LWN: Comments on "Firefox 38.0.5"

Firefox 38.0.5

mathstuf — Tue, 16 Jun 2015 14:17:18 +0000

> we recently added DuckDuckGo as a built-in option, and changed the defaults in the US, Canada, Russia and China to engines other than Google

I thank Mozilla for this. However, there is one problem: on Android, the search engine is *always* Yahoo! when launching from the long-press-home and dragging up to search. I haven't seen an about:config for it. Maybe it was just missed? It'd be really nice to use DDG there as well.

Firefox 38.0.5

sdalley — Fri, 12 Jun 2015 21:52:33 +0000

> [0] for lack of proper evidence alone; the idea itself is quite plausible...

Tell me this post is some kind of joke.

Seriously, dude. You need to get that tinfoil hat off before it completely fries your brain.

Firefox 38.0.5

Kluge — Fri, 12 Jun 2015 11:27:56 +0000

I'm not familiar with Pocket, but from what I've read here, it's not clear that it does anything that Zotero (www.zotero.org) doesn't do.

And for someone in a scholarly field, Zotero is the best thing since open-source sliced bread.

Firefox 38.0.5

pizza — Thu, 11 Jun 2015 23:58:36 +0000

> The excuse that they have no choice in the matter is either weak leadership or evidence of outright coercion.

I can't count the number of times I've seen "strong leaders" lead their ship off a cliff in the name of "not showing weakness" -- It takes a far stronger leadership to acknowledge highly uncomfortable reality and take an unpopular, but necessary approach.

Mind you, I'm not saying that Integrating a Pocket client was necessary -- but it's Free Software, and Pocket's API can be re-implemented by some sort of Free Software, assuming of course that enough people care enough to do so rather than just bellyache that someone else isn't doing work for them.

Firefox 38.0.5

ksandstr — Thu, 11 Jun 2015 16:10:21 +0000

>Why is this trojan horse being slipped into a minor point release?

So that it would be ignored, and so that opposition could be met with pooh-poohings as "gerv" does elsewhere in this thread. This strategy is known as "OTDL", i.e. on the down-low, and it sees quite a bit of use in circumstances where personal or political gain can be had from intransparency. You'll note that the answer to Pocket being a proprietary service by a for-profit corporation is "well you don't have to use it"; to Pocket being enabled by default (and thus the user having to opt out by semi-obscure about:config tweaks) it's a redefinition of what "enabled by default" means so as to confound discussion; and to accusations of Mozilla doing this for pecuniary reasons, an absence of even the faintest justification for why the specifics of the Mozilla-Pocket deal aren't being discussed publicly.

Note also Mozilla's endorsement of in-browser DRM as something mandated externally which they don't have any choice over. Umpteen percent of the browser market and they had "no choice" -- that looks like either sincere dicklessness, or conjoined commercial interests disguised as such for plausible deniability. The excuse that they have no choice in the matter is either weak leadership or evidence of outright coercion.

Someone else ITT pointed out Mozilla's ongoing infestation of the social-justice disease in relation to the Pocket affair, and while I disagree that these are directly related[0], the standards of discussion which that cancer supports[1] are ideally suited for hierarchical control by corporate america as well. It'd clear the air quite a bit if we were to know who made the call to include a Pocket client in Firefox, and what their reasoning was; or if, like certain trade agreements, the already-secret details of the contract must remain under non-disclosure until five years post-ratification. Right now there's a lot more activist PR in support of Pocket-in-Firefox than there is transparency in the matter, and as the situation goes on like this, Mozilla's implicit statement in the matter becomes effectively "put up or make your own", clearly a great way to increase Firefox adoption by overboarding the malcontents.

[0] for lack of proper evidence alone; the idea itself is quite plausible
[1] i.e. variations of "you're not allowed to say that", "how dare you", "you're a troll", "the problem is you", "believe or else", et cetera

Firefox 38.0.5

Cyberax — Wed, 10 Jun 2015 19:20:27 +0000

As a customer I very much prefer text. It's actually faster and easier to use.

Firefox 38.0.5

gerv — Wed, 10 Jun 2015 19:17:57 +0000

Advantageous for the business; not necessarily for the customer. But I agree, room for both solutions. Still, I think this use case has value.

I also think that people are yet to think up some of the cooler use cases for the instant ability to be talking to someone else over the web with voice and video.

