What’s everyone working on this week?

[Stephen De Gabrielle]

New week, new Racket! What are you folks up to? Answer here or

* https://racket.slack.com/ (Sign up at https://racket-slack.herokuapp.com/ )

* #racket IRC on freenode.net https://botbot.me/freenode/racket/

* Tweet @racketlang on Twitter

* Racket discord 

https://discord.gg/6Zq8sH5

Don’t forget to 

* sign up, or contribute, to Racket News https://racket-news.com/

* submit your links to Racket Stories https://racket-stories.com/

--

----

[Hendrik Boom]

On Tue, Dec 10, 2019 at 08:13:32AM +0000, Stephen De Gabrielle wrote:
> New week, new Racket! What are you folks up to? Answer here or

Still tinkering with Rackettown. I'd love to convert it to typed
Racket, but the association list mechanism is inherently typeless.
(Or, I suspect, only typable in some not-easily-decidable
type-theoretical logic).

Still looking at 3D graphics (you can find some of this in the
Rackettown file 3d-notes). There are a fair number of 3D-related
packages in the Racket package library, but far too many of them
currently don't compile or fail their test sets.

And there seems to be an opengl binding on its own, separate from the
one hidden away in pict3d. I'm no sure how the two are related.

I may find the best alternative for a 3D rackettown is to use openGL
directly. The buffer handling in pict3D and the triangle meshes in
3d-model may be useful. But it's not clear how to relate these meshes
to opengl.

-- hendrik

kk

[Ryan Johnson]

C programming >:(

--
You received this message because you are subscribed to the Google Groups "Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to racket-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/racket-users/CAGHj7-%2ByS-GzsJd7uThFem-ufrPkhSoX6MBifTRaPCG1%3D%3DNCpg%40mail.gmail.com.

--

Ryan Johnson

Free Software Foundation

Electronic Frontier Foundation

[Hadi Moshayedi]

Started creating some slideshows on postgres internals using #lang slideshow: https://github.com/pykello/hadi-slides

Not much content yet, as I've been struggling with getting things formatted as I like until now :) and this is my first time using slideshow, so I'm pretty slow.

[wanderley...@gmail.com]

Started to read HtDP and I plan to training the design recipe while solving AoC challenges.  So far, I am enjoying the systematic approach to right software.  It is funny to realize that you started to solve a problem with a partial understanding of it.

Playing a bit with rsound.  My goal is to have a Beginner Language like that manipulates sounds instead of images.  With that in hand, I could play (literally) in Racket with my musician friend that wants to learn how to program (that is one more reason why I am reading HtDP).

I might work in a re-write project.  My "first" serious program was a program to draw gates (those that go in front of houses).  I did it to my father 16 years ago.  And I decided to re-write it in Racket and see what I get.  This time the gift will go to my brothers who took over the family business.

On Tue, Dec 10, 2019 at 3:54 PM Hadi Moshayedi <ha...@moshayedi.net> wrote:

Started creating some slideshows on postgres internals using #lang slideshow: https://github.com/pykello/hadi-slides

Not much content yet, as I've been struggling with getting things formatted as I like until now :) and this is my first time using slideshow, so I'm pretty slow.

--
You received this message because you are subscribed to the Google Groups "Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to racket-users...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/racket-users/CAK%3D1%3DWqK3r9puu7JdM-aJ5HOE1FZSsrSwxaJvOa6eKELRTyu%3Dg%40mail.gmail.com.

[Christopher Lemmer Webber]

This week through the holidaze are a bit busy for me.

Nonetheless I am trying to understand CapTP, the capability transport
protocol (for a mutually suspicious networked programming environment):

http://erights.org/elib/distrib/captp/index.html

I am trying to port its ideas to Spritely Goblins. I expect the work
will be useful enough by the end of January but we'll see.

[Nathaniel Griswold]

I am almost done with chapter 5 of SICP which i have been working on in racket.

