Arc Forumnew | comments | leaders | submit | akkartik's commentslogin

I meant to submit this post when I first saw it, but forgot :)

reply

2 points by akkartik 30 days ago | link | parent | on: Nice work, malisper!

I don't think we can :) This forum seems slow enough that I figured we had time to discuss something off-topic.

reply

3 points by jsgrahamus 30 days ago | link

Well, that is certainly true; although I was surprised at the recent post where all those other folks showed up.

reply

3 points by jsgrahamus 30 days ago | link

http://arclanguage.org/item?id=20096

reply

3 points by akkartik 32 days ago | link | parent | on: At-sign in string

Yes, the atstrings feature is turned off by default in Arc 3.1 (and also in Anarki's official/stable branches). I think I'll turn off the default in Anarki as well, and just let news.arc continue to make use of it.

Edit 1 hour later: This is now done. Turns out Anarki never meant to turn on atstrings by default. This was a side effect of loading news.arc by default (mea culpa). I've changed things so we load news.arc by default but only use atstrings in news.arc by default. Keep an eye out for subtle bugs introduced by atstrings being turned off when functions in news.arc are called. If we see any I'll probably take out the ability to turn it on and off and just decide to keep it everywhere or take it out entirely.

https://github.com/arclanguage/anarki/commit/d0eb0dcf17

reply

5 points by akkartik 33 days ago | link | parent | on: How many people still lurk here?

Great idea, mikhail. Thanks for doing this. I now know of two readers that I wasn't at all aware of before! ^_^

reply


1. Oh, the clean implementation of classic Lambda Calculus syntax is awesome.

2. Your string interpolation implementation is just begging for an injection attack proof of concept :)

3. Why bother with continuations?!! Them's fighting words :)

reply

4 points by conanite 40 days ago | link

1. Thank you! I figured arc could have gone a lot further with special-syntax, by making it configurable inside arc code.

2. this "open source" has had only two eyeballs on it so far ... attacks welcome :)

3. The execution stack is entirely plagiarised from rainbow so it should not be hard to re-implement continuations in the same manner ... the need hasn't arisen yet though ...

reply

3 points by akkartik 41 days ago | link | parent | on: 2D syntax in Racket

Oh, this is cute.

reply

3 points by rocketnia 41 days ago | link

I don't know if I'd ever want to use this 2D syntax directly, but I'll probably "use" it as a thought experiment for non-hierarchical syntaxes, along with cellular automata and spreadsheets. :)

Can you imagine an 'unquote-splicing operation for this tabuleau syntax? Between lists, 'unquote-splicing can insert any list into any other, but the cells of these tables have externally imposed sizes and shapes, so for an 'unquote-splicing in a table (if we wanted one at all), we'd probably want to enforce that the sizes and shapes match up in a way that makes sense. What might make more sense is to do 'unquote-splicing in a full row or column at once, because at least we get one dimension of freedom that way. If the splice takes up a full row, then it can splice in any number of rows (of any height) to that location.

I'm probably a little obsessed with quasiquotation lately. I've been trying to write an implementation of a macro system where the meanings of notations like 'unquote, 'unquote-splicing, and parentheses themselves can be treated as user-defined macros, and where syntax-bound concepts like source locations and syntax highlighting are mostly managed automatically in the macroexpander rather than something every macro must deal with (except for the macros that do something unusual with them).

reply


Welcome back, conanite! I was coincidentally just thinking about you yesterday when I ran across your bug report at http://arclanguage.org/item?id=10485. Too bad that bug is still open in both Arc 3.1 and Anarki. I think I will try to fix it now that I've spotted it.

reply

4 points by akkartik 42 days ago | link | parent | on: At-sign in string

Congratulations, you've run into at-strings, one of Arc's more awkward features.

Prior discussions: http://arclanguage.org/item?id=10245; http://arclanguage.org/item?id=10485; http://arclanguage.org/item?id=18715

Interesting idea to quote a string. I'd somehow never thought of that!

reply

3 points by prestonbriggs 42 days ago | link

Got it, thanks.

I guess the quote idea came from me thinking that the strings were being evaluated. My untrained mind thought "strings evaluate to themselves", but apparently not. So I used the single quote to prevent evaluation.

reply

1 point by akkartik 42 days ago | link

And it worked! Nicely done.

reply

4 points by akkartik 53 days ago | link | parent | on: Anarki is now a Racket #lang

That is great! Welcome back :)

reply


Google Drive? You can share a document so anyone with the link can access it.

Another option is Gitlab or Bitbucket. Put the file in a repository (ideally with any of your code for the paper!) and push it.

I can do it for you if you like. Just email the paper to me. (My address is in my profile.)

-----

More