News for hackers. Stuff that matters.
Re:Show HN: Nyxt Browser 2.0.0 by ibraheemdev on Thursday May 20, 2021 @09:19AM (#27220731)
Why are most open source browsers based chromium as opposed to gecko? I would think much of the OSS community would be opposed to using google owned software.
Re:Show HN: Nyxt Browser 2.0.0 by pojntfx on Thursday May 20, 2021 @09:30AM (#27220873)
If I remember correctly, it is much harder to use Firefox's components outside of Firefox than it is to do with Chromium/Blink. There is a reason why there is no relevant Electron alternative based on Firefox/Gecko. On Android the situation is different but not on any other platform: https://mozac.org/
Re:Show HN: Nyxt Browser 2.0.0 by HeckFeck on Thursday May 20, 2021 @09:58AM (#27221172)
Is/Has there been much interest in doing this at Mozilla? More people using their renderer is good for them in the long run.
Re:Show HN: Nyxt Browser 2.0.0 by e3bc54b2 on Thursday May 20, 2021 @10:30AM (#27221557)
Historically, Firefox has been the only moneymaker for Mozilla. It is possible they didn't want to make it easy to embed so anyone can create second FF on top of their work and compete without paying dues. Electron and single applications built fully on top of browser engine alone is relatively new, and possibily out of Mozilla's field of vision back then.

But, that is speculation, and I have no proof whatsoever other than tangential conclusions from Mozilla's blunders in recent years.

Re:Show HN: Nyxt Browser 2.0.0 by whatshisface on Thursday May 20, 2021 @11:27AM (#27222395)
A much more realistic explanation would be that any software that isn't specifically designed in a decoupled way will be coupled at a million points and impossible to separate.
Re:Show HN: Nyxt Browser 2.0.0 by samatman on Thursday May 20, 2021 @12:39PM (#27223461)
It's probably both.

The only upside for Mozilla is that making Gecko available is the right thing to do. Which, for a nonprofit, matters, and I'm sure that having an Electron-equivalent is the kind of thing that would be considered cool by staff.

But there is the downside that it makes it easier to build browsers which would compete with Firefox for the shrinking niche that is "doesn't run Chrome or Safari". I guess you could make a case that such browsers would take at least as much share from Chrome as Firefox, maybe more even; but you could make the other case as well.

When you add the additional downside that it's probably many man-decades of work to do it at all, well, then it doesn't happen. Although the existence of GeckoView implies that Firefox and Gecko aren't as tightly coupled as they could be.

Re:Show HN: Nyxt Browser 2.0.0 by e3bc54b2 on Thursday May 20, 2021 @12:44PM (#27223529)
Isn't that the point? Making the engine decoupled and embeddable just wasn't on Mozilla's radar, and it suffered as a consequence.

Chrome team was advcied very early in their development by none other than Android team to go for WebKit because it was easier to work with. Not Presto, not Gecko. So at least since back then, thud feature was at least somewhat on Apple's list of priorities. I'm not sure if it got carried over from KHTML, but if it did, they didn't break it intentionally.

Back in IE horror days I remember using Avant Browser. It used IE to provide fancier features like tabs. I never remember Gecko anywhere in discussion for being a platform to build on top of, except for other Mozilla projects (FFOS, Thunderbird) that Mozilla dropped sooner or later.

Re:Show HN: Nyxt Browser 2.0.0 by meltedcapacitor on Thursday May 20, 2021 @12:25PM (#27223244)
Mozilla making money? LOL. Google just sends them spare change, pretending to pay for placement, to have an alt browser to show to antitrust authorities. Third-party use of Gecko satisfies that too.
Re:Show HN: Nyxt Browser 2.0.0 by e3bc54b2 on Thursday May 20, 2021 @08:40PM (#27229235)
So if someone else makes a browser off Gecko, then that one can satisfy Google's requirements, minus the special agreements Mozilla makes for search data, and suddenly there is mo reason for Google to pay Mozilla any more..
Re:Show HN: Nyxt Browser 2.0.0 by meltedcapacitor on Friday May 21, 2021 @09:19AM (#27233625)
If Google stop paying for Gecko to track Chrome, it will lose feature parity, become unusable and fall out of use, where they can again be accused of monopoly. It seem unlikely someone doing a wrapper UI on a browser engine can fund that sisyphean task.
Re:Show HN: Nyxt Browser 2.0.0 by aakkaarr on Thursday May 20, 2021 @02:17PM (#27224975)
I am sorry but you are wrong. I remember times when first company forked WebKit to make their WebKit based browser (it was apple creating safari, or if my memory is wrong it was Google and their chrome). Mozilla foundation was much upset they didn't choose Gecko. But the company (Apple?) said it was because easiness of own implementation. Gecko was too hard to adapt.
Re:Show HN: Nyxt Browser 2.0.0 by 0x_rs on Thursday May 20, 2021 @10:17AM (#27221376)
Not since the days of XULRunner, which is ancient nowadays.
Re:Show HN: Nyxt Browser 2.0.0 by smazga on Thursday May 20, 2021 @10:27AM (#27221519)
Conkeror (basically Nyxt on xulrunner…exactly what the upper comment is asking about) used to be my browser of choice. It was baffling to me when they deprecated xulrunner. Like, why?

