my friend. If we accept the premise that everyone should be forced to switch software every few weeks to retain access to useful content, even if doing so means losing access to other content in the process, then much of this time won't be looked back on at all.

Luckily, that's not how it is.

The thing is, I think that is exactly how it is.

For example, I'm pretty sure multiple decades of little visual aids developed by academic tutors as Java applets just became inaccessible to almost everyone. There was nothing wrong with many of those demonstrations in the content they presented, nor is there anyone responsible for replacing them. That valuable content has effectively now been lost.

There are discussions going on about Firefox 52 today on every relevant technical forum I follow online, and all of them (including this one) have threads showing a variety of other useful content or services that relied on these plugins.

There is no reasonable argument that this was necessary because maintaining support was too difficult. The same browsers eliminating NPAPI support are the ones adding things like WebAssembly.

There is no reasonable argument that this was necessary for security. Most of the plugins were click-to-play in all major browsers already, dramatically reducing the risk associated with them, and more importantly, there is no evidence to suggest that browsers won't suffer from a similar level of security vulnerabilities as they add ever more native functionality to replace what the plugins were used for before.

I realise I'm posting on HN, where the culture and interests are often more about building throwaway web apps and worrying about longevity later, but most of the Web and most of the world isn't like that. There is a price to pay when support for standards that have been around for 20+ years is dropped.

> I deeply miss the good old days, when I could trust that my computer would behave as I expected it to.

So don't update your browser. Or at the very least, use an ESR.

> So, you're saying that I should lose the ability to log into my router

You're using the wrong tool to connect to your router. The web has always been in flux. If your router only supports web login and requires mildly fancy web code which might get deprecated, that's a problem with it, not the browser.

Similarly, it's not the browser's fault that google has bad UX, and you can make sparse-looking pages with ye olde html as well.

You're conflating sooooooooo many orthogonal things here.

    So, you're saying that I should lose the ability to log into my router, all so you can use the latest template bullshit that will be abandoned in a year.

The problem here isn't that browsers are getting better. The problem is that you paid for a router that evidently uses some broken and unsupported web interface. Not blaming you; that kind of thing's happened to me too. I'm guessing it either relies upon proprietary junk (Flash, Java) or nonstandard HTML/JS, or it does something terribly insecure that modern browsers don't allow.

Non-broken HTML from the dawn of the web onward works fine in modern browsers. Had your router manufacturer done something sensible in the first place, we wouldn't be having this discussion.

    And it's not even that a major feature will be deprecated specifically, it's the threat and uncertainty that the constant update cycle creates. [...] The direction the web has taken recently is horrible.

Actually I totally agree! I think the web is worse than it was 10 years ago.

But, this is due to bad site design, not the fact that browsers are more capable now. You could argue that today's more-capable browsers enable and encourage this crap, which is kind of true. But in the old days site owners did the same horrible crap; they just used Flash and Java and ActiveX to do it.

    And I'm convinced that this time will be looked back on with disgust.

I hope so, because that'll mean it's gotten better.

    I vividly remember everyone saying how Flash needed to be pushed on users, and how HTML was obsolete. And that unskippable flash intro pages were cutting edge UX. And playing sounds in the background was a great idea. And that all users were 800x600 resolution.

Right. And those were all terrible and/or proprietary ideas, all of them contrary to the ideas (originally) underpinning the web. Those ideas were rightfully tossed into the trash bin of history. Mostly.

But, I don't see what harmful assumptions and proprietary closed technology have to do with one of the jewels of the open-source software world (Firefox) continuing to improve itself while remaining open.

> The problem here isn't that browsers are getting better. The problem is that you paid for a router that evidently uses some broken and unsupported web interface. Not blaming you; that kind of thing's happened to me too. I'm guessing it either relies upon proprietary junk (Flash, Java) or nonstandard HTML/JS, or it does something terribly insecure that modern browsers don't allow.

If a router manufacturer used completely correct and standard SSL and Java implementations in the past, it would likely be inaccessible today. A reasonable assumption should be made that if you implement something correctly today, that it will be accessible by a web browser in the future.