My plan now is to write a scheme compiler/interpreter with garbage collection that runs in native language in racket. I think i will use llvm.

[Hendrik Boom]

On Thu, Dec 12, 2019 at 10:49:01AM -0800, Nathaniel Griswold wrote:
> I am almost done with chapter 5 of SICP which i have been working on in
> racket.
>
> My plan now is to write a scheme compiler/interpreter with garbage
> collection that runs in native language in racket. I think i will use llvm.

So you'll be using LLVM directly? Or using Racket to generate LLVM code?

-- hendrik

[Nathaniel Griswold]

Racket to generate llvm

> On Dec 12, 2019, at 3:35 PM, Hendrik Boom <hen...@topoi.pooq.com> wrote:

> --
> You received this message because you are subscribed to the Google Groups "Racket Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to racket-users...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/racket-users/20191212213513.rqi6zplqs63vc4ty%40topoi.pooq.com.

[Sam Tobin-Hochstadt]

You might be interested in our project Sham:
https://github.com/rjnw/sham which is a Racket interface for easily
using LLVM.

Documentation coming soon, but you can see some examples here:
https://github.com/rjnw/sham/tree/master/test

Sam

> To view this discussion on the web visit https://groups.google.com/d/msgid/racket-users/0F71B923-2FEA-49AB-BA66-B6E31A723A6A%40gmail.com.

[Anthony Carrico]

On 12/12/19 10:05 PM, Sam Tobin-Hochstadt wrote:
> You might be interested in our project Sham

So is this compiling a state machine (or, two actually) to llvm and
running against a Racket oracle?
https://github.com/rjnw/sham/blob/master/test/automata.rkt

--
Anthony Carrico

[Sam Tobin-Hochstadt]

Yes, basically. I'd say it's showing how to translate that Racket code
(taken from Shriram Krishnmurthi's JFP paper/LL1 talk) into Sham.

Sam

> --
> You received this message because you are subscribed to the Google Groups "Racket Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to racket-users...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/racket-users/c619eae3-fac5-9a8f-a35a-dbf5a6921839%40memebeam.org.

[Anthony Carrico]

On 12/12/19 10:36 PM, Sam Tobin-Hochstadt wrote:
> Yes, basically. I'd say it's showing how to translate that Racket code
> (taken from Shriram Krishnmurthi's JFP paper/LL1 talk) into Sham.

Nice. Looking forward to less brief summary of the tool.

--
Anthony Carrico

[Nathaniel Griswold]

Regarding your question of using llvm directly. Do people do that? I am going to figure out as I go but was planning to write the garbage collection part of the kernel in C and the procedure application part in machine language. Could I just write llvm directly? (This is my first time using llvm)

> On Dec 12, 2019, at 9:06 PM, Sam Tobin-Hochstadt <sa...@cs.indiana.edu> wrote:
>
> You might be interested in our project Sham:

[Anthony Carrico]

On 12/13/19 12:26 AM, Nathaniel Griswold wrote:
> Could I just write llvm directly? (This is my first time using llvm)

Sure. You should probably play around and write some llvm assembly
procedures by hand.

I think the main issue with llvm is, or used to be, that you can "call
with arguments" (that is a normal C procedure call), but you can only
jump within a single procedure, so it isn't convenient to modularize
state machines like we can with tail calls in Racket.

Some compilers create special llvm calling conventions to deal with
this, but it would be nice if there was a standard "jmp with arguments"
instruction (which cleaned up the stack) that you could use to escape
into a stackless mode of operation.

--
Anthony Carrico

[Nathaniel Griswold]

Oh, interesting. Yes the tail call optimization for my interpreter was one thing I need to hand-code. It will be a good learning experience.

> On Dec 13, 2019, at 9:05 AM, Anthony Carrico <acar...@memebeam.org> wrote:

> --
> You received this message because you are subscribed to the Google Groups "Racket Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to racket-users...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/racket-users/6e4191ed-2a54-0a49-f6cb-e5912fa24183%40memebeam.org.