- From: Jon P. Knight <J.P.Knight@lut.ac.uk>
- Date: Wed, 1 Feb 1995 13:21:42 +0000 (GMT)
- To: uri@bunyip.com
- Cc: kball@kballuw.SJF.Novell.COM
In message <9501311929.AA04195@ka.SJF.Novell.COM>, kball@kballuw.SJF.Novell.COM (Keith Ball) writes: > > We have been working further on creating a single HTTP URL that > refers to a document on an HTTP server that is accessible via multiple > transports and is named within different naming services, such as DNS, > NetWare Directory Service (NDS) and X.500. Why don't you just define a new method such as nhtp: that is designed to work over IPX? Surely hacking in multiple transports into the existing URLs is a bad idea as you'll just have to do it all over again when you move to IP:ng along with the rest of us (you are moving right? :-) ). And not to mention that there's more to the WWW than HTTP; are you planning on coming up with IPX versions of FTP, gopher, NNTP, etc, etc and change those URLs as well? If you used a new nhtp: method to denote Novell's Hypertext Transfer Protocol, we Internet types could just set up proxy gateways that took your nhtp: URLs encoded inside a normal http: URL and handed them over to the Novell side of things for name resolution, binding and communication. You could even do the reverse for IPX clients (ie: configure them to bundle all the traditional URLs inside an nhtp: URL and shove it at the Novell/Internet gateway). You can stick what ever wild and wacky naming schemes and formats you like inside nhtp: URLs then without the rest of us worrying too much. As far as I can see that would work, would let you do caching in the proxy gateway at the boundary of the two universes, would be invisible to clients on both sides and wouldn't need to have the existing HTTP URL spec and WWW browsers fiddled with. Comments? Jon -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Jon Knight, Research Student in High Performance Networking and Distributed Systems in the Department of _Computer_Studies_ at Loughborough University. * It's not how big your share is, its how much you share that's important *
Received on Wednesday, 1 February 1995 08:58:17 UTC