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

What to do on byte ranges?

From: Martijn Koster <m.koster@webcrawler.com>
Date: Thu, 26 Oct 1995 11:22:21 -0700
Message-Id: <9510261822.AA15031@webcrawler.com>
To: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com

We noticed clients are sending out requests with byte ranges,
on GIF's etc. Apache takes them as part of the filename, as
the CERN httpd reportedly does also, and the request fails.

The byte range draft I found says:

> In a case when byte ranges aren't meaningful the client doesn't
> request them (why would it?), and the server should tag them as
> invalid requests (in case someone wrote them by hand). 

The thing is that in this case the request makes sense, but isn't
implemented. However, returning Not_implemented might be taken as
the GET method not being implemented for any resource, which
isn't quite right either. If I ignore the byte range and return the
whole thing, I'm breaking the draft's requirement to return a 
multipart MIME message (althoug I don't see why it couldn't figure
this out for itself). I feel like returning BAD_REQUEST because
the client is making up URL's I never published, and creates a
headache, and should stop it :-) Of course if I had lots of spare
time I could try and implement this draft, but that's going a bit far.

So, what is the correct response?

-- Martijn
__________
Email: m.koster@webcrawler.com
WWW: http://info.webcrawler.com/mak/mak.html
Received on Thursday, 26 October 1995 11:26:17 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:31:34 EDT