There was supposed to be an alternate way to do it (running with ‘firefox -app’ or something), but it was poorly documented and constantly breaking in new versions, so I eventually gave up.

Re:Show HN: Nyxt Browser 2.0.0 by JasonFruit on Thursday May 20, 2021 @12:02PM (#27222912)
I loved Conkeror, but I think this one might be even better, once I've gotten used to its quirks. I'm kinda stoked right now.
Re:Show HN: Nyxt Browser 2.0.0 by donio on Thursday May 20, 2021 @11:34AM (#27222501)
RIP conkeror, miss you dearly
Re:Show HN: Nyxt Browser 2.0.0 by kilian on Thursday May 20, 2021 @12:02PM (#27222920)
They have Geckoview on Android but from what I've heard there was no interest in making that available for other platforms, and this was before they fired a ton of their technical staff. A shame really!
Re:Show HN: Nyxt Browser 2.0.0 by avery42 on Thursday May 20, 2021 @06:17PM (#27228123)
Yes. There was an experimental project called Positron [0]: "a experimental, Electron-compatible runtime on top of Gecko". It was discontinued in 2017 [1].

AFAIK the only thing around nowadays is GeckoView [2] for Android.

[0]: https://github.com/mozilla/positron [1]: https://mykzilla.org/2017/03/08/positron-discontinued/ [2]: https://mozilla.github.io/geckoview/

Re:Show HN: Nyxt Browser 2.0.0 by pseudalopex on Thursday May 20, 2021 @07:36PM (#27228798)
There were older experiments too. Gecko Runtime Environment and XULRunner at least.
Re:Show HN: Nyxt Browser 2.0.0 by vfclists on Thursday May 20, 2021 @06:58PM (#27228486)
Mozilla has never been truly sincere in their claimed desire to create a browser builder toolkit.

Even their own developers were disappointed when they changed direction.

Re:Show HN: Nyxt Browser 2.0.0 by gostsamo on Thursday May 20, 2021 @10:38AM (#27221671)
I think that Mozilla has a project for decoupling the browser components but I'm not sure if it will result in further projects. The name was something like project Stylus, but I can't find it now and I might be wrong about the name.
Re:Show HN: Nyxt Browser 2.0.0 by jmercouris on Thursday May 20, 2021 @09:28AM (#27220841)
We are not based on Chromium, we are engine agnostic. As to why no Gecko support? Mozilla does not make it easy!
Re:Show HN: Nyxt Browser 2.0.0 by servilio on Thursday May 20, 2021 @09:55AM (#27221137)
Have you looked at Servo instead?
Re:Show HN: Nyxt Browser 2.0.0 by dijit on Thursday May 20, 2021 @10:19AM (#27221397)
I have,

I was working on replacing QtWebEngine(chromium) for Qutebrowser with Servo. It is very far from ready, most websites do not render correctly and JavaScript is very hit/miss with regards to updating the rendered view.

Simply put: Servo is not possible to use as a daily driver.

