In-response-to-base:
In-response-to:
Date: Fri, 20 Oct 1995 20:47:57 GMT
From: Linas Vepstas (linas@austin.ibm.com)

Resource Naming

HTTP/1.1 200 OK Date: Mon, 21 May 2012 09:14:21 GMT Server: Apache/2.2.6 (Fedora) Last-Modified: Thu, 30 Oct 1997 19:45:03 GMT ETag: "29c174-219-59212dc0" Accept-Ranges: bytes Content-Length: 537 Connection: close Content-Type: text/html; charset=UTF-8 Dave Ragget mentions use of HTTP to implement remote compute resources. Interesting idea ... but...

Say I have set up 100 web servers, each web server has a ray-tracer residing in it's cgi-bin directory.

How can I set up a CORBA/IDL/URI/URN/whatever, that would "name" all 100 of these resources with a single name? The goal is not to find a specific resource (a raytracer) but all resources of a type (all raytracers taking a specific format) ... so that I can implement a distributed processing "compute farm".

-linas