Then I started working with Zig[0]. It's about 90% the language I want and even though it's still immature and the syntax, std lib, and semantics are in flux, I still find the quality of life improvement so huge that all my current projects are in Zig.
Zig is a better C to Jai's better C++, I think. It's a small language, it favors simplicity and explicitness, it has entirely manual memory management to the point that there isn't even a default allocator (we pass one in to anything that requires one). It even lets you import C headers directly to interface with C libraries.
(skip the "stable" 2.0 download, it's way out of date from master)
A lot of language devs want to play with semantics but aren't into the whole "let's do the hard work of actually getting off of C as the baseline" idea so it's nice to have a project that does and has a fairly explicit goal of absorbing and replacing the existing infrastructure step by step.
I tried creating my own languages before, some 15 years ago. While I had fun writing compilers, I missed a lot more writing actual programs. One day I suddenly realized I don't really need new language, what I needed is simply context dependent syntax and code block organization, which can easily implemented on the spot. It become MyDef and I have been programing with it entirely for more than 10 years now and wondering how others don't go down the same path.
Actually in a broad sense, Knuth's literate programing adventure is very similar to mine. He also wondered how others were not as excited as he was.
So rather than a new language, try a new dimension.
Here is a short example:
page: test, basic_frame
module: c
$local int A[3]
$for i=0:3
A[i] = i
$dump A[0], A[1], A[2]
Despite the look, if you know c, you recognize it as c. And as you get familiar, you see it exactly as C. `basic_frame` is merely a custom template so one doesn't have to always write the main function signature. $for is shorthand for `for-loop`, `$dump` is simply printf with simple type translations.Now that you had this meta layer, there is no limit on how you want to dress up your code. Examples:
$local A[3]: int
$sumcode(3) A[i]=i+3
$foreach A
$dump $(t)
or page: test, basic_frame
$call declare
$call init
$call report
subcode: declare
$local pn_A[3] = {1, 2, 3}
subcode: init
$foreach pn_A
$(t) = $(t)^2
subcode: report
$(set:print_to=stderr)
$foreach pn_A
$print "%d: %d", $(i), $(t)
There are lots of idioms and conventions, but it is not a problem as long as you recognize them only as clothing.Lastly, when dressing up the code makes little sense or you simply don't know better way of dressing them, you can simply write vanila C (or any languages that you are writing). MyDef never tried to compile your language.
Like ?
Such as? We are always looking for for things that get in the way of use.
3, 4, and 5 are all pretty sore points. 2 would be as well if it weren't extremely likely that for any game you create roll some sort of reflection system. It would be a sore spot for small projects.
3 - No access to classes and polymorphism is a bit of a downer. Although I'm a big believer of data-oriented design, I think there are still many cases where modelling things as objects is the best way to go, especially for things like scene actors and game objects.
4 - No built in threading (core.thread) puts a hamper on any sort of distributed updating or performance optimizations you'd want to do. I'd imagine the workaround is external multi-threaded dlls or something, but it sounds like a pain to manage and deal with in a workflow.
5 - No dynamic arrays. This container is used abundantly in game development. It shows up 5 or more times for every heap/stack/queue. This may not be an issue if you want to write a special memory environment and your own containers, which is common in game development, but not having access to a standard one certainly hurts when it comes to smaller projects or graphics programming experiments.
I may be somewhat off base here since what happened is that I read about those being missing and decided not to try it.
Another note is that a big thing I'm looking for is massively improved compile times, which is one of the biggest things Jai looks to have over C++ in addition to its compiler hooks and #run.
You will say that you don't want a GC but in reality the GC can be optimized as much as you wan't, and in the mean time you'll get precious _developement time_ to optimize further.
If you can afford the D runtime by all means take it ; for some products I'm doing without the D runtime, and the surprise was that it did't make things any faster _at all_ without the runtime (we did it for other reasons). The only real gain is memory consumption of GC vs manual and that is achievable without going through extreme coercitive measures like -betterC.
