> Asm.js is not a very sound approach, especially compared to technology like Native Client.
Yes, it is. NaCl is not portable. PNaCl is tied to the nonstandard and Chrome-specific Pepper APIs, to say nothing of all of the issues of LLVM bitcode.
> Rust seems promising, but won't be seriously usable until at least Rust 1.0, which still seems a long way off. As the year progresses, I get more and more doubtful that it'll be out during 2014, like was claimed earlier this year.
As someone closely involved with Rust development, I do not share those doubts.
What's the current roadmap for getting Rust 1.0 out the door?
Last I saw, there were over 1500 issues in the project's GitHub issue tracker. Even assuming a lot of those may no longer be relevant, and assuming minimal growth, that's still not a small number of potential problems to deal with.
And we're still seeing disruptive/breaking language and library changes, like the box-related ones recently. I know, I know, I've heard the "let's break it now rather than later" argument from you guys before. But these kind of changes don't instill confidence.
2014 is nearly half over. I know that production-grade programming language implementations don't just happen over night. It takes a lot of effort, a lot of thorough testing, and many bug fixes to result in a truly stable release. All of that takes a lot of time.
While I'd like to be wrong about this by the end of 2014, the current state of affairs is not very encouraging.
> Last I saw, there were over 1500 issues in the project's GitHub issue tracker.
A tiny fraction of which are marked blocking 1.0.
> I know, I know, I've heard the "let's break it now rather than later" argument from you guys before.
And that's invalid why, exactly?
> 2014 is nearly half over. I know that production-grade programming language implementations don't just happen over night. It takes a lot of effort, a lot of thorough testing, and many bug fixes to result in a truly stable release. All of that takes a lot of time.
Backwards incompatible language changes don't set us back in terms of the 4 years of compiler stability and bug fixes that have been going on. Believe it or not, you can change the way a keyword is spelled without affecting compiler stability.
Anyone who has used Rust for a long time knows that ICE's are far less common now than they used to be.
Yes, it is. NaCl is not portable. PNaCl is tied to the nonstandard and Chrome-specific Pepper APIs, to say nothing of all of the issues of LLVM bitcode.
> Rust seems promising, but won't be seriously usable until at least Rust 1.0, which still seems a long way off. As the year progresses, I get more and more doubtful that it'll be out during 2014, like was claimed earlier this year.
As someone closely involved with Rust development, I do not share those doubts.