News for hackers. Stuff that matters.
Re:Firefox does not have a well-designed privsep model by benatkin on Monday June 11, 2018 @03:48PM (#17287875)
Chromium is an open source project, and we should treat it like one. Google paid for most of its development, but were able to pay many open source developers less because they were excited about working on an open source project (see the thread about compensation on today's Ask HN post about working for startups). Ditto for Android. Nobody should be shy about forking these projects. We don't owe Google for them, beyond what the license says. Fair is fair.

If we can treat Chromium like an open source project, maybe we can retire Firefox.

Re:Firefox does not have a well-designed privsep model by digitalzombie on Monday June 11, 2018 @05:20PM (#17288692)
Excuse me?

I love Firefox. I don't want it to retire and why would you want to limit choices for other people?

Also how do you make the case that Chromium can survive without Google continuous investment in Chrome?

Mozilla got money to continue Firefox to give competition. People can say whatever they want but Firefox is a good browser and the company have made many good moves to make the internet a good and open place where as others have attacked it. To tell a company to retire it main product that bring it money is beyond reason. Especially a company that aren't doing any harm to anybody.

Re:Firefox does not have a well-designed privsep model by benatkin on Monday June 11, 2018 @05:44PM (#17288889)
I'm exaggerating by saying "retire". I don't mean getting rid of it - major open source projects always live on. If things are as dire as this article, my experience trying to switch back to Firefox, and the "Through the Looking Glass" adware extension debacle suggest, a solid independent open source alternative could allow it to keep going where it unfortunately seems to be headed.
Re:Firefox does not have a well-designed privsep model by nbsd4life on Monday June 11, 2018 @04:48PM (#17288442)
Chromium is terrible for the BSDs, though - it requires something like 300 patches to build: https://github.com/freebsd/freebsd-ports/tree/master/www/chr...

Mozilla, unlike Google, is not hostile to the idea of receiving patches from BSDs.

Re:Firefox does not have a well-designed privsep model by 2trill2spill on Monday June 11, 2018 @05:54PM (#17288950)
You know why those patches haven't been upstreamed? Is the chromium project refusing to take the patches?
Re:Firefox does not have a well-designed privsep model by jacob019 on Monday June 11, 2018 @03:52PM (#17287908)
Some of us love Firefox and do not want webkit hegemony, even if it is open source.
Re:Firefox does not have a well-designed privsep model by digi_owl on Monday June 11, 2018 @05:19PM (#17288684)
IF that was true, then perhaps more effort should be directed towards making Gecko a drop in engine?

The big "benefit" of webkit is that it is a engine without the whole "browser" thing tagging along.

But the last time i poked at getting gecko to stand on its own, albeit some time ago, it involved grabbing Firefox and giving it some configure (or whatever that python thing that is masking as configure that Mozilla is using these days) options to have it barf out only Gecko.

Re:Firefox does not have a well-designed privsep model by genghizkhan on Tuesday June 12, 2018 @03:30AM (#17291374)
I think they're working on servo for that. It'll be able to be dropped in place of Chromium: it'll support the same API.

Or so I remember.

Re:Firefox does not have a well-designed privsep model by MBCook on Monday June 11, 2018 @05:02PM (#17288568)
It’s not even WebKit anymore.
Re:Firefox does not have a well-designed privsep model by seba_dos1 on Monday June 11, 2018 @03:54PM (#17287936)
Just from the usability and performance characteristics alone, I'd rather retire Chromium than Firefox.
Re:Firefox does not have a well-designed privsep model by bthrm on Monday June 11, 2018 @04:14PM (#17288124)
Firefox is more usable and performs better than Chrome? You can’t be serious…
Re:Firefox does not have a well-designed privsep model by seba_dos1 on Monday June 11, 2018 @04:32PM (#17288284)
It definitely does. Chrome is borderline unusable on 8GB RAM - more than a few tabs open make it go into swap hell, while Firefox works well with thousands (Chrome doesn't even have the concept of unloading open tabs!). Also, you can't even disable tab shrinking in Chrome, let alone move tabs to the left border of the window, as I do with Firefox. This results in an UI that's practically unusable with the only way to navigate the tabs being the mouse wheel.

