Maarten: > Has anybody ever looked into Grid computing with REBOL. I imagine a > dialect that can describe the computing resources, input/output > files/URLs, etc. should be easy to implement. Scheduling a bit harder > (because of the OS integration, but with the async-call port...).
Not done anything, but I have given the idea some thought from a different direction. I was looking for a way to safely run user-submitted code on a website. As you say, they would need absolutely secure sandboxing. Depending on how much I trusted your code, I'd have to limit or deny it access to things like: -- 'send -- any reading or writing outside of a sandbox -- quotas for total run time and disk space used -- restricted access to 'system so you could subvert any of the controls in place. I imagined there'd be some sort of "safe" version of do -- something like do-safe user-code [sandbox-dialect] where the sandbox dialect defined what access restrictions apply to the user-code, and what actions to take if it breaks them. The big problems I saw were: 1. We can't always trap all problems real-time (such as runaway code). 2. Running code on behalf of someone else may not be in accord with the current licensing of REBOL It'd be an interesting project to try. Sunanda. -- To unsubscribe from the list, just send an email to lists at rebol.com with unsubscribe as the subject.