Re:Show HN: Nyxt Browser 2.0.0 by lufte on Thursday May 20, 2021 @12:39PM (#27223456)
Interesting. Do you have that work available anywhere?
Re:Show HN: Nyxt Browser 2.0.0 by dijit on Saturday May 22, 2021 @11:52AM (#27247671)
Unfortunately since it was a non-starter I didn’t package the work nicely. Can probably package it up if you really want it. But like I said: the rendering engine was not usable. The overwhelming majority of websites did not function.
Re:Show HN: Nyxt Browser 2.0.0 by shakow on Thursday May 20, 2021 @10:25AM (#27221496)
Servo is not yet ready for everyday use. Moreover, Mozilla fired most of its team working on it, so notable progress is not to be expected in the short term :(
Re:Show HN: Nyxt Browser 2.0.0 by hawski on Thursday May 20, 2021 @10:21AM (#27221434)
Servo is not done.
Re:Show HN: Nyxt Browser 2.0.0 by ginko on Thursday May 20, 2021 @09:50AM (#27221079)
What I'd like to see would be more projects picking up Servo.
Re:Show HN: Nyxt Browser 2.0.0 by nvrspyx on Thursday May 20, 2021 @10:30AM (#27221563)
Perhaps when it's production-ready. There's still quite a lot of work to be done before then.
Re:Show HN: Nyxt Browser 2.0.0 by mavroprovato on Thursday May 20, 2021 @09:24AM (#27220793)
If I'm not greatly mistaken, there is no way to embed Gecko in a desktop application right now
Re:Show HN: Nyxt Browser 2.0.0 by BenoitEssiambre on Thursday May 20, 2021 @11:55AM (#27222818)
Imo, browsers need to have a standard specification and one that includes an open source versioned controlled reference implementation of the core parts. It seems chromium has become that standard (chromium is open source afaik, not owned by google).

I often argue against purely natural language specifications in favor code based specs. I just don't think human language is nearly precise enough to write an adequate specification. Natural language words are incredibly polysemic and contextual. Look, for example, at how many meanings the word "break" has: https://www.merriam-webster.com/dictionary/break

Kolmogrov has long ago suggested that fully specified information distills down to a computer program: https://en.wikipedia.org/wiki/Kolmogorov_complexity, https://en.wikipedia.org/wiki/Minimum_description_length

The ideal language for a pure specification might be a mix of natural language and pseudo code with a pseudo test suit. However, if you are writing that, you might as well go one step further and write working testable code.

I like the concept of Literate Programming (https://en.wikipedia.org/wiki/Literate_programming) and its descendants of having code with extractable inline comments that auto generate documentation. I would argue that modern pull request based workflows that tie discussions to version controlled code changes are also the progression of this line of thought. A cleaned up version of these might make sense for a specification.

And I get some of the concerns. While natural language under specifies, reference implementations over specify. This is more of a problem with low level languages however. Modern high level languages are getting fairly close to a form of pseudo code. I fully agree that the reference implementations shouldn't contain or should hide, low level optimizations. I also understand that reference implementations can unduly tie specs to specific hardware, OSs and platforms.

But to me, over-specification is less of a problem than under-specification and it can be mitigated by labeling particular functions or blocks of code as implementation specific and not part of the spec.

Without spec written in code, the different implementations always have subtle incompatibilities. I see egregious versions of under-specification in government where horrendously vague specs are created in order to issue RFPs for getting software built. They usually end up with non working software at mind blowing cost.

People have this weird misconception that you are contracting out to build software. You are not. Building software is really easy. You press the build button or type the compile command. Building software has been fully automated for a while now. What is difficult is designing software and specifying what it must do. This is because there is a vast jungle of protocols, business flows, hardware and software platforms that need to be interacted in different ways for different needs. This is what needs to be specified and only computer code can do it adequately.

I wish that Mozilla adopted the chromium core. We really need a well funded non-profit managed release of the reference browser.

Re:Show HN: Nyxt Browser 2.0.0 by cylinder714 on Thursday May 20, 2021 @01:22PM (#27224096)
This is really, really good, thank you.

My only concern is that one ends up with something like TeX, written in a long-obsolete dialect of Pascal, requiring a translation layer to convert the source into a more relevant language (C).

Re:Show HN: Nyxt Browser 2.0.0 by fmakunbound on Thursday May 20, 2021 @01:56PM (#27224636)
IIRC, Mozilla killed off embedded gecko years ago.
Re:Show HN: Nyxt Browser 2.0.0 by sequence7 on Thursday May 20, 2021 @09:22AM (#27220768)
I suspect the answer is because of compatibility. Chromium has become the industry standard, everyone tests only on Chrome (and Safari for mobile) so you can expect fewer issues if you're based on chromium.
Re:Show HN: Nyxt Browser 2.0.0 by graderjs on Thursday May 20, 2021 @09:27AM (#27220833)
But I think it's a valid question. Because for a niche browser (which most of these alternate projects are), mainstream compatibility for all websites doesn't seem like it would be the most important thing. I think the idea of using Gecko[0] is a solid one. :P ;) xx

[0]: https://github.com/mozilla/gecko-dev

Re:Show HN: Nyxt Browser 2.0.0 by sequence7 on Thursday May 20, 2021 @11:02AM (#27222012)
I agree it's a valid question and it seems my suspicion is not the main reason. Based on the other replies to the parent it seems that choosing Gecko requires more work.