Re: Proposal: Checksum header or method

According to Mordechai T. Abzug:
> On Wed, 26 Jul 1995 Harald.T.Alvestrand@uninett.no wrote:
> 
> > The content-MD5 document, RFC 1544, describes exactly such a header.
> > Is this what you are looking for?
> 
> Yes.  However, it's optional, and in fact, I've never seen it included
> with any HTTP-generated header.  It would be nice if such a mechanism were
> formally included as part of the reply to a HEAD request, not for purposes
> of error-chjecking, but merely to know if the document has changed. 
> 

Are there servers that produce an incorrect last modified date for
files?  I don't know of any, but if they exist then they are broken
and the proper solution is to have them fixed.  Of course the last
modified date for a CGI script may be problematical, but I don't see
how checksums could help this.

John Franks

Received on Thursday, 27 July 1995 07:37:21 UTC