Re: warning to information providers re HTTP/1.0

• Mail folder: WWW Talk Jul-Oct 1993
• Next message: Kevin Altis: "converting URLs in .html files"
• Previous message: Nathan Torkington: "Re: warning to information providers re HTTP/1.0"
• In-reply-to: Nathan Torkington: "Re: warning to information providers re HTTP/1.0"

Date: 30 Aug 93 17:02 -0700
From: George Phillips <phillips@cs.ubc.ca>
To: www-talk@nxoc01.cern.ch
In-reply-to: <199308302342.AA27591@kauri.vuw.ac.nz>
Message-id: <6202*phillips@cs.ubc.ca>
Subject: Re: warning to information providers re HTTP/1.0
Status: RO

Nathan Torkington said:
>Anyone want to write code that uses a `file' like database to work out
>the type of the file, so the server can correctly identify file types
>without needing to look at their extension?

Grab file://ftp.cs.ubc.ca/ftp/local/src/hfile
If you can grok perl, it will tell you how to identify lots of different
file types.  I'd really like to put this functionality into the
browsers.  Even though HTTP/1.0 will partially solve the problem, there's
still FTP sites and local files which have no other type information.

Another warning to HTTP/1.0 information providers.  There is a bug in
the libwww 2.0 libraries such that you better make sure you do a
fflush() after outputting your HTTP/1.0 reply header.  It only bites
if you're sending back a binary file (say, a GIF image) that has a
0 byte near the start.  I sent off a bug report about this, but
haven't heard if the bug has been fixed.

				-- George