[CM] Saving LISP objects to disk

[Ralf Mattes]

On Tue, 2008-03-25 at 11:19 +0000, Bret Battey wrote:
> One thing that has cropped up as I am trying to resurrect my PICACS
> code in SBCL is that the code I developed to save objects (and any
> subobjects) to disk is broken, because of all of the CCL calls I used.
> 
> I there a standard LISP way of writing objects to disk (and later
> restoring them), including all subobjects?

 Besides Rick Taubes answer: one place to look for such information
would be http://www.cliki.net/. You might want to search for 'object
persistence'. Elephant gets mentioned every once in a while in c.l.l.

> As it was, I solved it three years ago by writing code that determines
> the slots of any object, traverses the whole object and subobject tree
> of a given object and writes out to disk the code needed to recreate
> the objects.
> 
> Surely there is a better way, and one that doesn't require
> system-specific calls (?) -- and three days of hunting poor
> documentation to figure out how to translate my CCL implementation to
> SBCL?

You could read Arthur's essay-code on object persisctence here
http://www.pentaside.org/paper/persistence-lemmens.txt
http://weitz.de/eclm2006/rucksack-eclm2006.txt
I don't know the current status of rucksack but it looked very
promissing last time I checked.

 HTH Ralf Mattes

> Cheers,
> -=Bret
> 
> 
>