> From: Jukka.Korpela@hut.fi [SMTP:Jukka.Korpela@hut.fi] > > In fact, if frames had been introduced to HTML in a sensible way, > then it would have been done so that selection between a normal (noframes) > version and a frameset version of a page would be based on content > negotiation instead of <noframes> things. > [DJW:] You've still got the problem of local copies of the page, but more importantly, it is almost impossible to get people to use <noframes> properly, but even fewer content authors are aware about content negotiation, and many are using cheap web space where content negotiation is not available. The result of a distinction in the header would be 95% of frames pages being mislabelled in the HTTP headers! Note that a valid frames page can be identified based on the very first line, but again, it is rare for frames authors to put in a valid DOCTYPE (those with valid DOCTYPES probably avoid frames!). Isn't HTTP off topic?Received on Tuesday, 22 August 2000 06:19:52 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 6 April 2009 12:59:10 GMT