W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > May to August 1995

Re: Byte ranges -- formal spec proposal

From: Jon Knight <J.P.Knight@lut.ac.uk>
Date: Thu, 18 May 1995 16:32:56 +0100 (BST)
To: Gavin Nicol <gtn@ebt.com>
Cc: cshotton@biap.com, luotonen@netscape.com, www-talk@w3.org, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Message-Id: <Pine.SUN.3.91.950518161810.187F-100000@weeble.lut.ac.uk>
On Thu, 18 May 1995, Gavin Nicol wrote:
> Given that users expect them to be supported, they'll be surprised
> when a request for 500 bytes returns 50000 instead, or if the server
> returns an error code. They'll be more surprised to find that each
> time they request the same 500 bytes, they could, possibly, get 50000
> *different* bytes back. 

Given that you're basically saying that your server already reuses URLs
for different objects or representations of objects anyway, why would a
user be more surprised that a URL with some byterange magic on the end
returned a different 50,000 bytes each time they used it than a URL
without the byterange stuff?  Does DynaWeb support the NCSA httpd style 
~user stuff?  If not, do you get people emailing you in surprise when 
URLs containing that sort of thing don't work?  To my mind ~user and 
;byterange are the same kind of thing; handy conventions that some, but 
not all, servers implement.

If people know enough to be fiddling about with byte ranges then they
probably know enough to expect non-compliance or odd behaviour from some
servers.  People aren't generally stunned when a CGI script returns
different data for an identical URL on each invocation.  In fact I'd go as
far as to say that many people don't even understand URLs at all
(especially if places like libraries have turned off the browser's URL
display ``so as not to worry people''). 

I don't think most day-to-day users will give a fig whether your server
supports byte ranges or not quite frankly.  IMHO the vast bulk of users
either just click on hotlinks or type in what they're told to.  Those who
hunt about and fiddle with URLs quickly learn how far they can push things
and not to be surprised by anything on the Web. :-)

(In case anyone hasn't guessed, I think byte ranges are a good idea - I 
liked the spec.  Especially seeing as there was some thoughts about 
caching in there.)


Jon Knight, Researcher, Sysop and General Dogsbody, Department of Computer
Studies, Loughborough University of Technology, Leics., ENGLAND.  LE11 3TU.
*** Nothing looks so like a man of sense as a fool who holds his tongue ***
Received on Thursday, 18 May 1995 08:46:54 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:40:14 UTC