Re: Frames and Documents
On Thu, 29 Aug 1996, Stephanos Piperoglou wrote:
> On Wed, 28 Aug 1996, Steve Knoblock wrote:
> > It might solve the problem of bookmarking, but I hope you are not proposing
> > that frames cannot load independent files. Of course you could use the
> > single page for the initial load (introductory material and cover graphics).
> > This would be helpful---presently frames require you to create this kind of
> > material if you do not want blank frames on initial load---in gathering the
> > introductory material into the same place. On the other hand, perhaps as url
> > could be specified for the initial files with url(...).
> The main problem is that frames with the current Netscape spec are created
> with a frameset document and then contain independant documents. This is
> wrong. Technically you should view only one document at a time, and the
> various frames are part of that document.
No - this creates other problems. Certain applications *require* that the
content of frames be updatable independantly. Anything that attempted to
*force* all the content into a single document would make these
applications useless. See <URL:http://www.nihongo.org/english/chat/> for
such an application. There are two different things here - *presentation
partitioning* of the screen (which should be accomplishable for even a
single document with frames right now if you think about it - just use
named anchors to force different pieces of the same document to the
different frames and don't let them scroll. Hmmm...this would seem to be a
perfect application for multiple <HTML> sections in a single document,
actually...) and *content isolation* where content in different frames can
change without updating the content of the other frames.
I would _almost_ go so far as to say the only legitimate uses for frames
in fact are where content isolation is needed. Tables + CSS1 can
accomplish nearly all the other things frames are mis-used for in
presentation partitioning. Even scroll bars are defined for tables now.
All that is missing is the ability for a user to manually 're-size' the
individual cells if they don't like the output display. And that is
actually a presentation issue which could be fixed in the browsers without
changing the tables spec as I read it, since doesn't say anything about it