I think Jonathan Sewall wrote: [Charset iso-8859-1 unsupported, filtering to ASCII...] > Just to be sure, since I need backtracking for my algorithm, too, I want > to confirm my understanding that there is no generalized undo function > for Jess. That is, e.g., there's no built-in, general way to undo the > effects of an RHS firing. That right? As you might know by now, I use > bsave-bload for backtracking, but I appreciate other suggestions (like > this one). Thanks again, >
As an aside, I have to point out that complete rollback of a rule firing isn't possible because the RHS of a rule might do many things that aren't undoable, from printing a message to the user to exiting the JVM. But in any case, the next release of Jess 7 (tomorrow or Friday) has a mechanism for "marking" working memory and "rolling back" to that mark. All it can do, though, is retract new facts asserted since the mark, not undo modify or retract operations. Still, this marking capability is very useful if you have, for example, a few thousand "catalog-item" facts that you'd like to load just once, and then start just at that point for a series of per-customer transactions. It could also definitely be used for backtracking if the deductive rules didn't modify existing facts. --------------------------------------------------------- Ernest Friedman-Hill Advanced Software Research Phone: (925) 294-2154 Sandia National Labs FAX: (925) 294-2234 PO Box 969, MS 9012 [EMAIL PROTECTED] Livermore, CA 94550 http://herzberg.ca.sandia.gov -------------------------------------------------------------------- To unsubscribe, send the words 'unsubscribe jess-users [EMAIL PROTECTED]' in the BODY of a message to [EMAIL PROTECTED], NOT to the list (use your own address!) List problems? Notify [EMAIL PROTECTED] --------------------------------------------------------------------