Gerv

Firefox 38.0.5

zlynx — Wed, 10 Jun 2015 19:04:21 +0000

The HUGE advantage to using text chat over voice and video for tech help is that ONE tech can easily handle 5 different customers AND he doesn't need to speak English with the appropriate local accent as long as he can type it.

I much prefer text chats to voice. I have in fact sat down with someone and opened a notepad window so we could communicate more clearly. It isn't worth the time to work through accents or stuttering.

Firefox 38.0.5

gerv — Wed, 10 Jun 2015 16:44:04 +0000

The team notes that YSlow is currently broken on platforms other than Firefox, and seems not to be all that actively maintained. While forking the entire project is not a very attractive option, we are looking at ways we can bring the benefits YSlow provides to the built-in Firefox Developer Tools. However, the team which works on the tools is currently hard at work making sure they still work when we ship Electrolysis (process separation for Firefox), which is a fairly big internal rearchitecture. So it may be a little while before we get there.

Sorry it's not better news. If you need YSlow, I believe it does work with some not-too-old versions of Firefox.

Gerv

Firefox 38.0.5

pflugstad — Wed, 10 Jun 2015 13:18:10 +0000

I'm willing to buy off on adding audio/video functionality to Firefox as a way to leverage video chat and so on. I see value in that for a large number of users. But I think this thread was derailed a bit by that discussion.

You didn't answer (that I saw?) the specific question about Pocket. It had a working extension - it was NOT a "weekend-project" to include it in Firefox core. Pocket very much feels like something that worked best as an extension - entirely optional - and that also addresses any issue about it being a proprietary service as well. Why go to the effort to include it in the core, and probably what's on most people's minds: was there money involved? If so, that is the slippery slope, IMO.

Firefox 38.0.5

gerv — Wed, 10 Jun 2015 08:47:37 +0000

We don't maintain YSlow; but nevertheless, I've found some references for the problem you cite (I think...) and have passed them to our developer relations team, to see if they are able to get some action taken.

Gerv

Firefox 38.0.5

gerv — Wed, 10 Jun 2015 08:25:30 +0000

I'm not against self-hosting, or getting someone else to host for you. My point is simply that it's not reasonable to expect every Firefox user to do the work of setting up their own hosting, or searching out and making an arrangement with someone to do it for them. Having a service set up and ready to go makes sense from a usable product perspective. After all, we wouldn't ship Firefox with 0 included search engines (to be "balanced") and tell people they had to go and find and set up their favourite one.

I agree that a) the API the Pocket integration uses should be open, b) it should be fixed, and that c) it should be possible to configure Firefox to use a different API endpoint. a) and c) are certainly true; I'm not sure what the position is on b).

Gerv

Firefox 38.0.5

seneca6 — Tue, 09 Jun 2015 20:26:52 +0000

Not sure whether I'm considered to be the original questioner. I'm certainly no expert on Wallabag which I didn't know before reading this thread (as said, "print to PDF" fulfills all my needs), but it looks like a possible alternative which should be allowed to compete on equal grounds inside Firefox. Extension vs. extension, not extension vs. built-in.

There are public instances of Wallabag; the good folks from Framasoft run one. It seems to be at least partially translated in English. Certainly not good enough for the hundreds-of-millions-of-users-immediately criteria, but a start.

I also think that it is a false dichotomy to say: you can either use Google or other for-profits with big data centers, or you self-host (and are thus considered marginal). Even for the self-respecting geek, the Internet gets too complex to self-host everything from mail to voice to collaborative services. Given the ridiculous prices for renting dedicated servers, there is a middle ground that I call "friend computing" - you are hosted by the geek of your confidence. (I do know this does not work for everyone; some prefer being watched by a presumably anonymous NSA or other big-data to _possibly_ being watched by your friend. But it can work for some others.) And, on a larger scale, you have non-profits and communities like Framasoft, running public instances of Free Software services and asking for donations. In the long run, peer-to-peer web applications (where each browser is also a tiny server) might become another alternative.

Firefox 38.0.5

gerv — Tue, 09 Jun 2015 08:18:31 +0000

My intent was not to be condescending. The only other proposed alternative in this discussion on LWN has been wallabag, which describes itself as "a self hostable application for saving web pages." I was asking whether the original questioner's suggestion really was that instead of integrating Pocket, we should tell all Firefox users to buy some hosting and set up a wallabag server. Isn't it incredibly obvious that this is not a suggestion that will work for 99.99% of users?

I assume the original questioner's definition of Free service was that it has a free-as-in-freedom client and a free-as-in-freedom server. If that's the definition, Weave is Free.

