Tal is the programming language for the Uxn virtual machine (2021) | Hacker News Reader

Tal is the programming language for the Uxn virtual machine (2021)

wiki.xxiivv.com

84 pointstosh18 comments

This is a FORTH-like language by a programmer-artist who aspires to live a minimalist nomadic lifestyle of self-reliance (resides on a boat with his partner, no permanent home, sold most material goods), and who creates minimalist computer art(ifacts).

An attempt to build up a simple system with tools from scratch combined with a flair of "retro"/80s, but more out of artistic aspirations than to build something useful, as far as I remember.

Yep, they were the "Canadian couple" I mentioned in this comment: https://news.ycombinator.com/item?id=39586904

Artisanal software indeed.

Oh, did they win the “my shares have vested” lottery? I didn’t know that, honestly. It doesn’t seem like they are living any sort of luxurious lifestyle on that boat.

I dunno if they have or not, but they must have had a nice little sum stashed away to buy a boat and then retire on it, spending their days coding whatever they want. To me that alone is a luxury and they are in a very enviable position.

In any event, this thing they’re doing is cool. The world definitely needs alternative approaches to computing.

Devine had a great talk about the "why" of Uxn at Strange Loop last year:

https://www.youtube.com/watch?v=T3u7bGgVspM

And there was another talk that used it as an example:

https://www.youtube.com/watch?v=umSuLpjFUf8

Both were great. I'm a big fan of 100R's work. Another language of theirs worth checking out is Orca: https://100r.co/site/orca.html

I guess the original TAL is not making a comeback:

https://en.wikipedia.org/wiki/Transaction_Application_Langua...

I was wondering if somebody would remember the original.

I programmed in TAL many moons ago at a bank - a MIPS Tandem system running Guardian OS.

I did a tiny bit in the late '90s at Mastercard. Wrote my share of TACL macros before that.

When I first saw the "99 Bottles of Beer" site, I was working for a company the was using the POSIX-layer OSS and C. I was able to "drop down" into Guardian and re-learn enough to contribute: https://99-bottles-of-beer.net/language-tandem-tal-437.html

We ran BASE24 at the time (EFT switch) and customized parts of it.

Using a stack as input and a stack as output reminds me of the POP-11 language.

A high level value stack available as an input is a truly different way to code.

See other Hacker News articles https://hn.algolia.com/?q=pop-11

For me, it's a bit difficult to get the quick overview of what they are doing from their website. Much quicker to check the git page:

https://git.sr.ht/~rabbits/

Lots of very cool stuffs :D

We need some kind of ISO standardisation of language syntax. All new languages must have the exact syntax for standard constructs, so that anyone can write useful programs with the one syntax they know. The languages may differ around the edges.

It is getting out of control now. Every designer feels compelled to come up with a new way of defining a function, etc.

Why should we take that much care regarding the least important part of a language? Every syntax that was invented for actual use (not Brainfuck et al.) can be learnt in short time.

Language semantics and libraries are more important.

If it reduces cognitive load, why not?

A positive side effect could be that all those millions of tutorials that teach syntax in the name of teaching programming disappear.

You have a point here and I appreciate it.

But I still have to disagree: Standardization is a good idea for topics that are "done". But I strongly believe that we still do not know the best syntax, the best way of programming in the general.

Tal is obviously not a language that wants to replace your mainstream language of choice. Its syntax is not totally original, but is inspired from old languages like Forth. I believe we need such experiments to keep old but not (yet) mainstream ideas alive and to foster innovation.

Tal has advantages for the use case it was designed for, that would be difficult or impossible to achieve with more "standard" syntaxes. What is not to love about that?

Maybe experiments like this will someday lead to the inclusion of new ideas into our mainstream languages, like lambda expressions/closures have become normal over the last decade?

Good points. I have changed my opinion :)

Tal is the assembly language of the uxn CPU, it doesn't abstract any mechanical operation from the virtual machine, it's not a high level language but it could be a target for one. For instance, this is a C compiler that targets Uxn instructions, disassembling the artifacts will give you back uxntal code.

https://github.com/lynn/chibicc

Reply on news.ycombinator.com