I understand that these technologies have been found to be insecure. But some people still need them. And it is possible, and not even that hard, to run Applets securely. I do it. They can be sandboxed externally.

I'm actually thinking of putting my "distro" of Firefox on github, for running applets. Appletviewer is not enough.

These features are not removed for our benefit. They were removed under pressure from commercial entities. For example, Facebook pushed hard to kill flash with extreme prejudice two years ago [1]. If it had been up to them, everyone's Flash would have stopped working back then, and been very hard to re-enable. Facebook didn't do this out of concern for safety, they did it because it financially benefited them, and shrank their support window.

[1] https://tech.slashdot.org/story/15/07/14/0112229/facebooks-n...

> One thing is to keep supporting something that costs hundreds of dollars (like an iPhone), but there is very little sense IMHO to spend a lot of effort and—potentially—make everyone's experience worse just because some users won't bother downloading a 50MB free update.

My browsing experience has been made significantly worse than it was 5 years ago by websites abusing new browser features. I have had to start running uMatrix to disable JavaScript except on whitelisted websites, and have had to start using youtube-dl to download videos instead of viewing them in the browser. Firefox Reader View is a must because of bad CSS everywhere. I use Firefox for most browsing but keep Chrome around just to view problem websites.

Most websites serve up text, the only reason they "need" new features is for serving intrusive ads. I also regularly use Lynx and think it continues to be a good browsing experience. Most websites can be (and historically are/have been) very usable and quick to navigate in text browsers. I continue to design personal webpages to work in text browsers.

Taking it as far as saying that they need new features just to serve ads, wow.

The fact that web technologies have become powerful enough to make using complicated software a click away instead of forcing you to install a native app for every little thing..?

Take Twitter as an example - using older versions of Firefox.

They claim that "videos are not supported in your browser" but on occasion the video begins playing before being paused. Sometimes the "what version is your browser" check takes long enough for the entire video to finish playing. The videos are obviously supported because they work! I know this because the same video that tells me "it isn't supported" will sometimes partially or completely play (so it isn't that different videos are being served differently).

So why is Twitter telling me they don't work?

Sorry, what's your point in relation to what we're talking about..?

>Taking it as far as saying that they need new features just to serve ads, wow.

I'm agreeing with the GP comment - and that forcing updates on users that don't actually need to happen is for tracking/ad related purposes.

Few sites need the new features and many are abused for tracking/ads - which is largely why the Battery API was removed. It wasn't being used for its intended purposes.

I use the latest version of Lynx as a fallback on my OpenBSD laptop, but I don't expect to be able to watch Netflix on it. It's really good at fetching files from ftp sites that are hosted on http links, though.

Sure, but if you browse to Netflix on Lynx, you shouldn't be greeted with a dogs breakfast, you should be treated with respect as any user should, and informed of what you need to operate the site.

You might consider switching to the Firefox ESR (Extended Support Release) channel. It updates to a new major version every ~10 months (seven Firefox releases) but continues to receive security patches every six weeks in the meantime.

https://www.mozilla.org/en-US/firefox/organizations/faq/

Disappointing that a modern browser can only be considered secure for 6 weeks.

I can think of many more interesting things other than debugging a three year old evergreen browser.

Staring at a blank wall, for example.

Firefox went to an evergreen browser model and keeps an ESR around for people that don't want to upgrade. But that's at version 45 at the moment.

Firefox 33 is a three year old point release intended for eight weeks of operation, so things being broken shouldn't be a surprise. So sites have the choice of feature-sniffing and keeping an updated list of all the rendering bugs in all browsers going back ... what, a decade?; or kicking out people using old non-ESR browsers. They could be more gentle about it, but the choice is fairly obvious.

Firefox 33 was an official release. Things being broken should be a surprise in an HTML5 compliant browser.

If sites use bleeding edge features with no programmed fallback or error handling apart from "miserable failure", they have a responsibility to at least inform users of their limited browser support rather than pretending to be a website on the internet - which by its nature should be accessible to an official release browser from 2014 without failing miserably.

