W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > January to April 1999

Re: Resend: Re: IPP> Chunked POST: SUMMARY

From: David W. Morris <dwm@XPASC.COM>
Date: Tue, 26 Jan 1999 14:28:58 -0800 (PST)
To: Chris Newman <Chris.Newman@innosoft.com>
Cc: http-wg@cuckoo.hpl.hp.com, ipp@pwg.org, http-wg@hplb.hpl.hp.com
Message-Id: <Pine.GSO.3.96.990126141416.6495C-100000@shell1.aimnet.com>


On Tue, 26 Jan 1999, Chris Newman wrote:

> On Sun, 24 Jan 1999, Roy T. Fielding wrote:
> > >Should every
> > >application that uses HTTP as a transport build in its own HTTP
> > >implementation, even if one is already available on the platform?
> > 
> > No application should use HTTP as a transport.
> 

As an absolute statement, that's absurd. Everything done with HTTP is
running applications. It is myoptic to claim that a CGI program is
anything other than an application which uses HTTP as the transport.
The CGI program in conjunction with its host HTTP server must follow
the HTTP protocol to transport requests to the server and responses
back.

The issue is to define the characteristics of applications which are
appropriate for implementation using HTTP.

Conversely, given the availability of a high quality HTTP implementation
for the application context, finding a need to replace that
implementation would make me wonder if HTTP is an appropriate transport
choice.  OTH, there a lot of parameters to optimize and some 
applications may be better served with a custom HTTP implementation.
For example, if the application is to serve web resources based on a
RDBMS, one might make different implementation choices than if the OS file 
system is the data respository.

Dave Morris
Received on Tuesday, 26 January 1999 23:06:26 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:33:30 EDT