Gerv

Firefox 38.0.5

flussence — Mon, 08 Jun 2015 23:19:36 +0000

> What do you mean by "at least one Free alternative"? You mean the one you have to self-host the server for, that 99.999% of people in the world won't be able to use?

If we are to accept your condescending, snarky, rhetorical questions as an answer, then Mozilla's own Weave service isn't Free by that definition either, is it?

Firefox 38.0.5

bronson — Mon, 08 Jun 2015 18:41:04 +0000

Maybe the OP was talking about Mozilla's recent ad buys? Two dudes on phones with some sort of scary privacy implication (something about coat color? don't remember).

I can't find them now but there's a pretty clear SJW feel to them.

(the one time I actually WANT to see an ad on Hulu again and I can't figure out how...)

Firefox 38.0.5

bronson — Mon, 08 Jun 2015 18:29:34 +0000

That's the exact rationale EBay used when buying Skype. :)

Could be... I'll believe it when I see it. For now, to me, it just looks like gluing fun hacks into Firefox because that's easier than fixing bugs or keeping YSlow working. *shrug*

Firefox 38.0.5

gerv — Mon, 08 Jun 2015 18:29:05 +0000

It still lives :-) - http://www.seamonkey-project.org/ .

Gerv

Firefox 38.0.5

bronson — Mon, 08 Jun 2015 18:26:06 +0000

Netscape Communicator will rise again. :)

Firefox 38.0.5

gerv — Mon, 08 Jun 2015 18:01:13 +0000

How does inviting other people to create their own orgs if they want to, equate to a claim to "exclusive rights"?

The world needs Mozilla - and the EFF (and ORG, and La Quadrature, and...).

And if you did a little research on me, at least, you would find out that I don't fit the profile of an SJW one bit.

Gerv

Firefox 38.0.5

b7j0c — Mon, 08 Jun 2015 17:20:51 +0000

Your tone is consistent with so much coming out of Mozilla these days. We apparently "owe you one" because you hold exclusive rights to being the social justice warriors of the web. Mozilla has actually very little in this regard and your "open web" mantras have devolved to become little more than cheap slogans. I now realize my financial donations are better directed at an uncompromised org like the EFF, where they will be sent in the future.

Firefox 38.0.5

raven667 — Mon, 08 Jun 2015 16:07:12 +0000

> why build video chat into the browser?

The browser as an OS need to provide APIs which access hardware, like cameras and microphones, for video chat applications to work over the net without a separate client application. This is the same reason WebGL exists, or the touch API, to provide a hardware API that can be used by platform independent code in the browser OS.

Im not sure what hardware APIs are missing from the browser these days, there is support for local storage, GPU graphics, maybe not GPU computation (so no bitcoin yet), video and audio encoding as well as playback, client/server and peer/peer sockets, maybe not wireless configuration or bluetooth, and there still must be some work to be done on two-factor auth because the standards the browser does support don't seem to be widely enough used.

So to know where the browser is going to go, look at the negative space of where the browser isn't now, then you know where it is going to be.

Firefox 38.0.5

gerv — Mon, 08 Jun 2015 15:55:37 +0000

To start with, the web was about sending text backwards and forwards - HTML pages, and forms. It was 2-way communication - just. Then came the <img> tag and file uploads, and we got the chance to send pictures in both directions - albeit asynchronously, and (in practice) asymmetrically. With WebRTC, the web acquires the ability for any two people on the web to communicate with text, audio and video, in real time. Why audio and video? Because that's how people communicate best synchronously.

One application: have you ever been to a website where it offers "Having trouble? Click here to talk to a customer service person", and you click, and you get a puny little text box to connect to a guy in Nowhereistan who is probably talking to 3 people at once, and replies once every 5 minutes? What if that was an actual no-hassle voice and video connection? Much better and quicker to get your problem solved.

Gerv

Firefox 38.0.5

raven667 — Mon, 08 Jun 2015 15:42:55 +0000

> Is it [firefox] still about being a web browser?

Probably more than anything causing confusion is that the definition of a minimal web browser has increased in scope substantially in the last 10 years on its trajectory of being a standard OS platform, a scope which was predetermined the moment browsers (Netscape Navigator) gained the capability to run code client side. The fact that a browser is an OS platform all on its own is what scared MS so badly in the late '90s, they saw that the future was web applications, like WebRTC, and not native like Skype. It has just taken us a while to get there, to where something like a Chromebook is actually a viable computer, even with the rough edges that exist today the browser will only ever get more comprehensive to the point where traditional desktops and client applications will become a minority market rather than the dominant one.

