"Since the UI is written in major parts in JavaScript and CSS and an HTML-like language called XUL, anyone with webdev skills can improve it!"
I thought 57 was dumping XUL completely. Wasn't that a major point of 57?
There are some long-term investigations into using HTML/JS/CSS for the UI [1], but for the moment, the performance difference is still too much.
(I checked the browser.html issue tracker but most of their perf bugs seemed to be resolved.)
This is just a guess on my part.
That was a few years ago, results may be obsolete, of course.
Someone (working now at Apple) wrote a super fast list component in HTML for Firefox OS so this can definitely be done.
I seem to recall that one of the big "optimizations" that FF did compared to the mozilla suite (these days known as Seamonkey) was to translate a bunch of UI elements into native equivalents rather than drawing it all from XUL.
Thus replacing that with HTML may well be closer to reverting to the pre-FF XUL way.
Think about it this way: the transition away from XUL means the UI code (and more) is going to be very unstable for awhile. Thus addon authors would need to be continually rewriting their code for each update to keep up with the changes. It's better to bite the bullet now and switch to a more stable api.
Maintaining two interface languages doesn't make much sense in the long term.
XPCOM is another independent technology https://developer.mozilla.org/en-US/docs/Mozilla/Tech/XPCOM.
They are disabling the direct access from extensions to apis used by browser itself (that is access to XUL _and_ XPCOM), to be able to modify the browser ui without worrying about breaking extensions, with the goal to remove XUL eventually.
There are some people (like me) who use firefox because of addons that can't be supported on other browsers due to the lack of API support, so it's a bit disappointing if they're going to pull the rug from under those addons. If they do, they might see people leave for another browser.
For reference, the addon in question for me is Tabs Mix Plus.
But the core APIs are the same as Chrome's. The implementation might not yet always be perfect, but the APIs themselves are stable. So, for the vast majority of add-on developers, the experience will already be much less burdensome. Or it will be in the future.
Yes, some add-ons will not happen or might only be possible long after the release of Firefox 57. But they're not going back on that plan. They haven't just forbidden these extensions for funsies. They've very much already used the opportunity to rip out tons of legacy code and refactor things that they could have never refactored before without breaking tons of add-ons each time. At this point, there's very few legacy extensions left that still function in Firefox Nightly, even if you tell it to load them.
Not to mention that this has helped them to speed Firefox up by a significant factor. All the performance improvements since Firefox 48 would not have been made without this deprecation on the horizon. And considering that a significant chunk of Firefox users use no extension or just the obligatory ad blocker (or any set of extensions that's going to be easily replaceable with WebExtensions), this is also very much the decision that has to be made when asking yourself what's best for users.
As for people possibly leaving Firefox when their favorite extensions stop working, yes, that's possible, but Firefox's market share has been growing again since Firefox 48, and there's also just the fact that Firefox is still going to be the browser with the most powerful extension API.
You're not getting more extensions by switching. You'd have to be fine with even less extensions in exchange for maybe a particular feature that you like in another browser.
You have basicly told us all "your problems dont exist, dont matter, or we dont care, deal with it"
I'm done with firefox post 56.
And due to jilted lover syndrome, I feel compelled to post in threads like this one saying so. Heh.
I have no problem with you being sad about these add-ons going away. I do have a problem with you acting like what Mozilla does is completely unreasonable and objectively bad for users.
Its great to see they are going to drive firefox off a cliff with yet another anti-user decision.
The situation with popular extensions has improved, but there is going to be a lot of pain, especially with XUL functionality that they refuse to implement in WebExtensions.
It's true it's still there, but I think the plan is using it less and less. No new UIs are written in it and many things have been rewritten from XUL to HTML5. Eventually it could be removed entirely.
Its amusing to me that mozilla expects everyone else to be moved off XUL, but not their own code because it has legitimate reasons....
Hilariously, dispite the pro-privacy stance mozilla pretends at, they want to start tracking urls visited
https://news.ycombinator.com/item?id=15071492
Why keep using firefox? Because I love Mozilla so much when they clearly have zero interest in my needs as a user?
And no, it's certainly not just an administrative deadline. For now, you can still tell Firefox Nightly to install legacy extensions, the vast majority of them are just by now completely broken, because Mozilla is ripping out all of that legacy code and refactoring what's left. Here's a list of all of the things that they can get rid of: https://bugzilla.mozilla.org/show_bug.cgi?id=1347507
The question is how much the extensions can do, in the old model they could do anything firefox code could do, in WebExtensions model, they can do only very limited set of things.
On one hand it is quite sad to see firefox extensions go, on the other hand the old model was a maintenance hell both for firefox developers and for addon developers, because smallest changes in firefox ui were breaking bunch of addons, and even changes in addons themselves were breaking other addons, which all had incompatibilities and had to do lots of subtle hacks to work together.
I have worked on firefox addons for several years, and had to quit firefox when they started ignoring addons and investing in old addon sdk. But saying Mozilla is intentionally killing firefox is not fair the whole thing was shaky and full of hacks and would crumble by itself too.
What they should do now, is splitting their ui from the engine, so that advanced users can hack on the ui and get all the benefits of old addon system without the mess. If they fail to do that brave browser with https://github.com/brave/muon will take that niche.