Slashdot Mirror
Big Changes From Mozilla Mean Firefox Will Get Chrome Extensions
Mozilla announced yesterday a few high-level changes to the way Firefox and Firefox extensions will be developed; among them, the introduction of "a new extension API, called WebExtensions—largely compatible with the model used by Chrome and Opera—to make it easier to develop extensions across multiple browsers." (Liliputing has a nice breakdown of the changes.)
ZDNet reports that at the same time, "Mozilla will be deprecating XPCOM and XUL, the foundations of its extension system, and many Firefox developers are ticked off at these moves."
In 2-3 years Firefox will be nothing more than another shitty Chrome clone. I can only hope this absurd move leads to a serious fork of the browser that focuses on getting back to the original goals of Firefox.
That already exists in Pale Moon. It avoids the Australis interface that is just a clone of Chome's toolbar style, doesn't have the new start tile page with the "suggested" additions, or Pocket, or the earlier "social" additions.
What it needs is more developers and a plan on how to move forward and improve the browser until it stops being a tweaked third-party Firefox.
I think it's fair to consider Rust a "debacle" so far. Have you actually tried to use it? I've been following it for years. It took them fucking forever to get the 1.0 release out. Until then, they spent most of the time flip flopping back and forth between the different options for language features and library functionality. It wasn't just evolution, improvement, or rapid development. The language convulsed for years and years. You couldn't write code one weekend and reliably have it work the next weekend!
For a language whose web site proclaims it to be "a systems programming language that runs blazingly fast, prevents nearly all segfaults, and guarantees thread safety", the Rust implementation (which is mostly written in Rust!) is goddamn slow, and it's also full of bugs (over 2100 open bugs as of this time). Rust's supporters and developers will blame it on there being lots of "legacy code" and "new ideas" and other excuses like that, but I think the reality is that Rust just doesn't live up to the promises that were made about it and its capabilities. It doesn't help that the Rust project attracted some Rubyists, after it started to become clear that the Ruby and Ruby on Rails hype was rapidly falling apart and being associated with that community would soon be a liability. They ended up bringing their typical hype over to Rust, which has not been good for its community.
Despite all of the hype, Rust 1.x is actually a pretty awful language in a lot of ways. The semantics are mediocre. The syntax is worse than other C-style languages. The implementation, as I mentioned earlier, is slow and buggy. The standard library is lacking. The documentation isn't very good, either. All in all, it's a real disappointment, especially given how long we had to wait for it!
While Rust was sitting there going nowhere fast, we saw C++ come along and go through several major updates, which brought in some very useful language and library functionality. C++14 is a superb language. It's the most capable language out there. You can write low level code that gives you near-complete control. You can write mid-level code with ease. You can write high-level code using concepts borrowed from functional programming languages. You can mix and match all three approaches however you want to or need to. You can write code without garbage collection. You can easily write advanced code that never uses pointers, or uses them safely if they do need to be used, giving you almost all of Rust's safety without the burden of Rust. There are a huge number of libraries you can seamlessly use, too. And never forget that there are multiple, high quality C++ implementations. GCC and Clang/LLVM are two very capable open source systems. Then there are many commercial implementations, too. There's just the one shitty Rust implementation, and nothing suggests that's going to change any time soon.
Rust hasn't delivered at all, while C++ consistently has delivered. There's no sane reason to subject yourself, and especially clients or customers, to Rust. Mozilla, instead of wasting their time and resources on Rust, should have just started gradually using the new C++ functionality in Firefox and their other existing products written in C++. They didn't need to develop a whole new language that falls significantly short of C++! So, yes, Rust can be considered a "debacle".