Well actually mobile phones are the dominant model of computer now, in 20 years there will probably be adults who have never used a traditional desktop OS.

Firefox 38.0.5

bronson — Mon, 08 Jun 2015 15:18:36 +0000

I *know* it's cool. But why is it in Firefox? It doesn't improve my browsing experience at all.

This is my question: why build video chat into the browser? Why not a bitcoin blockchain, or instant messaging, or bittorrent, or todo lists, or email (ha), or Dropbox, or blog publishing, or anything other truly awesome, groundbreaking thing? What is the vision?

Firefox 38.0.5

jezuch — Mon, 08 Jun 2015 09:54:28 +0000

> an open source search engine (is there even one?).

http://www.yacy.net/en/ ?

Firefox 38.0.5

andrel — Mon, 08 Jun 2015 05:39:22 +0000

I had no idea right-click worked on those icons! Thanks for the tip.

Firefox 38.0.5

gerv — Sun, 07 Jun 2015 12:43:49 +0000

Why not try using the service before commenting on the user experience? The string "Save for Later" is not actually used in the UI (hence my "or something like that").

If the Pocket button isn't appearing on the toolbar for you, open about:config and flip browser.pocket.useLocaleList to "false". At the moment, it's only enabled for certain locales, which does not include (e.g.) en-GB.

Gerv

Firefox 38.0.5

gerv — Sun, 07 Jun 2015 12:41:04 +0000

You don't need about:config - https://support.mozilla.org/en-US/kb/disable-pocket-firefox .

Gerv

Firefox 38.0.5

gerv — Sun, 07 Jun 2015 12:20:13 +0000

"This is a strawman. Mozilla has lost its way somewhere between .org and .com. I knew trouble was ahead when they tried endowing an open source project with the headcount and headquarters of a .com"

You think Mozilla would have been more successful at preserving and protecting the open web, and building sufficient market share that people had to take notice of us, if we'd remained an all-volunteer organization?

Having the resources to pay people to spend 40 hours a week achieving your goals, instead of having to take whatever evening and weekend time they can spare, is a wonderful opportunity. (Of course, in order to get people to come, you need to pay salaries that are competitive and have working conditions that match. We already lose too many good people to places that pay better or offer stock options.)

"the project seems hopelessly compromised."

That's a pretty vague accusation. But, if you don't like what we are doing, why not start your own open web-preserving and promoting organization? The more the merrier.

Gerv

Firefox 38.0.5

gerv — Sun, 07 Jun 2015 12:10:34 +0000

A video chat system capable of working at Firefox scale is not a "random weekend project". We now have an alternative to the proprietary Skype client that anyone in the world can use without needing to install additional software, get an account or reveal any information about themselves, it's based on open standards, and it's end-to-end encrypted using standard protocols. I think that's an awesome thing. Why is that not an awesome thing? If it were an addon, lots of the "it just works" advantage would be lost. And if Firefox is going to support WebRTC anyway, which it is, then the added code from Hello in the client is not that big.

Your slippery slope is not very slippery, IMO.

Gerv

Firefox 38.0.5

bronson — Sun, 07 Jun 2015 07:16:53 +0000

> it's the idea of including more and more into the core what used to be outside.

Absolutely right, and I'll say it more directly: I have no idea what Firefox is trying to be anymore. Is it still about being a web browser?

Today, Pocket and Telefonica. Tomorrow, who knows?

(Seriously, if anyone knows, I'd love to hear. These core features look like random weekend projects to me.)

Firefox 38.0.5

b7j0c — Sun, 07 Jun 2015 02:14:51 +0000

> I sometimes think we can't win. If we add a useful feature Chrome has, people say "oh, you're just copying Chrome". If we add
> a useful feature they don't have, people say "oh, bloatware" or "proprietary webservice" (as if they are all in the 0.00001%
> of people who don't use a website unless the code and content are all openly licensed).

This is a strawman. Mozilla has lost its way somewhere between .org and .com. I knew trouble was ahead when they tried endowing an open source project with the headcount and headquarters of a .com...it just doesn't work and the project seems hopelessly compromised.

