Hacker News new | past | comments | ask | show | jobs | submit login
Mozilla launches voice search, file-sharing and note-taking tools for Firefox (techcrunch.com)
386 points by denchikceo on Aug 1, 2017 | hide | past | web | favorite | 204 comments

Some more information in the official blog post: https://blog.mozilla.org/blog/2017/08/01/new-test-pilot-expe...

I was especially interested in the Voice Fill (speech recognition) technology. Landing page: https://testpilot.firefox.com/experiments/voice-fill

It seems the project is here: https://github.com/mozilla/speaktome/

This seems as if it actually is a webservice. From the code (https://github.com/mozilla/speaktome/blob/master/extension/c...), I see: const STT_SERVER_URL = "https://speaktome.services.mozilla.com";

Actually, I think this can be very easily done fully client-side, with good accuracy. Even on Android, the voice recognition can run client-side / offline.

I wonder if the project is in any way related to their DeepSpeech project (https://github.com/mozilla/DeepSpeech). Maybe they use DeepSpeech on the server-side? At some other place they call it Pipsqueak, not sure if this is yet something else.

And maybe also related is their common voice project (https://voice.mozilla.org/). Recent discussion here on HN (https://news.ycombinator.com/item?id=14794654).

Some more information also here: https://research.mozilla.org/machine-learning/

"this can be very easily done fully client-side" : maybe, if you have the voice model and and inference engine that runs well on devices. Mozilla doesn't have that yet, so this experiment uses a backend running a Kaldi server and model that uses too much memory to run locally.

Once DeepSpeech is ready I'm pretty sure they will switch to that, and ultimately to on-device voice recognition with PipSqueak (PipSqueak is expected to be an inference engine usable on devices). Unfortunately none of these projects are far along enough to be usable.

Common Voice is mostly related to DeepSpeech as this will help getting data to train the engine.

Here is the backend: https://github.com/mozilla/speech-proxy (which talks to Kaldi via I think https://github.com/api-ai/asr-server)

> Actually, I think this can be very easily done fully client-side, with good accuracy. Even on Android, the voice recognition can run client-side / offline.

I'm not sure I'd say it's easy; you will certainly trade off accuracy versus a state-of-the-art server model. Among other things, Firefox users are not going to download gigabytes of recognition model, so it'd have to be a lot smaller than the server ones would be.

Very possibly it will be slower too, since the servers would most likely be using GPUs for at least parts of the recognition, but it might not be easy to ensure the same on all the millions of PCs Firefox runs on.

It's interesting if Mozilla is running their own speech recognition system. I wonder whether it would actually be usable in practice... The problem is, I couldn't find any kind of online demo.

Browsers should be browsers, not a way to get a whole bunch of trojan horse software installed on your machine.

If I wanted voice search, file sharing and/or note taking I would find the appropriate piece of software for that task and install it separately.

All this does is increasing the attack surface of the browser in a ridiculous manner providing features that only very few will use. File-sharing? What could possibly go wrong?

And in the meantime Thunderbird, a stand-alone program that had a decent following got the axe in favor of more browser bloat.

I really no longer understand Mozilla's mission.

The title is misleading. These are not part of Firefox. They are in the Test Pilot program, which is a way for Mozilla to gather feedback on what features they should include, how they should change them if they do include them and all in all is completely open-ended. They might also release these as extensions or somesuch. They are all coded as WebExtensions (except for that file-sharing thing, that's just a webpage), so that's very well a possibility.

Also, browsers already have code to record from your microphone, due to the WebRTC web standard, so it's not like this opens an entire new class of potential vulnerabilities.

>The title is misleading. These are not part of Firefox.

Doesn't matter. They shouldn't be part of Mozilla's output either...

> If I wanted voice search, file sharing and/or note taking I would find the appropriate piece of software for that task and install it separately.

The good thing is that these are all separate browser extensions, none of which are bundled by default. So that's exactly the case.

For now. And then because not enough people install them they will be bundled by default. This has happened over and over, for instance with sync, pocket and so on.

The whole idea behind test pilot stuff is to trial stuff intended for future inclusion in the main distribution, getting it force-fed to be able to claim a large installed base is the logical next step.

You say that, but I've heard nothing about Tab Center (the Test Pilot vertical-tab sidebar) being folded into newer versions of Firefox. Luckily, a community member forked the old code and ported it to be a WebExtension so it still works with Firefox Nightly, but it really needs to be built-in to the browser for maximum polish.

There is the perfectly functional tree style tab extension. That's another waste of resources.

Really, Mozilla resource allocation is a mystery to me.

They are trying to move away from the old extension API (because maintaining that ate up a crap ton of resources, so don't come up with that move being a waste of resources) and Tree-Style Tabs is not yet compatible with this new API.

The Vertical Tabs test pilot thing was them demonstrating that it could be done (including to themselves, they need to test these APIs) and giving the community something to work with, should no other tree tab extensions crop up before the deprecation of the old API.

>The Vertical Tabs test pilot thing was [...] to test these APIs

Except that Tab Center wasn't a way for them to test the new WebExt APIs, as it was still based on XUL. See here: https://discourse.mozilla-community.org/t/whats-all-this-abo...

>maintaining [the old extension API] ate up a crap ton of resources, so don't come up with that move being a waste of resources

Arguably not a waste if that meant preserving the most powerful extension environment out there, and also one of the very few reasons for people to stay with Firefox.

Signed: a Firefox user

The old extensions were not compatible with e10s.

If the Firefox team wanted to improve speed/security/modularity, etc. they had to ditch the old extension model.

It sucks, but necessary.

"Not compatible" is just another way of saying "we don't have the resources and/or the inclination to support your programs". Look at what Microsoft does to ensure backward compatibility. They simulate specific windows bugs for specific old games that were coded to expect the original bug in Win95. Just so users' programs keep working. It's a matter of priorities and resources. And I believe both are out of whack at Mozilla right now.

