4 ms·
It is a hard problem, and Mozilla has put a ton of effort into Rust, but at this point they aren't supporting any of their current extensions (XUL) and WebExten
by trome 10y ago
It is a hard problem, and Mozilla has put a ton of effort into Rust, but at this point they aren't supporting any of their current extensions (XUL) and WebExtensions will not cover most use cases that XUL covered. This isn't a Python 2 to Python 3 event, this is Mozilla going from the equivalent of C to Java.
Most of your favorite extensions will die, and it is directly forcing me to drop firefox for chromium at work, which I abhor. Sadly the feature I need (decent kiosk mode) will die in its current form (as a XUL addon) only a year after deploying this solution, versus with chromium its built in, and not going away any time soon.
- sametmax 10y agoYes they had to make a some very hard choice on this one. It's going to piss of a lot of people on the way. But on the long run I think it will result in a better browser. With less versatility, yes. But I got an entire OS at my disposal for advanced used, it's rather have a very good browser doing browser things. I'll the tab group extension though. I loved it and it won't be possible with webextensions.
- homarp 10y ago>With less versatility, yes. isn't/wasn't that the point of firefox though. >a very good browser doing browser things. Chrome is the competition. Quite good at 'browser' things
- sametmax 10y agoYes. But do in need sqlite manager to be a firefox extension ? I can probably live without it. Do I need it to be at least as fast as chrome ? Yes.
- mercer 10y agoYour comment is a good example tightrope that Mozilla is on. I've switched to Firefox a few times over the past years, but usually I'd switch back to Chrome because, whether true or not, it felt more performant/cleaner/better. The main reason I bothered trying to switch back to Firefox was always the versatility or 'freedom'. Stuff like the tree-style tab plugin. Realistically, for me to break my usual patterns and try Firefox again would have to involve at least one of the following: - equal (perceived) performance to Chrome - freedom/versatility in regards to plugins - some as-of-yet not seen before advantage With all else being equal, or even ever so slightly worse than Chrome, I'd go for Firefox simply because I distrust Google. Thing is, I'd characterize myself as much more security- and privacy-conscious than most people, and yet I'm still on Chrome despite giving Firefox a shot at least once a year for the past few years or so. I catch myself forgetting to test the stuff I make on FF occasionally (I'm a web developer). While I'm by no means a paragon of sticking with my beliefs, I'm still a bit of a privacy-weenie. The fact that I'm not using Firefox worries me a bit. The fact that I can't quite explain it frustrates me. If I had to give a reason as to why I use Chrome instead of Firefox, I'd say there are two primary reasons and a very vague third one: 1. Chrome is just more familiar at this point, especially its dev tools 2. Firefox feels like one of those Java applets: slow to boot and alien to my OS (moreso than Chrome). 3. Chrome feels snappier
- Certhas 10y ago> WebExtensions will not cover most use cases that XUL covered. This is just FUD. Especially as Mozilla is working closely with the developers of the most popular add ons to make sure all capabilities they need will be available (see also webextension experiments). Of course not all capabilities of XUL will make it, but as Chrome has a vibrant extensive add on community around WebExtension the idea that most usecases are affected is implausible.
- sametmax 10y agoApparently, some very popular extensions still have problems with it. The one that bother me the most is tab group: https://github.com/Quicksaver/Tab-Groups/issues/535 https://github.com/Quicksaver/Tab-Groups/issues/535 It was an feature of Firefox that I loved. They removed it. So I now use the extension. And the author is now announcing he won't be able to port it.