I'm guessing the reason you don't want to upgrade is because the new worse search bar that was included in Firefox 34? You can install a new version of Firefox and use Classic theme restorer[0] to get both the old UI and the new features. At least until Firefox 57 ships in November which will only support web extensions. It will be sorely missed.

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

> Removed support for Netscape Plugin API (NPAPI) plugins other than Flash. Silverlight, Java, Acrobat and the like are no longer supported.

That's really, really bad. At this point Firefox was the only browser which was supporting Java on Linux. I am forced to stick to the 51 version (52 64bit ESR does not support Java also).

You may ask, why on earth do you need Java inside your browser in 2017? My company, which does not support Linux, uses Java applet to create system-wide VPN connections to the client's networks and Firefox was the only choice.

It's frustrating that multi-process is still not enabled on my touchscreen windows laptop. The about:support page says "Disabled by accessibility tools". Feel like I've been waiting for this feature to arrive for years.

You can force-enable it if you want to see for yourself whether it interferes with your Accessibility settings. https://wiki.mozilla.org/Electrolysis#Force_Enable

Multi-process support for Windows touchscreens should be coming in Firefox 53, according to the release notes. 53 will enter the Firefox Beta channel this week or next.

https://www.mozilla.org/en-US/firefox/53.0a2/releasenotes/

Since "Enabled multi-process Firefox for Windows users with touch screens" is in the Firefox 52 release notes, it should be enabled now. I don't understand how else that note could be interpreted.

I'm the developer responsible for multiprocess on touch screens. There was a mixup in the release notes.

This is what the realities of the situation are:

Touchscreen Windows users running 32-bit Firefox 52 should have multiprocess.

Touchscreen Windows users running 64-bit Firefox will not have multiprocess until version 54.

Thanks for the clarification. And thank you for the work.

Are you running a 64-bit build of Firefox, by any chance?

Yes. 32-bit is too unstable. 64-bit seems to be a bit better but still becomes unbearably slow after a couple days of usage.

That's why. 64-bit touch multiprocess isn't ready until v54.

Here's a more hacker oriented write-up at https://hacks.mozilla.org/2017/03/firefox-52-introducing-web... (published today). Discussion (if it comes) at https://news.ycombinator.com/item?id=13812145

How is Firefox Android working for everyone? I can't seem to get video player working on some website like bloomberg.com, otherwise I'm a fan of it.

I avoid videos on mobile like the plague anyway (because of shitty data plan), so I can't comment on that, but otherwise I'm a happy user of Firefox for Android, esp. because it has uBlock Origin available.

I use as my main browser too and I do not have specific problems with it. The greatest advantage is being able to use the same extensions as the desktop version.

Can you? I miss NoScript... :-/

EDIT: looks like it's available, installing... Thanks!

Another fan here. There are some rough spots and there doesn't seem to be much done to fix them. I have already cloned the source but haven't found the time to build it yet.

My main gripe is that it doesn't treat closing the window as "exit", which means you must go to menu and select Exit if you want to remove session cookies and similar.

Another one is that selecting text is clumsy because standard Android lens is not used.

But I still prefer FF to its "all-your-data-are-belong-to-us" competition.

actually a whole session is the phone turning on and off. So I like this session behavior better. Otherwise unlike apps I have to login again and again to some websites.

You may need to add a user-agent spoofing extension. Certain sites will not serve plain H.264 video to Firefox correctly, unless you pretend to be Safari mobile.

It's been good for me! I don't miss Chrome at all. Plus the "send tab to device" is pretty sweet (if not exactly new or original).

What happened to TLS 1.3 which was supposed to land in Firefox 52? (At least according to https://threatpost.com/mozilla-turning-tls-1-3-on-by-default...)

Not that I'm ungrateful; Firefox is my preferred browser.

It is annoying that I have to install pulseaudio if I want to continue using Firefox for media consumption.