>The old extensions were not compatible with e10s.

Addons based on older APIs could be made compatible with the newer multiprocess design (see https://developer.mozilla.org/en-US/Add-ons/Working_with_mul... ) and for a while Mozilla tried to persuade developers to update theirs (see https://blog.mozilla.org/addons/2016/08/02/multi-process-fir... )

It was only not long after that when Mozilla's fresh direction decided to start pushing the WebExtensions bandwagon only while dropping everything else, in spite of the thousands of man-hours invested by the addon developers (many of which, you can understand, were not happy; see http://fasezero.com/lastnotice.html )

>they had to ditch the old extension model

To me, that's more like a political decision they made at an arbitrary point in time.

EXACTLY. This is just like that little IM Client 'ello' or whatever that came bundled in the last version. I used it for a few days and said exactly what you did as I uninstalled and went back to a different browser.

I wonder which browser you're using, because there are plenty that come rather close to your definition (e.g. Midori), but odds are really high that you're not using them - because nobody is.

Thing is, if Mozilla wants to stay relevant, it has to find out what users want and deliver that. These experiments are one way to do that.

Nobody uses them because they're broken and slow.

Lynx works and is fast.

I just wish it was .. faster.

Lately I've been moving away from Google everywhere I can. I moved everything but Google Voice. Yes, even Google Search - I've moved to DuckDuckGo. On windows however, I had to fall back to Chrome, because I was just shocked at how slow Firefox was.

Opening pages like Twitch.tv proved to be shockingly slow. Furthermore, my habit of opening many tabs in the background like I do in Chrome/Safari was massively slower in Firefox because while Chrome doesn't autoplay new-hidden tabs, Firefox does - I imagine Chrome feels faster there because it's not running nearly as much stuff at once.

Pretty much everything of Firefox felt slower for me. And this is from someone that really wants to get away from Chrome! On OS X, I've long switched to Safari and DuckDuckGo, and been quite happy. I've had zero complaints about performance with Safari.

So.. I don't know what they need to do, but I'm really hoping they do something.

It's getting there. They recently added multiprocess support. A new CSS engine is coming. In the further future we're going to get more and more pieces of their new rendering engine in.

Personally, I dropped Chrome 2 years ago or so, for the same reasons (moving away from everything Google that I reasonably can). Firefox is more painful to use, for sure, but it's gotten "okay" enough that I'm willing to keep using it in favor of its benefits (privacy).

By the way: In Firefox, open `about:support` and check that "Multiprocess Windows" says something like "2/2 (enabled by Default", where the numbers can really be anything but 0. If it's disabled, that means your Firefox isn't using the new multiprocess support, most likely because you're running an incompatible addon.

Is there a way to figure out which addon or addons are the cause, without disabling them all and enabling them one by one?

Personally I've found the recent Nightly with the new Rust CSS engine to be pretty snappy. I'd say anecdotally faster than Chrome.

There was also a recent post regarding having large numbers of tabs open in Firefox (see https://news.ycombinator.com/item?id=14823807) and it did pretty well.

Yea I'm definitely going to try switching to Nightly based on the comments here. As well as adjusting some of the configuration. Thanks!

Would you comment back here and gives your first impression on it? How fast is it compared to stable.

I have Nightly in one machine (Linux) and the R55 in a Windows machine. It's way fast than the previous versions I used to have and looks definitely snappier than G. chrome. Don't have any benchmarks with me, purely based on perception/ felt response times.

Not the OP, but I switched to Nightly a week ago and unlike stable I would recommend it to anyone. Not just to people who don't want to use anything Google. It's fast and doesn't hog memory. A far better experience than before.

Over the course of the next two weeks, they'll be merging some major UI overhauls, as well as remove swaths of legacy code, so things might be a bit rockier for some time and I wouldn't necessarily install it on my gramma's PC right in this moment, but other than that, yeah, it is pretty crazy how big the difference is, with Nightly being just three months ahead.

Okay, fair point: To everyone who understands that nightly means "unstable at times".

Just downloaded Nightly and it's a much better experience than the beta channel I was using before. Snappier page load, cleaner UI (compact theme).

However it still seems to lack in the animation department. CSS transformations and transitions, particularly if affecting complex elements, just aren't as smooth as in Chrome or Safari. Your own menu animations, like the global drop down menu, are finally fluid and smooth though.

I switched to Nightly based on comments on this thread and it's great.

Much more streamlined UI, and much faster. I already preferred Firefox because of better handling of large numbers of tabs, but Nightly beats Chrome even at smaller number of tabs so far (I synced my FF account so my extensions, etc got reproduced, and I manually replaced legacy ones with newer ones). I still have to see how it holds up over a period of use, but so far with my data synced it looks great.

I just installed the nightly, (and with several privacy extensions).


I have Chrome as reference to my FF here and since their beginning, it never was faster again. As in every FF post, there will be people saying that and the opposite. I guess the difference depends on other factors like Addons or Plugins as well as the probably narrow difference between both.

I never really came upon a reason to leave FF behind. It works good, has all the addons, and I can still make it look like a real window with options and things where they should be. I also have a "special" relationship to Chrome due to this "bundling to freeware" they use to push the browser onto people who don't want it.

@your tab problem: I just googled this up https://addons.mozilla.org/de/firefox/addon/load-tab-on-sele... No idea how good it is though since I never felt I'd need that.

There is a solution/addon for everything ;)

Google does not want you to move from their services and APIs.

Chrome/Chromium, as a recent example allowed serviceURI in Web Speech APIs for third party recognition to be plugged in. Here: https://developer.mozilla.org/en-US/docs/Web/API/SpeechRecog...

