Re: ATS3 and stuff

Timmy Jose

unread,

Jun 28, 2021, 8:58:55 AM6/28/21

to ats-lang-users

Hello,

I had dabbled with ATS2 a couple of years back, and I happened to have a look back at the ATS ecosystem a couple of days back. It is interesting to see ATS3 being actively developed, and it's also exciting to see that ergonomics is a key focus for this version of the language (this is what I gleaned from some of the posts here).

About learning the language (I've unfortunately forgotten most of it), I looked at the learning materials again, and they are of very good quality, but I was thinking that someone from the community could create something from their own perspective, for beginners. That would also help get people interested in the language - a more gentle introduction to the core selling points of ATS (or at least the basic viable features of ATS). What do you think?

I know the community is small, but time and again I keep thinking that ATS has so much potential that it's a shame that it's not getting much more attention from people (a lot of it has to do with the fact that most people are not even aware of its existence).

Cheers,

Timmy

Timmy Jose

unread,

Jun 28, 2021, 9:00:00 AM6/28/21

to ats-lang-users

Also, in any case, I think I'll wait for ATS3, and this time go through full learning process, and start doing some projects in it. I'm excited, and hopefully it won't be too long coming!

gmhwxi

unread,

Jul 4, 2021, 5:27:11 PM7/4/21

to ats-lang-users

Thanks for your interest in ATS3!

I am still in the midst of implementing ATS3. More precisely, I am

in the midst of implementing a type-checker for dependent types

and linear types. The required effort seems to have far exceeded

what I expected.

Up to ATS2, ATS was largely a vehicle for experimentation.

If I remember correctly, ATS0 (that is, ATS/Proto) is probably the most

feature-rich one among all editions of ATS so far; it contains many features

(e.g., OOP and typed meta-programming based on macros) that are removed

in later editions of ATS.

As I feel that ATS3 is very likely the last edition of ATS that I will ever be implementing,

I tend to think a lot more before I actually write down code. So the implementation

does take a lot longer this time. The good news is that I should be able to move a bit

faster once I finish this type-checker, which undoubtedly is the most challenging part

of implementing ATS3. Very optimistically, I expect to see some version of ATS3 that is

able to compile to C/C++. Keep tuned :)

Cheers!

--Hongwei

gmhwxi

unread,

Jul 4, 2021, 9:24:19 PM7/4/21

to ats-lang-users

>>Very optimistically, I expect to see some version of ATS3 that is

>>able to compile to C/C++

Sorry, I meant to say that I expect to see it by the end of this year.

Timmy Jose

unread,

Jul 5, 2021, 11:37:42 AM7/5/21

to ats-lang-users

Hi Hongwei,

Thank you for the detailed response!

This part especially fascinated me:

>ATS0 (that is, ATS/Proto) is probably the most

>feature-rich one among all editions of ATS so far; it contains many features

>(e.g., OOP and typed meta-programming based on macros) that are removed

>in later editions of ATS.

> Sorry, I meant to say that I expect to see it by the end of this year.

And I absolutely understand (though I cannot appreciate it fully since I am not very familiar with the domain) the amount of time and effort involved in the project, so much respect! Excited and eagerly awaiting the release of ATS3! :-)