Hi, I just saw this thread and I have to say, your requirements are really interesting.
On 09/23/2013 03:08 AM, Richard Kulisz wrote: > Hi there, > > I'm investigating whether this wiki engine is right for me. I'm > looking for three specific design features: > > 1. able to attach autonomous agents to pages > > 2. transclusions / includes. > > 3. capability security model rather than ACLs Zed Shaw discussed this in "the ACL is dead" IIRC, he was using snippets of ruby (obviously touring complete) to express security rules. Bitcoin is of course another notable example since the actual payments are expressed as bytecode and the "payment claims" are inputs which cause that code to return true. > > features #3 + #1 can easily be made to enact: > 3a. patron blindness - users can become patrons of others and their > own patron is blind to that, but must deal with the entire > sub-hierarchy of users as a collective > > which makes this other feature rather important: > 3b. patron multiplicity - users can acquire multiple patrons just in > case their first patron decides to eradicate them I find this use case particularly interesting since it seems you're trying to implement the same friend-to-friend model that's done with ISP peering in Internet routing. This is a very scalable model which I have spent some time studying before beginning on cjdns. The reason I think why these things don't exist is because they can't outperform dictatorships at small scale. When a programmer implements one of these brilliant ideas, the first words out of the customer's mouth are "how can I make a rule so only management group can read these pages?" and we end up using this technology to emulate the ACL. What the customer comprehends is what he asks for, what he asks for is what the programmer must provide and what the programmer provides defined the customer's view of what is possible. Of course it's a bit of a catch 22 but still the programmer has a window of opportunity to define new possibilities through his solution to the customer's problems. So then the question becomes how to redefine the features needed to solve these problems which happen at scale in terms of use cases which people know they need. As far as actual answers I'm afraid I don't have any. Thanks, Caleb > > Oh and I suppose: > 4. automatic garbage collection and revisions even of deleted pages > > > If you know of another wiki engine that supports these better, please > let me know. I know that Wagn supports transclusions but I'm rather > doubtful that anyone supports a capability security model. Even though > ACLs really should have died out in the 70s. > > > Thanks for reading this far, > > Richard > _______________________________________________ > users mailing list > users@xwiki.org > http://lists.xwiki.org/mailman/listinfo/users > _______________________________________________ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users