It was dropped in Chrome 49 and we're now all stuck with using Google. There has been little coverage on this, some speculated it was hardly implemented, or was dropped because lack of a standard API format.

Whatever the case. Google / Chrome choose not to fully develop that feature and its now gone. Which ultimately works in Googles favor.

I mean, I don't care what Google wants lol. I've moved everything except Chrome on Windows, and Google Voice because of no replacement.

I'm not complaining because I'm locked in by Google, I've switched to Safari on OS X with no issue at all - I don't miss Chrome in the slightest. Yet, on Windows, Firefox is proving to be a hurdle.

It's on Mozilla, not Google. Imo

What does "faster" mean? honest question. I've been using Firefox since its early days and never jumped on the Chrome bandwagon simply because I didn't see the need to do so. When Firefox felt slow, it usually was the whole operating system that's slow and neither Chrome or any other browser could solve that. Nowadays, all the machines I use are powerful enough to not feel any such slowness and I just don't understand what is this "speed" people keep attributing to Chrome and what is so "slow" in Firefox (regardless of that fact they're working on speeding it up).

In addition, Chrome recently crashed on me multiple times while using Google Spreadsheet which is supposed to work better with Chrome than other browsers.

Not OP, but what I think is that FF renders the twitch.tv page very slow. On macOS, not just the tab but the whole FF app locks up, showing a spinning beach ball. It's the only site that I found that does that.

I don't personally visit twitch.tv but I just did and it loaded normally and streaming auto-played without enabling Flash (although the site requested it for some reason).

Speaking of Flash, some sites request Flash on Firefox but not on Chrome for example (although it's totally unnecessary). This is becoming less frequent recently but it shows that a lot of websites are not taking advantage of Firefox's full features as they do with Chrome (something to consider as well).

Option to disable autoplay in background tabs exists in Firefox for so long that I forgot it isn't enabled by default.

Go to about:config and set: media.block-autoplay-until-in-foreground = true

In older versions they used another config variable media.block-play-until-visible .

There is also media.suspend-bkgnd-video.enabled to stop decoding video in background tabs after set amount of time (10s default).

I find FF actually faster than Chrome. But maybe that's because I actually open multiple tabs which load in background, then I just switch between them. Though to be fair, I never knew about this difference... It just felt that Chrome is slower somehow.

I wonder if they are both the same, it's just that us users have learned different workflows based on strengths of each browser?

This has a lot to do with the ever increasing incompetence of web devs than it has to do with Firefox imho. I do performance testing for a living and Firefox creams Chrome on sites where the devs know what they are doing. I don't think of Chrome as faster, but better designed to hide developer mistakes. Chrome basically keeps the mediocre developer who can't rtfm in business just the same way Microsoft did back in the day. Same pointlessness on Android. I call it the Trump definition of success - pander to the lowest common denominator and pretend there isn't a price to be paid.

speaking as a web dev who's obsessed with payload size, TTFB and web app performance, i find your statement to be patently false.

i have well-written code that runs very fast in FF, and it always runs even faster in Chrome. not just JS, but also repaint and layout. FF does handle some absurd cases better: giant dom trees & scrolling, RAM usage, lazy tabs.

as a long-time Firefox bug hunter and nightly user, i hope Firefox/Quantum & Servo can reverse this pattern, for sure.

Firefox stable feels relatively slow. Firefox nightly is really snappy, especially with Stylo enabled.

Also, go to about:config, and make sure preloading is turned on. I checked it last week and for some reason it was turned off, even though I don't recall ever doing that.

> I just wish it was .. faster.

They're working on it, to see the progress, install nightly (which is already a huge improvement) and turn on servo CSS for an extra boost. It feels much faster.

When did you try?

I've been on Firefox developer edition for awhile now, because it's just SO MUCH FASTER than chrome, especially on high tab volume.

seconding Firefox developer, although still have problems when it has been open for a while, evidently because it needs to update.

Try disabling HTTPS Everywhere, if you have it. It made everything many seconds slower for me.

You might as well open it's options and disable SSL Observatory since it's exactly what cause slowdown.

Hmm, interesting, that sounds plausible, thank you.

The auto playing issue is fixed and making its way down towards general release.

Weirdly enough - I found gmail was running VERY slow on Chrome. I switched to the Brave browser and it's been blazing fast.

Weird that Google products don't even seem to run the best on their own products.

Isn't Brave based on libChromium? At least on Linux

Do you like Brave? I've not done too much research on alternate browsers, but at this point as long as I trust the company I'd be willing to try it.

Cliqz is another one I thought about trying, but it is based on Firefox so I'm a bit dismayed haha.

I do!

It has some issues with sites because of cookie storage + I believe it disables scripts - but it's really fast :)

Routinely have similar issues with gmail/chrome whereas I never have any issues on Firefox. Odd.

I think part of the issue here is that it just visually seems slower. In my experience it isn't actually doing things much slower, it's just a little visually clunky in comparison. And as far as resource usage goes, RAM is way down now because their multiprocess implementation is a lot less intensive than Chrome's. I switched for that reason and I'm enjoying it a lot.

Oh c'mon it's not that much slower. Some things feel chunkier (scrolling) but JS performance is 80%-90% of the way there.

Oh so apart from the one thing that needs to be perfect, everything is fine? Sorry but if scrolling is not smooth af you don't have a viable browser.

Someone should dump all y'all in some redneck craphole out in the middle of third-world nowhere for a few years. With that environment you'll find out just how important "perfect scrolling" is when most of the web loads for shit under your spotty 3G connection.

I was born and have lived my whole life in India :/

~as an asside

Its not a real solution, but for twitch, which i agree is essentially unusable, ive replaced it with streamlink + mpv + a couple of scripts to notify me of who i follow is live. Twitch chat works in irc too.

Speaking of speed, or lack thereof on Googke properties, has anyone else noticed how insanely slow the Google Admin interface is?