I have one question: when I hit "." in twitter.com will the whole thing hang whilst it crunches through the updates?

I moved from chrome to firefox and so far this is my one complaint. This and I believe the CPU usage is higher for the same number of tabs open.

Why does Mozilla make it hard to run the sync server (easily) on our own? Just give a docker file. Instead you need a sync server, your own account server.. such poor technical design. These are small features that will make many stick to Firefox.

I self-host my own weave server which is significantly easier to set up and run. They haven't dropped support for that yet.

Try connecting to it with a new profile on a recent Firefox without jumping through hoops.

Yeah, I know the process isn't exactly straightforward.

If you are currently hosting a Sync 1.1 server I strongly encourage you to migrate to 1.5, we removed the legacy parts of Sync in https://bugzilla.mozilla.org/show_bug.cgi?id=1296767 which means it will not work anymore on Firefox 54.

I'll investigate that, thanks. However, the last time I checked I'd have to set up and run a separate account server service as well and since weave still works I had no reason to invest the effort into switching.

Regardless, I'll likely be switching to FF 52 ESR so the point will be moot.

"This is an all-in-one package for running a self-hosted Firefox Sync server. It bundles the "tokenserver" project for authentication and the "syncstorage" project for storage, produce a single stand-alone webapp."

https://github.com/mozilla-services/syncserver

This doesn't contain the accounts server, only the sync: https://docs.services.mozilla.com/howtos/run-sync-1.5.html

Thank you for pointing that out, I don't use the sync feature because I don't like giving too much info to 3rd party services but I never considered running it on one of my servers.

I've just skimmed through this guide: https://docs.services.mozilla.com/howtos/run-sync-1.5.html#h...

It doesn't look too difficult and it says that running your own account server is optional. Or does it mean that you can run your own sync server using a mozilla account? I'm not sure I understand how sync server and accounts server interact exactly...

This is exactly the problem. They have just made it hard for no reason. I want concrete information as well with an easy to run guide. It's not clear why I need to use Mozilla accounts.

> for no reason

What's the reason that they should accommodate you and write you a dockerfile?

You can run a sync server that accepts tokens from Mozilla, but I'm not sure it makes any sense that way. The only difference is that you'll have some audit over sync blob store access logs, but that's about it.

Did they bump the minimum Android version required?

I do not see the release via the Play store app, but the web site shows it.

For a project that seems to bill itself as being for the people, they seems awful quick at dropping backwards compatibility.

What are the supported operating systems, I don't see any list of supported OS's ? i did the upgrade and it just crashed on startup (on windows 7 enterprise edition), Installing from scratch didn't help either.

https://www.mozilla.org/en-US/firefox/52.0/system-requiremen... has the list. Windows 7 is supported...

woohoo CSS Grid! It's happening!

I hope this update doesn't break Vimperator again.

The removal of XUL addons is scheduled for Firefox 57. https://blog.mozilla.org/addons/2017/02/16/the-road-to-firef...

Vimperator is gonna break for good by the end of the year. It's been encouraging that the VimFX developer has been so involved in the Web Extension discussions about keyboard input [1]. Unfortunately, the last few updates don't look super promising [2].

1. https://bugzilla.mozilla.org/show_bug.cgi?id=1215061

2. https://github.com/akhodakivskiy/VimFx/issues/860

Oh gosh :(

VimFx is enough for me, maybe it's for you too. https://addons.mozilla.org/en-US/firefox/addon/vimfx/

Thanks! I hadn't seen it, but Vimperator was what finally got me to switch from Chrome/Vimium. If it's customizable enough that I can remap `d` to close a tab and `u` to reopen, I'll be happy... but then I guess I don't need to be happy, as I won't have a choice soon enough :)

Oh, I feel you. I had a hard time remapping Vimperator "Ctrl-u/d" to VimFX "u/d" and Vimperator "u/d" to VimFX "x/X" in my muscle memory.

On the other hand, some parts of VimFX's language are more logical, e.g. "yf" instead of ";y" to select a link into the clipboard.

What's the subset of Vimperator's features supported by VimFx?

