Re: redo-safe-variables and redo-safe-parameters

[Stefan Israelsson Tampe <stefan.itampe@gmail.com>]

> Daniel Writed
> Dynamic states are not suitable for the purpose.  They have nothing to
> do with compenstating for the inability of continuations to backtrack
> _through side-effects_.  I believe this will be obvious if you
> consider the problem of side-effects generally, rather than focusing
> only on variable assignment.
Yes, I try now to avoid dynamic states. I was a bit ignorant of the
subject. But after some reading and thought your point has become 
clear to me.

> Backtracking is typically handled (at least, in part) by the
> evaluator, by either:

> - explicitly tracking side-effects, so that they can be reverted in a
> sensible manner; or

> - state-copying, that is, non-mutable environments.
Yea I have thise two features on my mind when trying to design a spec.


> I do not see how you can hope to marry the concepts of continuations
> and backtracking side-effects without modifying the evaluator, at
> which point you have continuations and an evaluation environment that
> is not Scheme, although perhaps very Scheme-like.
Note that there is a guard that checks if you should restore
or not. If that executes always to #f then everything is scheme, so it
is an extension. Actually I'm not sure if I need to change the
evaluater to get redo-safe-variables I beleve that you can get the
features by simply modify call/cc with current scheme. But for 
redo-safe-parameters I do not know if the same holds. Also the latter
part of the spec handling set! and set~, AS Noha stated in his last
mail. We might just drop that part and have a recomendation for macro 
writers to follow a certain pattern instead.

> It seems your real objective is to extend Scheme-embedded logic DSLs
> by supporting continuations and non-functional Scheme code within
> them.  I appreciate that you have some experience in the area, can you
> point to any papers that discuss anything similar to what you are
> trying to achieve?  (Not the Scheme modifications, but the logic DSL +
> side-effects + continuations).
I'm afraid that I have not seen any papers on this. But I'm
not in academia and have not a great overview of the subject. Maybe I
should write a paper about guile-log, maybe I should try to dig up a
references. Maybe I should documnet guile-log at a very detailed level. 

> Back to the Scheme modifications.  Perhaps I do not understand that
> problem space as well as you, but when I look at this I see a
> premature attempt to solve a problem that is _hard_.  There is also no
> precedent for continuations that backtrack side-effects in any Scheme
> or Lisp I know of, and noone will miss that if you do not acheive it.

> Clearly you are spending some effort on this, and I do not like to see
> anyone wasting efforts.  IMO this specific path is unproductive.
Yea that path is unproductive, right now I pursue other paths.

To be onest, I do this mainly to learn. But also because I find it 
interesting and useful. The hard part actually seams not to implement 
something that are redo safe. The hard part is to do it in a way so
that you will get easy to reason about code and possible efficient 
imlementations. Much of the confusion is attributed to this.

/Stefan