Pale Moon, been using it for a few years OR Chromium.

It doesn't help that they keep cramming features into Firefox. Chrome feels pretty bare in comparison.

Many of said features are just bundled web extensions, add-ons, so they don't pollute the main code base. Please, don't spread FUD when it is not needed.

Chrome has consisted of significantly (i.e. millions of lines) more code than Firefox has for quite a while now.

I feel like I heard that a Chrome install was comparable in size to an average Linux install a year or two ago.

The features they're 'cramming' into Firefox don't slow down the browser.

I always admired Mozilla for their efforts in making Firefox better. But I switched to Chrome many years ago. For me, the problem with FF was UI and UX.

Chrome and Safari just feel more "smooth and sleek", and UI elements are consistent (Look how Chrome buttons are rounded-rectangles).

The other reason I switched from FF was that Chrome has always been simply better. Sure, FF can handle many tabs, but I save that for times I want to work with Selenium or something. For an ordinary user (and even pros), Chrome just beats FF.

A redesigned UI is coming to Firefox with version 57. Maybe you'll like that one better.

Can't exactly argue with your other point, since you don't explain how Chrome beats Firefox, but as many people in this thread have already said, Firefox is currently picking up a lot of speed. You can try Firefox Nightly, if you want a taste right now. Or wait until Firefox 57, which will hit stable on Nov 14.

Didn't they say not a few months ago that "the reason Firefox has fallen behind is that we focused on stuff that wasn't browser-related like MozillaOS etc."?

Learning from own mistakes anyone?

Quoting from the article:

"These are obviously experimental tools and there’s no guarantee they will ever make it into a release version of Firefox. Indeed, the idea behind Test Pilot is to allow the Firefox team to test new concepts."

The efforts on making Firefox faster and more responsive have been going on for a few years (e10s/electrolysis being one of them) and are appearing in the releases as fast as the team can develop, test and deliver them. I personally believe Mozilla has a highly capable core team on this, and to insinuate otherwise, to me, seems like accepting commercial interests of other browser makers as higher than FLOSS interests. I don't disagree on quality of the product being a huge aspect in value judgments, but it cannot and should not be the only one.

It doesn't matter if they are experimental or not. It doesn't matter if they will make it into the release or not. There is one very certain thing: these experiments divert resources and people.

"Fast and more responsive" is just one aspect. Because "we're getting faster" and "these are just experiments" lead to a situation when "Browse against the machine" [1] uses FF for work and Chrome for play, and pretends it's ok.

[1] https://medium.com/the-official-unofficial-firefox-blog/brow...

More people won't necessarily help solve any problems though. (Insert obligatory Mythical Man Month mumble mumble, hand-waving, etc. here.) Perhaps additional resources could help.

In general I agree that Mozilla has lost its way with its current head count; successfully funding a company with advertising revenue tends to do that.

This is how they are adding features to the browser. This is a smart way of doing it. Make it experimental, gauge the feedback, and integrate it if the feedback makes it seem useful.

As an example, they tried vertical tabs, and it appears it didn't work as well as they needed it to be so it has been left out. On the flip side, the empty tab page was also an experiment, and that worked well and made it to the default browser.

These experiments are work on the browser. But instead of adding them as features, introducing them as experiments and then adding the polished, stable, feedback adjusted version as a true feature that can be marketed effectively.

Vertical tabs used to work so well on Firefox that I held on for just that, even when everybody had already left for Chrome. Then they killed it. Now you're saying they dropped it because they couldn't get it to work again. Kind of sad.


Vertical tabs have never been built into Firefox.

Unsure what you mean by "they dropped it because they couldn't get it to work again" -- it was never built into Firefox.

I have a plugin for vertical tabs. Works great. Wouldn't want to use my 28" 4k monitor without it.

>I personally believe Mozilla has a highly capable core team on this, and to insinuate otherwise, to me, seems like accepting commercial interests of other browser makers as higher than FLOSS interests.

There's only one interest in the browser wars: prevail or perish. If Firefox cannot secure a large enough market share, it will be insignificant, lose funding, and die off.

Seriously. They're behind in speed, security, quality of the developer tools etc. and introduce new useless toy features and one re-design after the other.

I appreciate a lot of stuff Mozilla do outside the browser for the open internet, but the development of Firefox is pretty bad and it looks like the project leaders are incompetent.

Firefox is a sad example of how a successful software can be overtaken by a new competitor within only a few years due to a lack of focus.

>due to a lack of focus.

And due to their competitors ability to advertise on the majority of people's home page. For most people, chrome wasn't a firefox alternative, it was an internet explorer alternative.

I'd like to know how much of an impact that had, but nonetheless Chrome is technically in many areas the best browser today and Mozilla, being the clear market leader at one point, could have done a lot more to prevent that.

Firefox was in need of technical improvements for some time, but they only started acting when Chrome already had a noticeable market share and that's not how you stay at the top.

I would buy that if FF would be technically the best or most secure browser. But for a very long it hasn't been.

For every 100 experiment maybe one will be fruitful.

No one is criticising Mozilla for coming up with Rust, yet if they didn't take the time to experiment Rust wouldn't be around.

So yeah, good on them, I'm happy for them to try voice recognition and other endeavours, who knows what they might discover?

Rust is a technical improvement, so it's exactly what many users are asking.

>who knows what they might discover?

They'll discover, just like Opera, that an abundance of useless features doesn't make up for a browser that is lacking speed, security, a clear design, ease-of-use and configurability.

Rust, at least, had a clear design goal.

> speed, security, quality of the developer tools

Coincidentally, all of the examples you mention have been a major focus for far more people within Mozilla than their experiments are, and improvements in these areas have really been noticeable if you've been using Firefox lately.

Yes, but these changes take time and they only started working on them when users started leaving for Chrome.

