Re: DSSSL extensions for XSL

[christo@xxxxxxxxxxxxxxxxxx (Frank Christoph)]

> > > An upwards-only continuation is one that can only be used to return to a
> > > stack frame in the current call chain. When the call/cc call that
> > > created a continuation has returned, the continuation can no longer be
> > > called.
> > 
> > Then that's call/ec.
> 
> I've never come across that name.  What Scheme systems provide a
> procedure named call/ec?

Hm, well I don't claim to be a Scheme expert and certainly not a historian.
Perhaps my impression that it was a widely used name was mistaken.

I know that Rice's MzScheme, which is the underlying interpreter in PLT's
DrScheme package provides it.  I tried checking a few other implementations
like Chez Scheme and Gambit over the web, but couldn't get a hold of their
library manuals online.  The other implementation I have installed here,
SCM, doesn't seem to provide it.

> > I don't think it makes sense to call this "upwards-only"
> 
> I didn't make this term up; I'm only using the terminlogy that I've come
> across in the Scheme community.  The Kawa docs use this term, and that's
> certainly not the only place I've come across it.  It makes a certain
> amount of sense to me because the limitation on upwards-only
> continuations is that they can only be used to return up the stack.

I'm not saying you made it up.

I did a little check of the literature.

Luc Moreau uses "upwards" and "downwards" in the sense I used them in both
"An Introduction to Continuations" and in his thesis, "Sound Evaluation
of Parallel Functional Programs with First-Class Continuations".

Kent Dybvig uses the terms "inwards" and "outwards" in his dissertation,
"Three Implementation Models for Scheme".

Daniel Friedman, Mitchell Wand and Christopher Haynes use "upwards" and
"downwards" in the sense I used them in "Essentials of Programming
Languages".

I think that Matthias Felleisen and Robert Hieb use the terms in my sense
in "The Revised Report on the Syntactic Theories of Sequential Control
and State," (but I'm not sure).

I didn't come across any papers or references that use the terms in the
opposite sense.  (I didn't check Kawa.  I trust you.  :)

Well, whatever... it's a trivial point.  The best thing is probably to
distinguish between the two as "escaping" and "first-class" continuations.
(Although, there is a problem even there because Reynolds originally
introduced first-class continuations using the keyword "escape"!)

--FC

 DSSSList info and archive:  http://www.mulberrytech.com/dsssl/dssslist