- From: M.T. Carrasco Benitez <carrasco@innet.lu>
- Date: Fri, 28 Feb 1997 09:15:13 +0100 (MET)
- To: Drazen Kacar <Drazen.Kacar@public.srce.hr>
- cc: Larry Masinter <masinter@parc.xerox.com>, misha.wolf@reuters.com, www-international@w3.org, unicode@unicode.org
> As far as I understand the situation, it would be nice to have that > information extracted and put in the header for the HEAD request. > I don't know which type of client could benefit from this, but perhaps > somebody else does. I was thinking about robots: A robot that only look for German docs, for example. There are probably other applications. > > The "charset" information is NECESSARY to interpret the > > body (of text/* types) and thus MUST appear in the header. > > Yes, but... > There are enough web admins who don't want to know anything about their > servers. That means the users can't put charset info in the header. > META tag is the only thing that remains, at least for text/html. > Disclaimer: just describing current practice... The charset must be in the HTTP header and inside the doc in <META HTTP-EQUIV=Content-Type ...>. This META should be as near to the beginning of the doc as possible, as this is a catch-22 situation: how do the program reads the charset if it does not know the charset in the first place ? The doc management system could pre-process the docs, so the server does not read many information at the precise moment of serving the docs. But this is a doc management system factor and one could consider that the server is "virtually" reading the information (language label, charset, etc). Tomas
Received on Friday, 28 February 1997 03:13:49 UTC