Re: MDMA server at UNC/SunSITE is broken

Roy T. Fielding (fielding@avron.ICS.UCI.EDU)
Thu, 15 Sep 1994 19:33:30 -0700


charlesa@cosmos.learned.co.uk writes:

> Anyone played with HyperG? Not that it would solve all you mention here, but
> certinly addresses many of the maintaince issues we all are facing, perhaps
> exemplified by this post. I'm wondering who has any experience with HyperG,
> and comments re: WWW.

Yes, I have read the technical specs and the main article on Hyper-G.
Although it looks like a fine system (and may be ideal for many environments),
I do not believe it will scale as well as the WWW has, nor does it support
out-of-context referencing as well as URLs do.

Although the defficiencies with the MDMA server would not have occurred
with Hyper-G, that is only because Hyper-G is much more fragile than WWW
(i.e. it would have fallen on its face, or just refused requests, rather
than allowing the most common requests to succeed) and also has a much
more definitive specification of the protocol than we have for HTTP.

On balance, we are much better off with a loosely-coupled system like WWW.
Simon is a great guy (I'm sorry my original message was so harsh) and I am
very confidant that MDMA will eventually be a full-featured, fully-compliant
HTTP server.

......Roy Fielding ICS Grad Student, University of California, Irvine USA
<fielding@ics.uci.edu>
<URL:http://www.ics.uci.edu/dir/grad/Software/fielding>