Re: multi-host virtual sites for HTTP 1.2

Paul Hoffman writes in <v03007607ae08a187ad5b@[165.227.113.247]>:
>This misses a fair amount of the advantages suggested by Paul. For one,
>every host in the round-robin would have to have all of the same content;
>Paul's suggestion allows partial redirection. Thus,
>http://www.megacorp/root1/... could be redirected to one host that has just
>that content, while http://www.megacorp/root2/... could be redirected to a
>different host with different content.
>
>Further, Paul's suggestion makes it easier to add redundancy.
>http://www.megacorp/root1/... could point to two or more hosts that are at
>very different locations. If host1 is down because of hardware or network
>failure, a good client would try host2 and so on. Of course, these hosts
>have to try to have the same content, but it would be much less (and
>therefore more likely to be the same) than for the entire MegaCorp tree.

This feels to me like an extension of the UNIX filesystem paradigm to Web 
document hierarchies ("The UNIX Time-Sharing System", D. M. Ritchie and K. 
Thompson, The Bell System Technical Journal, July-August 1978):

     "A hierarchial filesystem incorporating demountable volumes"

except that the document hierarchy  would be spread among multiple disk 
drives, computers, and geographic locations -- altogether a Good Thing.

Although this would gain the Web some of the advantages of URNs, it still 
would not provide the "eternal global document ID" feature that will be a 
central advantage to URNs.  However, I suspect the URN researchers will be 
able to leverage off of the knowledge to be gained by implementing Paul et. 
al.'s ideas for redundancy and document hierarchy spreading.

<rant>
Something that I have sorely missed in my use of VMS, Microsoft DOS/Windows 
3.1, and MVS is the transparency of "a hierarchial filesystem incorporating 
demountable volumes".  Users (including programmers) should just not have to 
care about precisely where their documents are located, except in 
low-resource situations.  If it could be managed, they shouldn't have to 
care about precisely which processor is serving their needs at that moment, 
either.  The original idea behind filesystems was to allow programmers and 
users to access their data without knowing precise platter/track/sector 
locations of their data.

Although multi-host virtual sites (MHVS?) has some aspects of a research 
project, the proposed implementation is simple enough that it could be 
implemented within 1 or 2 browser revisions (as the browser does the work). 
 Further work along this line could result in a world-wide document web that 
also looks like a set of hierarchies, where the hierarchies differ from each 
other depending on where you start (two hierarchies might even contain each 
other).  This would require an extension of MHVS to permit referrals to 
different domains, which would involve some sort of cryptographic 
authentication scheme.
</rant>


As Paul Leach said in his original message:
>This proposal is nothing more than the application to HTTP of the tried
>and true referral mechanisms of distributed directory and file systems.
======================================================================
Mark Leighton Fisher                   Thomson Consumer Electronics
fisherm@indy.tce.com                   Indianapolis, IN

Received on Friday, 12 July 1996 05:19:10 UTC