Firefox 38.0.5

k8to — Sat, 06 Jun 2015 23:26:30 +0000

I don't think "Save for Later" is very clear that it means "save on a proprietary cloud service for later". "Save" traditionally meant a local operation which was fully under local user control. I don't think it's hard to see why not everyone is happy about this type of change.

Firefox 38.0.5

lacos — Sat, 06 Jun 2015 11:19:21 +0000

Bad idea. No content filtering is infallible. The default must be to give users unmodified pages.

Firefox 38.0.5

lacos — Sat, 06 Jun 2015 11:12:45 +0000

Thanks for making DuckDuckGo a builtin.

Firefox 38.0.5

rghetta — Sat, 06 Jun 2015 08:42:35 +0000

I second that. With all respect to the hard work of Mozilla developers, I also find that building the pocket extension into the core of the browser is worst than offering an official Mozilla extension, especially since Pocket is not a Mozilla service. But which Mozilla ends up advertising and "locking" its users into. As other comments pointed out, there was an official pocket extension, AFAIK already working. Is the new core option better ? There was something an extension cannot do and this pocket interface needs ? If there are technical reasons to have a pocket client in core code, IMHO a better move would be defining a standard interface aimed at extension writers, perhaps it will be used not only by pocket-like services. Otherwise is difficult to understand what users gain from this change (apart not having to download the pocket extension) and why pocket and not, say, Dictionary.com, ebay or whatever else users request. Looking at bugzilla entries, it appears that pocket integration was a big work, it has even its own telemetry. Frankly, all this just to integrate a read-later third party service seems a bit overkill. Again, why not simply offering an extension ?

Firefox 38.0.5

seneca6 — Fri, 05 Jun 2015 20:43:02 +0000

Gerv, you have been active in Mozilla discussions for a long time. I do not doubt that you and everyone are doing their very best, and that the trade-offs to be made are difficult. I can imagine that continuous criticism from all sides might make one quite weary in the long time. Positive voices and congratulations should come more often.

But when there is criticism about features - apart from trolling, of course - it is because people care, they like their browser, they want it to improve, and for my part, I get unhappy when they see it moving in the wrong way. It's not just about Pocket in particular, it's the idea of including more and more into the core what used to be outside.

Which brings me to the second part. Sure, the web was different 10 years ago (and particularly 15 years ago). What has not changed - rather, it has grown - is the extreme diversity of people using it. As you see in these discussions, everyone wants his browser to be something different. The idea of a minimalist core with plenty of extensions is brilliant and timeless. It has the potential of making everyone happy. So one can have a "Firefox Jeeves Edition" that bundles extensions, which might even be the default download! And at the same time, one can strip them off or get a minimal build for those who do not want _all_ of the bundled extensions.

But now? If I want to get rid of feature X (yesterday EME, today Pocket), I have to resort to alternative builds. What if I then want _one_ of these features? There won't be any extension to add to my minimal build. All or nothing.

Firefox 38.0.5

gerv — Fri, 05 Jun 2015 14:17:32 +0000

"Always lovely to hear how marginal one is. If your requirement is that the service can immediately host data for hundreds of millions of users, for free, well, I don't see how the user will not have to pay with his privacy for that."

Right - and yet the Pocket service complies with Mozilla's privacy policy. So we've managed to find a way to provide a service that people want to use, at no cost, not even in their privacy.

"The promise of Firefox was a minimalist browser with lots of easily installable extension."

The web was a very different place 10 years ago; is it at least possible that what people expect of a browser has changed? If you want to know what I think about what a browser should be, see: http://blog.gerv.net/2015/05/the-jeeves-test/ .

Gerv

Firefox 38.0.5

seneca6 — Fri, 05 Jun 2015 13:48:45 +0000

Always lovely to hear how marginal one is. If your requirement is that the service can immediately host data for hundreds of millions of users, for free, well, I don't see how the user will not have to pay with his privacy for that. Another alternative is "print as PDF", which is what I've been using for years.

I'm not saying anything of what you cited, and I certainly do not insist that people go "without". It's the "included by default" that I think is a pity. The promise of Firefox was a minimalist browser with lots of easily installable extension. Including Pocket by default does not really fit in that picture.

Go on, I'm sure Mozilla will find plenty of other "widely-useful" features in the future. Then again, you said you deserve the support of everyone in LWN's readership, and that the state of the Free Web would be worse without Mozilla. Certainly! but not every minor issue that people complain about - and rightly so - questions the very existence of Mozilla. You cannot ask for full support of all your decisions from all of LWN's readers. And nothing worse than some criticism is what happens here.