Does anyone have enough experience with a Crystal to say if it's

File: Screen Shot 2017-06-07 at 13.29.58.png (23KB, 502x142px) Image search: [Google]

23KB, 502x142px

BUMP
Also interested in this.

>>60791205
I hate the syntax. I can't really see why it is listed as a pro, but maybe that's me

>>60791205
bump

>>60791205
>good language to learn
That's very difficult to say. Typically when people think "good language to learn" they mean "employable", which it absolutely isn't right now.

I haven't used it a whole pile but it seems just like C++ in a lot of ways. A lot more needless keywords than say, Python, fast and optionally object oriented, but with a faster turnaround time and better toolchain. And less fun and games with memory obviously.

File: 1439815808390.jpg (36KB, 500x498px) Image search: [Google]

36KB, 500x498px

monitoring
there's gotta be a catch

>>60792449
A catch this instant? It has absolutely no parallelism whatsoever and from my own testing the concurrency is a bit screwy. You can have multiple "threads" a la Python, but when I tried it I found that if one "thread" doesn't have a sleep in it, it will literally never be switched out for another. It's not a serious issue, but I doesn't inspire confidence right now.

It also has exception handling and operator overloading if those are things that you hate.

File: 1451364096672.jpg (964KB, 1920x1080px) Image search: [Google]

964KB, 1920x1080px

>>60792481
i guess those things don't really matter to me if i don't know any programming in the first place

>>60792481
Found some other stuff it doesn't have that you might want:
Function throw specifications.
User defined literals.
Any way to define casts (seemingly? I can't find one anyway).
Concepts (i.e. template member constraints).
Compile time assertions.
Friend classes/methods.
Any sort of using namespace declaration.

Also while you can have constants, they're made by starting the variable name with a capital letter, which I find a bit weird.

>>60792819
All C++ crap you don't need.

>>60792836
I'm inclined to agree for most of it, but there's nothing wrong with concepts, throw specifications or static asserts.

The language does have functionality for something similar to concepts already, but you have to check every function individually with #responds_to?, it would be far easier to check "inherits from <interface class>".

I've written in it before, made a project and so far I'd say it's pretty good. I've not had any problems with it and found nothing it missing for me.

>>60791205
Garbage collected garbage

>>60791205
It's not finished yet and there are quite a lot of rough edges, in particular regarding the compiler.

That said, what is there is quite nice. It's not perfect, but if you come from a dynamic anything language and want a comfy RAD language, it's probably your best shot once it's ready.

>Yet another C++-killer with GC
Will they ever learn? As if Go, Nim and D weren't enough to show how stupid the idea is.

>>60793038
>it's probably your best shot once it's ready.
And to clarify, other RAD languages would be
- golang, but they are obviously too incompetent to add a sane type system or metaprogramming features
- nim, but they bought too much into the compiling-to-other-languages meme which comes with too many disadvantages
- swift, which seems like a glorified wrapper over ObjectCs standard library
- modern, garbage collected Ada, which has had its chance

>>60793069
Is it supposed to kill C++? I thought it was supposed to be more like the other scripting languages, but less horrible to work with (i.e. competing with Python/Ruby/Javascript rather than with C++).

>>60793069
It was never meant to be a C++ killer. And golang, nim and d failed because of some other obvious flaws.