- From: Klaus Weide <kweide@tezcat.com>
- Date: Fri, 5 Sep 1997 10:12:17 -0500 (CDT)
- To: Larry Masinter <masinter@parc.xerox.com>
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
On Thu, 4 Sep 1997, Larry Masinter wrote: > Removing the restriction on the sender unfortunately adds more > complexity to the recipient, if the recipient must be able to > process mandatory header-fields in the trailer. And implementors > of recipients and senders seem to be in different groups, even > if they work for the same company, so we can't rely on implementor > ingenuity to get this one 'right'. > I was thinking of something like: > "Any header-field which the sender MUST send in a request should be > in the header and not in the trailer." > > although that is not quite right. > > (I will note, parenthetically, that the ability to say the previous > sentence is critically dependent on separating "header-field" from > "header".) That is far too cryptic, and I am not sure what it means. The BNF currently doesn't define "header-field". "header field" means the same as what is more sloppily referred to as just "header", in my understanding. So maybe you mean "header field value" instead of "header-field"? Klaus
Received on Friday, 5 September 1997 08:15:32 UTC