Chromium used to perform visibly better than Firefox with somewhere around one to five tabs open, so sometimes, if I didn't have the browser running and didn't want its full session to load, I've started Chromium just to quickly check one URL - however, recent versions of Firefox made it comparable in such cases, so the only use case for Chromium I have now is to deal with badly coded websites that assume it's the only browser out there.

Re:Firefox does not have a well-designed privsep model by bitmapbrother on Monday June 11, 2018 @05:59PM (#17288989)
>Chrome is borderline unusable on 8GB RAM - more than a few tabs open make it go into swap hell,

I'm currently using Chrome on a 4GB system with 10 tabs open with no performance issues or fans screaming. To claim that Chrome is borderline unusable on 8GB of RAM is a lie.

Re:Firefox does not have a well-designed privsep model by bwbw223 on Tuesday June 12, 2018 @12:06AM (#17290636)
> is a lie

Or maybe people have different standards/expectations?

Re:Firefox does not have a well-designed privsep model by taborj on Monday June 11, 2018 @05:15PM (#17288659)
I think this is dependent on many factors. Right now I'm on a laptop with 8GB of RAM, I have a VMWare VM running locally taking up 3GB of RAM, and Chrome with 6 tabs open. Yes, I'm getting near swap territory, but everything is cruising along nicely...
Re:Firefox does not have a well-designed privsep model by bthrm on Monday June 11, 2018 @04:37PM (#17288334)
I’m sure they have telemetry and know that almost nobody leaves hundreds of tabs open.

When you tell me you open thousands of tabs, I wonder if you’re just stress testing the browser. I expect software not to work well when I stress test it.

Re:Firefox does not have a well-designed privsep model by seba_dos1 on Monday June 11, 2018 @04:42PM (#17288382)
I'm sure nobody leaves them open with Chromium as it's impossible.

My common browser usage is to have low hundreds of tabs open - right now it's a bit higher than usual, 1086, although it's not the record. It's just regular browsing, topic research, some social media and stuff left for later. Works well - under Firefox with Tab Center Redux, that is. I've tried switching to Chromium in the past, when Firefox used to be visibly slower, but I've never could use it for longer, always went back to Firefox as Chromium's UX was horrible. With recent versions, I don't even feel the need to check out Chromium ever again.

Re:Firefox does not have a well-designed privsep model by genghizkhan on Tuesday June 12, 2018 @03:31AM (#17291381)
It has been for a while now. In particular, it has been since they started going all-in on Quantum from FF 57.
Re:Firefox does not have a well-designed privsep model by burnte on Monday June 11, 2018 @04:14PM (#17288117)
Firefox is an OSS product, and has been around longer than Chromium. Why not "treat it like open source" too and contribute?
Re:Firefox does not have a well-designed privsep model by benatkin on Monday June 11, 2018 @04:16PM (#17288141)
Because I agree with Theo de Raadt.
Re:Firefox does not have a well-designed privsep model by burnte on Tuesday June 12, 2018 @06:33PM (#17298679)
I don't supposed you'd like to expand on that response for my personal edification, would you?
Re:Firefox does not have a well-designed privsep model by TheDong on Monday June 11, 2018 @03:54PM (#17287940)
It would still be terrible for the web, as a standard, to have less diversity. Chrome+chromium barely count as diversity since the standard-related code (rendering, js engine, layout, etc) is identical.
Re:Firefox does not have a well-designed privsep model by sli on Monday June 11, 2018 @06:27PM (#17289167)
That, and Chrome tends to make some hardline, opinionated changes that aren't reciprocated by other browsers. "It works in Edge and Firefox, but it's broken in Chrome" is a a fairly common thing to hear where I work.

This includes things like weird rendering bugs, and extra "Chrome-only" features.

Amusingly, as of last week, Chrome Experiments worked fine in Edge and Firefox, but was broken in Chrome. Because of Chrome's opinionated, built in whitelisting for media content forgot to include the Chrome Experiments domain name.

Yikes. It's starting to sound a lot like old IE.