Like many others, I used Firefox for a long time despite it being technically inferior to Chrome, but at some point you say 'fuck it' and make the change.

Mozilla, given their market share, should have been first to offer 64-bit support, separate processes for tabs, sand-boxing and a lot of other stuff, but they've been asleep, because for a long time there was no strong competitor.

Those are accurate comments about Mozilla in the past, but you can't say development currently is bad when they're working exactly on what you say they should be working on (and some other things).


I've been waiting for working and full printing support something like https://www.smashingmagazine.com/2015/01/designing-for-print... for years now. Printing articles from the browser is the same pain as 10 years ago.

But yes this is not as fancy as voice search or file sharing (or their single sign on, identity and all the other failed stuff).

Mozilla has been a vehicle for people who play out their tech and political dreams and then move on. And FF is crumbling all the way.

But this misery is soon over, we're in the end game of Firefox.


(Written from FF - a browser I've used since Mosaic)

The linked page seems to indicate Chrome is leading because people are browsing on their Android phones. So it would seem if Mozilla should concentrate on anything it would be improving the mobile experience.

That is actually the one area where Firefox is already better for me, because of extension support. Having an ad blocker on mobile devices is priceless.

I use FF only because of extensions right now.

Sadly Mozilla seem to be trying to kill off development of Firefox on Android...

Mozilla should focus on building the browser instead of building app that competes with other developers. They need to start listening to the developers community and stop being arrogant and ignorant. You build the foundation, and developers build the app.

That's a weird take. Mozilla has a wide variety of developers, working on all sorts of things. One of those things is experiments to figure out what the browser "foundation" should actually include.

I like that they're experimenting to figure out what good usability looks like and what the next generation of "browser" should deliver. Some things should just be there when you want them...not an app you install separately or a website you go to for the functionality. I don't know if all of these new things fit that category (voice stuff probably yes, notes probably yes, large file sharing maybe?), but I know I want Mozilla to keep trying new things.

Also, if Mozillians aren't building user-facing things with the "foundation", they don't actually know what the foundation should look like for developers. You have to be a developer using the platform daily in order to understand what it does right and wrong. I don't know how they could do that without actually building things that sit atop the platform.

I just found it damaging to their vision and brand. Take Hello and Persona as example. User focused? Yes. Failure? Absolutely. They need to be developers focused. Less and less developers are testing their apps with Firefox as the primary platform.

That's becuase of their declining marketshare (which is mostly due to strategic disadvantages in distribution channels and marketing budget) not their brand or split focus.

Budget has always been their disadvantages, yet in the pre-Chrome days, they were core product focused and won many developers support. Their vision was once inspiring and admirable. This was how they won over IE. However, ever since Chrome was introduced, most of their innovations are either catch up or gimmicky features - not to mention the political turmoil/non sense. I really wish they can focus on their core product again.

They got kind of lucky too. Their main competition abandoned development of their browser for about 4 years. Now, there are four browsers in active development each regularly communicating with web developers.

Given their pick it's obvious web developers would rather focus on the browsers with the most share of users.

>large file sharing maybe?

From what I can tell their new file sharing feature is actually just a mozilla hosted web app.

see: https://send.firefox.com/

There was a complaint in this thread that it isn't peer-to-peer, which is...kinda weird. Requiring that your recipient have the device they want to receive it on running at this very moment is kinda a non-starter for a file sharing tool, IMHO. Doing that without it being a Mozilla-hosted app isn't really a thing I'd care about. P2P is neat, but it's not a very good user experience, especially for non-technical users.

A lot of companies built the platform, and populate it with some of their own apps to set a minimum bar of quality, like with Apple (Reminders) or Google (Keep). One might say the same discussion applies to language creators -- bake it into the language, or have 3rd parties do it? Not always an obvious discussion, and so perhaps the statements about ignorance and arrogance come too quickly.

Some things are too important to not at least put out a standard-bearer.

Perhaps I was being critical, but it sadden me to see Mozilla is making one missteps after another. I think they are missing the point. Instead of competing with other browser vendors for developers, they choose to go the other way.

Sorry, but I will not use any speech recognition service until it becomes a pure on-device service.

I hope Firefox succeeds and destroys Chrome.

After the 54 release, it is the better browser frankly, low latency, and not a resource hog. Plugins, extensions, privacy and security defaults all seem to be better as well.

Using firefox focus on android is pretty great too. The normal firefox on android is just slow, not sure why.

because its webview which is just chromium. They don't use the gecko engine like in desktop firefox

I believe Firefox Focus (on Android and iOS) uses the OS native web view (effectively, Chrome or Safari), but the actual Firefox for Android uses Gecko.

^ This seems right from my quick googling, there is this project for a gecko webview, which needs work: https://wiki.mozilla.org/Mobile/GeckoView and mentions firefox focus' gecko build option.

Firefox on android is too slow to use a lot of times. Something mozilla should be able to do better.

Mozilla just don't care about Firefox on Android, the naive observer would be forgiven for thinking they're actively sabotaging it.

It won't happen unless Google decides to abandon Chrome.

Mozilla really needs to outdo themselves just to regain the market share they lost to Chrome. Compared to IE, Firefox was clearly the better browser, but it took many years just to become significant.

Slightly off-topic, but does anyone use a Vim plugin for Firefox? I keep trying to switch, but the scrolling of VimFX drives me crazy. Here's an issue outlining the behavior: https://github.com/akhodakivskiy/VimFx/issues/830

Are there any other Vim plugins without this issue?

Vimium-FF <https://addons.mozilla.org/en-US/firefox/addon/vimium-ff/> is the only Vim-keyboard-binding plugin I know that will work with Firefox 57 and above. I haven't noticed any scroll jankiness as you describe, but it's possible I have my key-repeat-delay set low enough that it doesn't affect me.

