IRC channel logs


back to list of logs

<taylanub>Heh, R7RS-small sadly rejected the idea of depracating call/cc. -large will at least have delimited continuations at all.
<cluck>i really hope everybody gets aboard r7rs even if they don't think it perfect all this fragmentation is just silly and self hindering we all gotta scheme together against the lesser langs
<taylanub>I'll "release" "1.2" soon I guess:
<cky>cluck: I'm all for promulgating delimited continuations, but there should be a standard, common name for its procedure. :-)
<cky>cluck: You see, my car plate is CALL/CC, and if I'm going to upgrade it with call/cc's successor, it needs to be just as instantly recognisable as call/cc. ;-)
<taylanub>I really like Guile's primitives for delimited continuations.
<taylanub>shift/reset and the like never stop messing with my head, because it's right "in the middle" of the code, you know ?..
<add^_>I thought shift/reset *was* delimited continuations.
<add^_>or rather: is
<taylanub>It's one way to do delimited continuations, but not Guile's most primitive way of doing them.
<add^_>You mean the building blocks for shift/reset?
<taylanub>Guile has `call-with-prompt' as a primitive. I don't care about the term "prompt," but I like the way the function works.
<add^_>the % and co
<add^_>"prompt" comes from that it looked like a shell-prompt, or something like that.
<taylanub>The concept is similar to a shell prompt: when a program exits, the user, or "control," is thrown back to the prompt.
<add^_>Oh, no
<add^_>% was shift/reset
<tupi>hello guilers, nice: guile-2.0-libs 2.0.9+1-1 just may its way into debian testing
<add^_>call-with-prompt, make-prompt-tag, default-prompt-tag, abort-to-prompt, call-with-escape-continuation and let-escape-continuation. I think those are them.
<add^_>tupi: nice!
<taylanub>Yeah, escape continuations are also very neat. `let/ec' is just like CL's `block'.
<tupi>add^_: made yes :)
<taylanub>Hrm, do you think this will help with Guile's publicity/popularity ? :)
<add^_>Nope ;-) But it makes it easier for debian users ;-)
<add^_>I can't say I don't wish that though. I wish it got more popular, but on the other hand, there would be more pressure on the maintainers.
<add^_>Which is both good and bad :-/
<taylanub>Well yeah, when talking about popularity I'm actually hoping for more contributors, first and foremost.
<add^_>Ah, right
*add^_ thought block in CL was like a let for a function, apparently remembered it wrong.
<add^_>Like flet in CL
<add^_>Oh well.
<add^_>Gotta go soon.
<add^_>taylanub: so what are you up to?
<tupi>taylanub: it helps developers too because guile users report bugs that are [very often] already fixed in git and latest release tarball ...
<add^_>Well, gotta go now, cya guys later!
<taylanub>Hrm, if we had `bytevector-slice', that would be great. I guess it can be implemented via pointers, as it stands.
***bitdiddl` is now known as bitdiddle`