langreiter.com plain, simple: vanilla-mesh

LANGREITER.COM plain, simple

START

INDEX

vanilla-mesh

CREATED BY james • LAST EDITED BY james 8277 days AGO

a community of linked vanilla sites.

goals:
reduce duplicate information
introduce a form of automatic concept linking between sites

why not look here for a demo?

So this is like www.c2.com/cgi/wiki?InterWiki?

not quite. the idea (as far as i see it) is that most vanilla authors have their own seperate views on certain subjects, but links to other authors' view can be created automatically. the technique, given appropriate query techniques, can even suggest other snips which might be related - in a sense creating more of a vanilla-internet (linked but seperate sites) rather than the InterWiki described above (clear partitioning of subjects, portalling/load balancing). - james

I would like to make my tool vanilla-mesh compatible, what do I need ? I talked to a friend some days ago about wiki linking and finding related wikis in other spaces. What a coincidence.
- funzel

it's evolving, but basically the idea is to allow different sites access to meta-information from other sites. for instance, lazyatom.com asks langreiter.com about the existence of any snips with the same name as the snip currently being displayed, and if one exists, it creates a link.

the only change need to the vanilla core is an additional selector - snip-query (please note - all these names are stuff i just pulled out of the air, there are probably better names and better ways to do all this :). currently in the ultra-alpha implementation we have, it just returns a rebol series holding all the metadata for that snip, if it exists (or [none> if the snip doesn't exist)

next, what should be done is the creation of a dyna which uses rebol's read rejoin ["http://www.langreiter.com/exec/vanilla.r" "?selector=query-snip&snip=" snipname] to determine whether or not the exists on the remote site. easy-peasy.

of course, this could be made so much better - restrictions on the metadata returned might be desirable, along with different types of query - the remote sites might be able to suggest other snips which are relevant to the query, or anything you can imagine!
- james

Well, the only problem is that our tool is not Vanilla (but written in Java, borrowing heavily from vanilla) :-)
- funzel

that's not a problem :) so long as there is some shared data interchange format then any site should be able to query any other site. it's a bit like some mad active RSS or something.

What do I get as a result of the URL lookup ?
- funzel

currently, it's just the metadata of the vanilla snip, i.e. [displays 0 last-store ......]. if the snip doesn't exist, it returns an empty REBOL list, i.e. []. but any suggestions for some universal format (XML, you say? RSS, do i hear? ;) are most welcome - this 'thing' is totally organic, just an infant of an idea...