I am for example quite reliant on .vimperatorrc for custom commands and set! about:config preferences.

Still no hardware accelerated video on Linux. Everytime Mozilla adds new features, I am reminded of this:

https://xkcd.com/619/

Does Pentadactyl still work in this release?

What percent of Google Voice users will abandon Firefox now that v52 breaks a key component of their workflow?

https://gsuiteupdates.googleblog.com/2017/02/google-hangouts...

Ironically, Mozilla has stated that breaking all these user plugins is part of some strategy to regain market share and "influence web standards". Now that it's actually happening, I'm hoping more people will see how it's not going to work.

I didn't even realize google voice had hangouts integration in the browser.

Why is Google relying on an NPAPI plugin for Hangouts, rather than using WebRTC?

This is the old version of hangouts, so I guess a) because WebRTC didn't exist/was too new at the time, and b) hats

Hangouts switched from NPAPI to WebRTC in Chrome back in 2014, but still uses plugins (NPAPI or ActiveX) for other browsers. Hangouts depends on some non-standard WebRTC functionality in Chrome. Google has had almost three years to adapt to other browsers' WebRTC stacks, but it is not a priority because the plugins have still worked (in Firefox < 52 and IE).

https://plus.google.com/103171586947853434456/posts/39TCW3Pc...

Except Mozilla had planned and announced this would happen at least a year in advance. On top of that Google also removed the same thing from Chrome a while ago citing the same reasons, it's just that they shoehorned in a fix for Chrome and IE when they did so.

(a) None of the replies changes the fact that Firefox will lose users over this. A harbinger of what's to come. Too bad Mozilla's wack leadership can't solve their problems by downvoting the users who leave Firefox because they will no longer the browser that works with a rich extension ecosystem (in general).

(b) If the one of the largest software companies out there, which is apparently acting in good faith here (since they intend to release a new plugin), can't keep up, then in what universe do all the rest?

All that being said, who knows what it going on with the new Google Voice. While seemingly improving the UI, they made it very difficult to place outgoing calls to people one has not received calls from already. So maybe Google is getting ready to kill Voice off.

a. It seems you're conflating the deprecation of XUL/legacy extensions and the NPAPI plugins here. This current release removes NPAPI support which affects plugins only. Extension changes are coming in later releases.

b. If they're acting in good faith why did they misrepresent when Mozilla announced they would be deprecating the NPAPI plugin API[1]? It seems more likely that they were simply lazy and instead of making the solution they used for Chrome (which already removed NPAPI support, mind you) work across the board, they waited until the last moment and realized Mozilla was actually going to go through with making the same changes they did.

[1] https://blog.mozilla.org/futurereleases/2015/10/08/npapi-plu... which is a full year before Google mentioned Mozilla made the announcement ( https://www.fxsitecompat.com/en-CA/docs/2016/plug-in-support... )

Hangouts switched from NPAPI to WebRTC in Chrome back in 2014, but continues to use plugins (NPAPI or ActiveX) in other browsers. Hangouts depends on some non-standard WebRTC functionality in Chrome.

https://plus.google.com/103171586947853434456/posts/39TCW3Pc...

Yes, I think I am conflating the issues of XUL/legacy extensions and the NPAPI plugins.

All I can say is that while I didn't take a lot of time to parse this, most of the people impacted by this issue will take even less. They will just understand that Firefox worked with Google Voice and then stopped.

So it does seem to me like part of an overall theme. Firefox used to let users decide what was safe, now it's decided for us.

> Ironically, Mozilla has stated that breaking all these user plugins is part of some strategy to regain market share and "influence web standards". Now that it's actually happening, I'm hoping more people will see how it's not going to work.

I never thought I would see a defense of NPAPI, of all things, on HN. NPAPI is a gaping security problem, and the only fix is to remove it entirely.

What is your proposal for making NPAPI secure?

Many people think this is a Google problem, not a Mozilla one:

https://news.ycombinator.com/item?id=13763759

Reply on news.ycombinator.com