I have my doubts on Jai, besides being built towards game development, from what I read/watched about it, it has 2 or 3 meta programming capabilities, like comptime, macros, etc it feels too much of the same, also Jai is not built towards correctness or safety, John mentality is that he knows what he is doing, so he doesn’t need those guardrails and he wants instead expressiveness.
Also Jai is like C++ in complexity, while Zig is similar to C, very simple language.
Carbon is vaporware so far, there’s no language that could be used yet, because they first need to solve the C++ interop and fast compilation times, that is what will shape the language, so no one is using it, because it doesn’t exist yet.
>Also Jai is like C++ in complexity, while Zig is similar to C, very simple language.
And most importantly, Zig is aiming at being a C++ replacement with the simplicity of C, it is not trying to replace C.
I think you meant to say Jai, not Zig.
Good luck with that, it is basically Modula-2 with C like syntax, and we aren't even getting into the whole ecosystem that it is missing on.
Any C++ or C replacement will need to win the earths of mainstream OS and game console vendors, otherwise it will remain yet another wannabe candidate.
Those have already their own languages, alongside their own C and C++ compilers, and are only now starting to warm up to Rust.
Zig or any other candidate will have a very hard time being considered.
So no one should even try because they will never win over all of the C/C++ crowd so are doomed to fail and forever to be a wannabe? I think Andrew has gone about things in a good way, going back to C and exploiting hindsight, not trying to offer everything as quickly as possible. Extend C but keep C interoperability and do both better than C++ instead of trying to be the next big thing and he goes about it in a very deliberate and calculated way. He may not succeed, but the effort has given us a great deal.
One should try, while being aware of the realities of language adoption.
I disagree Zig is that great deal of a language, it would have been if we were talking about 1990's programming language ecosystem, not in 21st century.
Use-after-free problems should not be something we still need to worry about, when tooling like PurifyPlus trace back to 1992.
I don't think Andrew believes Zig is going kill C or C++, he probably has hope but I think he is aware of the reality. He found a way to make a living on something he was passionate about.
Use-after-free is a fact of life until something kills C, but the realities of language adoption are against that. Zig seems interesting and worthwhile in offering a different perspective on the problem and does it in a way more agreeable than Rust or the like for all those who love C and are adverse to large complex languages. The realities of language adoption are as much for as against Zig, large numbers of people are still getting drawn to C and Zig seems to do a better job addressing why so many are drawn to it than the alternatives.
For that to matter OS vendors that only care about C on their platforms, have to also care about Zig.
Otherwise the only users are going to be the ones happy to do some yak shaving instead of the actual application code with the vendor tools.
It also ignores that C doesn't stand still, the competition is C2y, not C89.
Remember that it was Jai that inspired all these new languages. When you talk about the capabilities like comptime, that's all from Jai and why John no longer does public presentations(as people keep stealing his unfinished ideas).
Your comment about gamedev focus makes no sense as that it the most hardcore segment of all the programming there is. So if a language is good for gamedev, it's good for everything else - with high performance.
I'm still in the GC camp with Go and don't see myself leaving any time soon but Zig is just rust-fugly and takes for ever to complete(it started 10 years ago, mind you). Odin is essentially complete, just lacks official spec. I like it but can' bring myself to use it as it lacks methods and I won' be going back to writing a procedural code like its 2002.
I'm curious to see Jai being released, despite having no use case for it. My initial post is merely about purposefulness, or the lack of, for named programming languages as nowadays John's name will carry more weight than Zig could ever have. so without Zig being 1.0 after a decade, and having no competitive advantage over Jai, it has no chance to survive after Jai is released. As I said, Odin will likely will as it is quite simpler, more niche language. Zig just goes directly against Jai and it will lose.
> When you talk about the capabilities like comptime, that's all from Jai
You really should learn a few new languages if you think that’s remotely true. For example, Lisp macros are the distant ancestors of most metaprogramming techniques, dating from the 60s. But there are many similar techniques that precede Zig comptime and Jai, like D’s mixins and templates, which together have very similar power to comptime. There are also Nim macros which I believe also precede Jai. Even the C preprocessor could probably be considered an inspiration for anyone creating a C competitor before we need to invoke Jai ideas.
> Remember that it was Jai that inspired all these new languages
You’ve made this claim on here before and when I asked you to substantiate it with a source you were completely unable to.
> why John no longer does public presentations(as people keep stealing his unfinished ideas)
That's your opinion or you have a source for that?
he said it in a video, i think it was in the recent wookash podcast. although you might fing a short clip of the segment somewhere.
The idea of languages "stealing" ideas from each other is not something anyone building a language cares about. I'll just charitably assume you've completly misinterpreted something he said.
> Remember that it was Jai that inspired all these new languages.
Not really. Rust was a thing long before Jai.
Not sure why this is being downvoted, Blow started working on Jai in 2014, by which point Rust was already nearing stabilization with 1.0 shipping in 2015.
In fact Rust was specifically discussed as a possible alternative to the C++ status quo in Jon's initial "A Programming Language For Games" talk which roughly marks the inception of his current / upcoming language.
> as nowadays John's name will carry more weight than Zig could ever have.
I'm sorry but Zig has been used to create actual production software for many companies whereas Jai has been used maybe once for a mediocre game.
You sound like you're in a cult