On Fri, 23 Jul 2004, Timwi wrote:
Michaƫl P. wrote:
Has someone a suggestion about how-to produce a good Id element?
If it only needs to be unique and is not subject to any other constraints, wouldn't the curid resp. oldid plus the date/time of the revision be enough?
This should help:
http://diveintomark.org/archives/2004/05/28/howto-atom-id
--jeremy