RE: [xml-dev] MS thinks HTTP Needs Replacing???

From
Bill de hÓra <>
To
Date
2002-02-27T15:34:12
ID
<002d01c1bfa4$38e9e940$887ba8c0@mitchum>
Thread
RE: [xml-dev] MS thinks HTTP Needs Replacing???
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


> -----Original Message-----
> From: Simon St.Laurent [mailto:] 
>
> If you ask whether HTTP is an implementation of RPC, a 
> protocol built on RPC, no.  I'm not talking about hypertext - 
> I'm talking about the foundations on which the HTTP protocol 
> is itself built.  I find those foundations to look 
> suspiciously related to RPC, however loud the denials of their
> fans.  

Sensible engineers build RPC on top of asynchronous unreliable
protocols. So you start with with async/unordered base and layer
sync/ordered on top; ie RPC can be baked as a set of guarantees and
constraints about messages between processes. If similar guarantees
can be identified in HTTP then you've got a case.

Bill de hÓra

-----BEGIN PGP SIGNATURE-----
Version: PGP 7.0.4

iQA/AwUBPHz8cOaWiFwg2CH4EQLMCgCfUnBj055KSS8J85yOZhiSlLUHZTwAoPDN
i1mYU6CifK73W9THHK1dewaP
=QtDr
-----END PGP SIGNATURE-----