On 21 May 2005, at 13:19, Graham wrote:
The appropriate answer to this is to say comparing an id from a
document retrieved from one URI to an id retrieved from a second
URI is not reliable. This makes feed:ids largely useless. The
primary use, of comparing with a previous version of a document
retrieved from the same URI, is fine.
Graham
That makes sense. But then it only gives one a very limited ability
to move an entry from one place
to another. If the entry has to be located in the same feed, then
presumably that means that it
would be difficult to move the entry from one domain to another.
Or is the following a necessary sequence for moving an entry from one
domain to another:
1. Entry e_id1 is referred to in feed f1 on domain d1 where it has
location d1:loc1
2. e_id1 is placed on domain d2 at d2:loc2
3. feed f1 now points to d2:loc2 as the new location of e_id1
4. e_id1's new representation in domain d2 somehow points to its
new trusted feed f2
also located on d2
It is complicated, but I suppose it could work.
Henry Story