Gave this a try, and it works! No weird jump after the first scroll.

Also, for anyone using 1Password, they have a beta of a multiprocess compatible FireFox extension that seems to work fine.


I use VimFx on Firefox, I guess that scrolling issue doesn't bother me because I hadn't noticed it until I visited the link you shared.

I've started looking for VimFx alternatives since the following doesn't sound great:



I'm still using Pentadactyl, which is incredible, but it's almost certainly doomed. You'll need to use the signed version on github: https://github.com/willsALMANJ/pentadactyl-signed You might consider Vimperator, which is similar, but inferior iny opinion. It has better support, though, and is more likely to survive into the future.

If you're looking for something simple, VimFX works, but it's too basic for me. Not sure of the future of VimFX.

Can't say it has ever bothered me, but the future of Vim extensions on Firefox once XUL in extensions is phased out for good is dire. Right now, there is a port of Vimium (which I suspect gets my FF 55 to hang). I don't think any of the other extensions are likely to make it.

I use keysnail for the emacs key bindings but I think it has a vim mode too

I find it hard to believe that speed is really the issue here. I suspect it's more likely it's the aggressive advertising scheme in which Google push for Chrome.

I think these projects which Mozilla test/launch are an indication that even they don't know how to stop the eroding of Firefox market share.

If I were Mozilla I'd be seriously worried about Electron.

Speed is definitely an issue on the mobile app. Personally I think Mozilla should try harder with the mobile market. Providing an ad and tracking blocking browser is still desirable on Android. Firefox is just too slow at the moment.

Doesn't Firefox focus fill that gap? https://www.mozilla.org/en-US/firefox/focus/

Super speedy as it uses an Android Webview and privacy focused.

I'd forgotten about Firefox mobile. I agree, it's very slow.

I really hope Mozilla take Servo and prop it up as an Electron competitor, as it might actually make a chance there and its incomplete compatibility with hacks used in the old days of the web are not that much of a problem there.

Isn't Electron based on Chromium? Also, why would they be worried about it?

Chrome cannot control, for example VS Code.

Because developers will code apps around Electron -> Chromium. It'll get more focus because of that.

I wonder what happens to the previous Test Pilot experiments.

I loved the experimental home page and the Tab Center [1]. I really hope it continues to live.

[1] https://github.com/bwinton/TabCenter

Tab Center Redux is the official successor to Tab Center: https://addons.mozilla.org/en-US/firefox/addon/tab-center-re...

It's now a WebExtension, which means it will work in Firefox 57+ (and possibly eventually Chrome, if they end up supporting the same APIs), and uses the new Sidebar APIs that the new Notes experiment also makes use of.

If that doesn't fit your use cases, there are also quite a few new tab sidebar WebExtensions popping up that aim to replicate existing more sophisticated tab management extensions like Tree Styles Tabs (https://addons.mozilla.org/en-US/firefox/addon/tree-tabs/), and others that try to innovate on their own terms, such as Sea Containers (https://addons.mozilla.org/en-GB/firefox/addon/sea-container...), which makes use of the new Containers experiment.

Overall there is a huge amount of very promising new development in the WebExtensions space, and this is exactly the kind of innovation I had hoped WebExtensions would encourage.

On the main https://testpilot.firefox.com/ page, scroll down and click "view past experiments". Each one gets a write-up when it's over.

Hey, we're a little behind on writing up the last few experiments unfortunately. We'll be adding full reports over the next few weeks.

If with experimental home page you're referring to Activity Stream [1], something similar cropped up in Firefox Nightly around a week ago.

Looks like this: http://i.imgur.com/nhGpHYt.png

Each section can be hidden individually, so if you just want the search box or just the websites you frequently visit or don't want stupid articles recommended on your newtab-page, those are all things you can do.

Seems like they took out the recent activity thing, but no idea if this is finalized already.

[1]: https://testpilot.firefox.com/experiments/activity-stream

I don't know here Mozilla got their inspiration, but Send looks pretty similar to our E2E encrypted file sharing app Whisply: https://whisp.ly

PS: Whisply even has more features and a detailed description how its encryption works: https://whisp.ly/static/whisplyTechnicalOverview_20151201.pd...

There's many such tools available, RiseUp runs one called Up1. See https://share.riseup.net/

It's quite nice, has integrated image and video viewers as well as a pastebin all end-to-end encrypted.

Up1 looks nice, but is internally using SJCL which is slow and does not support bigger files ("This is not a problem with sub-10MB images"). Send - and Whisply - are built on the new WebCrypto APIs which are faster and allow bigger files, up to 1 GB in both cases.

I've seen all the promo videos, and some questions remain unanswered:

1. What's Send encrypting with? It doesn't show anything related to setting up a password during the video.

2. What's Notes syncing with?

3. What's Voice using to analyze the voice commands?

>What's Send encrypting with? It doesn't show anything related to setting up a password during the video.

Likely a random string that is added to the download URL as Fragment (the part after # that doesn't get sent to the server).

The download URLs look like this:


I agree that there is very little explanation beyond "look at this cool thing".

2. Notes source code is at https://github.com/mozilla/notes, the panel.js source[1] is the one that saves and load content. It uses the Storage API from Web Extensions[2], but it uses it only to save/load locally. It is not currently sending any data anywhere. So, sync is not enabled. If you click the sync button, you're greeted with a message[3] saying it is not implemented yet (they do count it and metrify it, probably gauging interest).

[1]: https://github.com/mozilla/notes/blob/6bae20ceb58c5f1487885f...

[2]: https://developer.mozilla.org/en-US/Add-ons/WebExtensions/AP...

[3]: http://take.ms/ieu4n

3. Kaldi

Wowsers. The Nightly version of Firefox is FAST.