[0] https://p0nce.github.io/d-idioms/#The-impossible-real-time-t...
- std.experimental.allocator
- @nogc to ensure, well, no GC usage
- various things built on placement new (`emplace`)
- GC profiling with -profile=gc, will display number of allocations and size
None of this existed when people were already making high-performance systems with D...
--- A popular example is minecraft that uses mainstream G1 GC.
I was suspicious about #3 since classes in D don't necessarily need the GC. For starters D supports C++ classes, so this works:
https://gist.github.com/atilaneves/b69ed0399efac21bb7a977afd...
I used `scope` with `new` to allocate on the stack, but `emplace` should just work as well to "placement new" a class in malloc'ed memory.
#4 is annoying but one could always use the OS threading API. It's doable to write a library solution to do that akin to boost or Qt before C++11. You do it once and it's there.
#5 No dynamic arrays are also annoying. But... writing a clone of std::vector isn't hard, and there's a unique array in my automem library that _should_ be usable in betterC (but I haven't checked): https://github.com/atilaneves/automem/blob/master/source/aut...
> Another note is that a big thing I'm looking for is massively improved compile times
You get that with D.
Those rely on the garbage collector. If you are willing to do C style memory management, they are all available with betterC.
If you want to write games, use your favorite language, and just START. All limitations can be circumvented somehow.
The only nagging point is the copy-paste compatibility with C. :\
I find Swift to be the best so far, its strong ties to the Apple ecosystem notwithstanding.
Having tinkered with various environments since the Sinclair Spectrum's BASIC (though back then I had no idea what I was doing) I've successively fallen in love with Visual Basic (6.0, not the .NET impostor) and then C#, and now I'm smitten with Swift.
In Microsoft land I was let down by their usual dichotomies that did not allow you full access to all their APIs (or hardware performance) from anything but C++ (though I did manage to make playable DirectX games in VB6). I don't know how that has changed in the last decade since I jumped ship to Apple, but .NET was treated as a second-class citizen for too long.
With Swift, not only do I have access to everything that Apple's operating systems offer, it's also a modern language which feels naturally suited to game design patterns. Although until Swift 2 I often had to fight the language, since Swift 3 and especially now in Swift 4.2 I find it effortless to reason about most game architectures while sticking to the language's idioms.
I also appreciate it for enforcing good practices and getting rid of most sources of ambiguity (such as the ++/-- operators and the legacy `for` syntax, or being specific about %/modulo and so on), even if it may feel annoying in the beginning if you're coming from "looser" languages. It also compiles to native code and can easily interop with C [0], but best of all: I am glad to be able to finally leave semicolons behind. C# honestly feels archaic in comparison now.
As for the elephant in the room – that Swift is only good for iOS, macOS, tvOS and watchOS – well, those platforms have a combined userbase of over a billion users, and the most successful app market [1]; I don't feel that it's such a downside to be limited to those platforms. I do wish that the Apple TV came with a proper gamepad so we could only target that without losing users. It could be a serious contender to existing game consoles if Apple took that route.
If you're in love with Swift too, I'm writing an open-source game engine which offers an Entity-Component-System architecture on top of Apple's APIs [2].
[0] https://theswiftdev.com/2018/01/15/how-to-call-c-code-from-s...
To the point the Windows UI team does all their graphics demos in C#.
I’m too deeply entrenched in Swift though and disgusted by the sight of semicolons in a language anymore :P (maybe I’ll dip into VB.NET and see how it’s doing.)
Most new languages just don't get the things right that C++ did, which is why despite it having a ton of deficiencies when it comes to game development, it's still what everyone in AAA is using. Companies in AAA spend so much money on software to even just somewhat improve places where C++ has gaping flaws. A better language should have been developed by now and I don't even want to think about how much money C++ has cost the industry as a whole, even just purely on iteration speed.