i think patterns.scm should be a pure scheme except for record definitions that
(at least at one point) had to be added from the c side. any code you convert
to pure scheme ill add into the repo.
On Sep 18, 2012, at 10:29 PM, Fernando Lopez-Lezcano wrote:
> On 09/18/2012 10:17 AM, Heinrich Taube wrote:
>>> , I'm trying to bring old code up and running and it needs patterns
>>> ("make-item-stream"!). It does not look like I can get patterns.scm to load
>>> on a plain s7…
>>
>> make-item-stream! wow that really is old!! :) hmm cm2 has (new<cycle>
>> …) maybe you mean that?
>
> Nope, make-item-stream 'items 'heap '( things ...)
> :-)
> Or just (items x y z in heap)
> Very old.
>
>> in the best of all worlds one could simply load up cm2 code into s7/grace,
>> but cm2 used a tiny-clos'ish object system that chicken, gauche and stklos
>> managed to provide in one form or another. if i could mimic this with s7's
>> object system that would work, but i don't know yet if thats possible or
>> what sort of speed issues would arise. its on my todo list.
>
> I ended up hacking patterns.scm enough so it loads and runs (very limited
> testing just for the patterns I was using) in an s7 interpreter. It was not
> that difficult. It would be good to have an "official" version that loads
> into plain s7...
>
> The piece (from 1994) is starting to make noises again with much help from
> Bill.
> -- Fernando
_______________________________________________
Cmdist mailing list
[email protected]
http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
