- From: Roy T. Fielding <fielding@apache.org>
- Date: Wed, 28 Aug 2002 17:42:55 -0700
- To: Norman Walsh <Norman.Walsh@Sun.COM>
- Cc: www-tag@w3.org
On Wednesday, August 28, 2002, at 02:38 PM, Norman Walsh wrote: > / Tim Bray <tbray@textuality.com> was heard to say: > | denumerable. If you could arrange for a finite-length way to encode > | irrational numbers (aside from special cases such as e and pi) you'd > | be right, but I'm pretty sure that in principle you can't. Because if > | you could then they'd be denumerable just like URIs. -Tim > > I don't think the proof hinges on irrational numbers. The > diagonalization proof that real numbers are not denumerable relies on > the fact that you can always manufacture a new real number that is > demonstrably not in your list. Whether or not that number is > irrational is incidental. I think. :-) Right. Any given URI is finite. However, there is no finite limit on the length of a URI. Therefore, given any list of URI one could construct a URI not on that list ... In short, that footnote would be wrong even if it were stated correctly. We should just delete it and all memory of the concept. ....Roy
Received on Wednesday, 28 August 2002 20:43:07 UTC