Somehow, I assumed "Notes" would be "notes for this page". But, it's just a notepad...no awareness of where I am. That's much less interesting/useful than I thought it would be.

Is there a browser plugin out there that provides a notes sidebar for the page you're currently on, and allows flipping through the notes with a link to the pages they relate to? Because that'd be awesome. Googling only reveals some kitschy "post-it on a web page" type note things, which is definitely not what I want.

They've launched these specifically to gather feedback, so do put that into the feedback form. These are in no way finalized and it's definitely possible that they'll take that into account.

I believe the initial inspiration for this was people asking for a place to quickly jot things down that they'll then copy over into other applications, so might be a bit awkward to get those two use-cases under one hood, but I'm sure they'll try, if the demand exists.

Look at Hypothesis, which does something better : it allows you to annotate your web pages.

Note that Hypothes.is notes are public.

Doesn't Edge have this built in?

I don't know. Is there a Linux version?

Opera 12 called and it's rotating in its grave. Unite was such an amazing feature. And notes were available in Opera even before I started using it. Opera, I miss you, it still hurts.

I wish they'd stop focusing on the flavor-of-the-week technology gimmick bullshit and focus on making Firefox the better performing browser. Chrome outperforms it in nearly every way by margins that grow with every release. My reasons for sticking with FF grow fewer every day.

Try developer edition or beta to get a taste of the work they've put into speed. I ditched chrome for FF and haven't looked back because even a year ago, developer edition was so much faster... Especially on heavy tab loads.

Iirc the speed changes can't start landing in stable until November, because of the shift to Webextensions.

Are Mozilla paying for people to do these experiments or is it people sending patches? I don't know but I'd imagine a lot of it is the latter.

Have you tried a recent Nightly? Firefox is getting faster: http://www.techradar.com/news/firefoxs-blazing-speed-with-hu... [1]. As Mozilla continues to pull in code from Servo and implement e10s, Fiefox will perform better and crash less.

[1] discussion: https://news.ycombinator.com/item?id=14848836

> Are Mozilla paying for people to do these experiments or is it people sending patches? I don't know but I'd imagine a lot of it is the latter.

As far as I'm aware, it is mostly developed by Mozilla employees, but heavily used for new devs to get their feet wet. Not only is it not mission-critical when they make mistakes there, they also get a ton of feedback on their work.

Agreed. This plus a number of plugins disappearing in FF57 with the new WebExtensions have forced me to switch to Safari.

On the plus side now, I get my tabs across all my devices.

Voice fill is awesome feature. Kudos to Mozilla :)

Is Mozilla committing publicly to supporting these new features for a significant period of time? I've been slightly wary of integrating new Firefox features into my workflow ever since they cut off support for Firefox Hello in Firefox 49.

No, not at all. These are in the Test Pilot program, which is specifically for trying new things and getting feedback if and how they should be included in Firefox.

So, most of these are not here to stay for the long term.

They might be included, they might be published as an extension either officially or just forked by someone, but overall, no guarantees on anything.

These features are only in the Pikot program...a Mozilla extension.

When is the end of life for these features?

These Test Pilot experiments usually last for a few months, but there's no fixed end date, I don't think. It's mostly just a matter of having gathered enough feedback.

But well, if it seems like a feature that people want, then it gets included into Firefox itself and stays for longer.

Then, again, I for example can't at all imagine them including that file-sending mechanism. I doubt they even have the server capacity for that...

Looks like a good waste of resources to implement things people don't want in a browser anyway, in a desperate attempt to get people to consider using Firefox again.

If I was Mozilla I'd allocate programming time to catch up with Chrome now that they don't have the phone OS thing anymore, instead of creating more distractions.

They really blew it. I remember a few years ago when I used both Firefox and Thunderbird as my default browser/email client...

I feel like I must be reaching the "get off my lawn" phase of my life, because I can't even comprehend a situation where I'd ever want to talk at my browser... maybe when I'm at home alone and, couldn't type for some reason? Certainly not in the office. Am I missing a use case, or am I just old now?

Actually, people who are very deep into the "get off my lawn" phase of their life, that is old people, are generally not the greatest typists. For them, the ability to talk instead of type, everywhere and not just in google search (which already has this) seems like a huge help.

I'll never use it. It's completely useless to me. But if my grandma actually did anything else than Facebook on her computer I'd show this to her.

> that is old people, are generally not the greatest typists.

Well.. and we're just fine with it. Simple works better somethings and why be in such a hurry sometimes. I look at these young whipersnappers who fly around the computer with windows attacking from all directions. They just seem to make things complicated without actually meeting the goal any faster.

Lots of people are faster at speaking than typing. Or maybe you're reading from a document (or pill bottle) and don't want to type with one hand or whatever. But what I usually use it for is if I'm having a conversation with someone and we want to look something up. It feels more anti-social to type silently for a while than to make the computer part of the conversation.

I work in the online learning industry, mainly with higher education institutions. One of the top features required by instructors is the ability to provide audio feedback to students. And this is really a huge benefit for them because most instructors are not fast typists and they have hundreds of assignments to review every once in a while.

Faster than typing would be great. What's not so great (or so fast) is trying 3 or 4 times by voice and having to go type it anyway because the accuracy's not there. Sooo, like my old-aged ancient geriatric friend in the parent, I too still preferentially go to the keyboard, since I figure I'll probably end up there anyway. If it's a social situation I usually preface the typing with some variation of "All right let's see here..." which sounds like I'm trying to figure out something hard or ask something of the Great Oracle, but really it just means "I'm about to type."

Maybe once speech recognition advances a bit more...

But it's only for searching after you've already navigated to a search engine, right? You'd have to be extremely slow at peck-and-hunting for it to save you much time (especially as search strings are so short) - and in the video you have to click "submit" to even do the search (and then presumably navigate normally with mouse). I can't see many instances where this could be worth it. Maybe I'm actually too young to see why it's useful!

