- From: Danyel Ceccaldi <dceccald@elaine.crcg.edu>
- Date: Thu, 21 Sep 1995 17:53:55 -0400
- To: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
- Cc: luotonen@netscape.com
> From http-wg-request@cuckoo.hpl.hp.com Thu Sep 21 15:37:05 1995 > From: Ari Luotonen <luotonen@netscape.com> > Subject: Re: domain-name? > To: fielding@beach.w3.org (Roy Fielding) > Date: Thu, 21 Sep 1995 12:33:41 -0700 (PDT) > Cc: montulli@mozilla.com, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com > > > So we're going to waste terabytes of bandwidth over the coming months > (and years) before we even have URNs that would need the Orig-URI > header instead. Just the Netscape home server alone would receive 10 > GB of useless data that it doesn't need nor want in one year. Does it matter ? Why ? Are there no other wasting sources more important to remove ? How did you got that numbers ? What about using a base element as recommended in RFC1808 ? base-header = "Base" ":" "<URL:" absoluteURL ">" If you have the possibility of providing some sort of a base-header, you could at least specify the host, if the base is not already given by the document. I was wondering, I thought the base-header is inside the draft, but it isn't. Example: If a www-client wants to get the resource pointed to by http://www.foo.bar/any he could send: Request to the IP-number got by resolving the domain-name 'www.foo.bar': GET /any HTTP/1.0 Base: <URL:http://www.foo.bar> ... (other headers) Was it the intention of RFC that the base-header is used in the Http-protocol ? By Danny
Received on Thursday, 21 September 1995 13:56:49 UTC