HTTP/1.0 and noncompliant servers...

• Mail folder: WWW Talk Jul-Oct 1993
• Next message: Steve Romig: "HTTP/1.0 and noncompliant servers... "
• Previous message: Kevin Altis: "Re: converting URLs in .html files"
• Reply: Steve Romig: "HTTP/1.0 and noncompliant servers... "

To: www-talk@nxoc01.cern.ch
Subject: HTTP/1.0 and noncompliant servers... 
Date: Wed, 01 Sep 1993 10:21:29 -0500
Message-id: <8996.746896889@apricot.ucs.indiana.edu>
From: "William M. Perry" <wmperry@apricot.ucs.indiana.edu>
Status: RO

Hello all,

   This is just a simple request for everyone developing/running
HTTP/0.9 servers that are semi-broken.

   There are several HTTP servers out there that don't understand
HTTP/1.0 requests, and don't strip off the HTRQ from the GET command.
(Most notably the SLAC searchable index, and the info gateway at
ohio-state)

   With the movement of all known browsers to HTTP/1.0, this will get
to be a problem pretty quick.  The line mode browser, lynx, and my emacs
browser all support HTTP/1.0, and soon mosaic will.  It will be
annoying and confusing to the users when all of a sudden they can't
access these servers (lynx returns a blank page when trying to access
http://www.cis.ohio-state.edu:85/info/dir, same with any http/1.0
browser - it thinks you are trying to access HTRQ.info).

   All that is needed is for the servers to look for the first white
space in a request, not the end of the line.

   This will make the transition to HTTP/1.0 much easier - right now I
store a list of 'bad' servers that don't understand HTTP/1.0, but I
don't think that is the right way to go.

   Thanks,
      Bill P.