Maybe the search engine is your homepage, or you just have to hit "g" and enter to get there, or it's bookmarked. Maybe you don't have hands, but you have an easy way to move a mouse around and click. Maybe you're eating a sandwich with one hand and don't feel like putting it down. Maybe you won't use it for a short query like "weather", but you will for sentence-long queries. Maybe you want to search "linux won't recognize wireless keyboard".

It's very easy to think of instances where this will be helpful.

There are many use cases for this, among them:

* Mobile version of Firefox. It might be easier to talk to fill a form on mobile than type on glass screen.

* Accessibility. There is a huge under served demography of people who will welcome this as typing might be hard for them.

You may not be on any of the groups above but that doesn't invalidate the features.

In fact, mobile Firefox already lets you speak to fill in the address bar (which doesn't automatically launch a search).

"Firefox! I've fallen and I can't get up!"

Just to clarify these features are not yet on the release.

But why?

It's a browser. All it should do (and do well) is display content.

I don't want a bunch of extra garbage tossed in because it sounds good on some marketing slide.

Chrome has been and continues to eat FF's marketshare alive because it has been both faster and lighter. More junk is not going to help FF beat them.

> Chrome has been and continues to eat FF's marketshare alive because it has been both faster and lighter.

I believe Chrome got a larger market share also because it had the giant force of Google marketing it heavily on its properties for years, whereas Firefox only had some support of that sort from the Yahoo search deal (IIRC). More people land on Google's online properties than on Yahoo's.

I personally don't agree that Chrome is faster and lighter. For me (as a heavy multi-tab user) it has always been slow as molasses and a laggy browser.

Firefox is as fast and lighter, and has been for a while (the exception being Googles own webapps that are heavily Chrome optimized). This is stuff they are testing out in TestPilot, so it's for people who want to go add the TestPilot add on and then selectively enable these features.

So your complaint really makes zero sense.

I just want to browse the web!

What's stopping you? Optional experimental plug ins?

what will happen to test pilot after the non-webextension are not allowed?

As far i know must of these feature are implemented as extension and i don't thing webextension have the api to do a lot of thing that are in test pilot.

We're moving to a webExtensiony future. The idea here is that we'll be a first party consumer of the webExtension APIs so we can help test/drive/expand capabilities.

Note that Voice Fill is actually a WebExtension : https://github.com/mozilla/speaktome

They will have to adapt. For now they're using data from the experiments to guide API development. For example, the Tab Center experiment used old APIs, but now they've expended the new API and are building a new Tab Center Redux extension at the same time. https://addons.mozilla.org/en-US/firefox/addon/tab-center-re...

I don’t need any of this in a browser, and if I did, it feels like extension territory. Weren’t Mozilla over this? Here I am, looking forward to Project Quantum technologies.

What would be useful in FF is a content blocker, especially in the mobile version. Not pocket, file sharing apps, voice, notes. There are web apps for that.

Go into about:config and set "privacy.trackingprotection.enabled" to true. This is default-enabled in Private Browsing.

Also, uBlock Origin and similar are perfectly functional extensions. Takes just three clicks to install uBlock Origin on a fresh desktop Firefox installation.

If you're wanting to tell me that this content blocker should be default-enabled, good luck explaining to webpage owners that they won't make ad money off of Firefox anymore, but should still support it.

I already have uBlock Origin on Firefox on the desktop. It shoudn't be enabled by default but a content blocker would be an useful addition to a browser rather than notes, voice, pocket, chat and who knows what lass than useful feature they might think of next.

can't you install umatrix/ublockO in mobile FF?

I was interested in Send until I realised it wasn't p2p.


I only want a sweet WaveNet-based TTS voice. Baidu's DeepVoice seems promising.

I don't know about other users but I use FF less and less. I don't need all this additional stuff. And it was painful how we lost features like the firebug plugin.

It's a pity considering a few years back I was the greatest FF fan.

send.firefox.com says that my browser (Safari) is unsupported. Anyone know what particular "web technology" the site uses that Safari doesn't support?

This is the only one I can think of, not sure why that site would require it though. https://caniuse.com/#feat=input-file-directory

Found it. It's checking for window.crypto.subtle. Looks like Safari TP supports this. I believe the problem with Safari 10 is that it implemented an older version of the web cryptography standard.


This browser is more and more bloated... :-(

These featured come in an optional extension and your comment is fud.

Mozilla is doing everything but what is supposed to do:make Firefox blazing fast.

They actually did, the recent switch to multiprocessing made a massive difference. I've been happily using Vivaldi for a few months now, but I'm tempted to go back to FF as the speed gap has been wiped out of late - only cold startups are a bit slower, actual pageload is absolutely on par or better. This on Mac, at least, but I expect it will have been similar on other platforms.

Last time I tried it still was slow... I'll give it another try.

When was the last time you tried? R55 is way snappier than Chrome. Nightly is also fast.

Yeah, a lot has happened in the past few months. And there's still more things scheduled for the next few versions...

None of these will get me to switch. Chrome is still faster than Firefox, sorry Mozilla.

Firefox is the browser I really want to like, but can't use. I recently tried to switch back, I kept running into problems with webpages not working, figured it must be an add-on, removed them all, still frozen pages persisted. A web browser that chokes on pages frequently is dead in the water for me.

Then there's the lack of compatibility with their mobile browser for extensions, so most of my extensions didn't work on Android tablets. It was definitely faster than it had been the last time I used it, but at this point, it needs to be significantly better than Chrome to win back users and it's not there except for a small group of people.

I do feel that the Voice Fill is a step in the right direction. I'd prefer that the notes application was a dedicated product with an add in that tightly integrated with the browser, a webapp in a wrapper would be fine really.

Registration is open for Startup School 2019. Classes start July 22nd.

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