- From: Xiaoshu Wang <wangxiao@musc.edu>
- Date: Tue, 08 Jan 2008 11:54:37 +0000
- To: John Cowan <cowan@ccil.org>
- CC: John Black <JohnBlack@kashori.com>, Pat Hayes <phayes@ihmc.us>, noah_mendelsohn@us.ibm.com, Mikael Nilsson <mikael@nilsson.name>, "Sean B. Palmer" <sean@miscoranda.com>, www-tag@w3.org
John Cowan wrote: > John Black scripsit: > > >> Isn't this assumption false with respect to the web? In general, I have no >> idea what states the resources I visit may be in. Any given resource may be >> a novel, a song, a picture, an email message, a movie, a blog post, etc., >> etc. How could I possibly know the N possible states that 8 billion >> resources can be in? >> > > We can if we set some reasonable upper limit to the size of a > representation. For example, if we say that 1 TB is such an upper limit, > then the resource is in one of about 2^(10^15) states, given that there > are about 10^2 media types and there are ~ 10 bits/byte. > Still, this is the wrong assumption. JB is right that Shannon's theory is only meaningful when there is a context. The same bit can mean different things in different context. The implication of your suggestion is that we can assign every possible states a bit pattern. But if cap the size, the implication is that web is closed. If you don't, then the problem comes back to you. Xiaoshu
Received on Tuesday, 8 January 2008 12:00:27 UTC