Antone Roundy wrote:
On Wednesday, April 27, 2005, at 04:21 AM, Brett Lindsley wrote:
One reason for title only feeds is to address bandwidth limited devices. The server I set up provides the same feed in two different formats - one title only and the other with title/summary/etc. The end client can decide which feed it wants to
work with based on its capabilities;
This raises a few questions:
1) How does one indicate the existence of variants?
[EMAIL PROTECTED]"alternate" or using a custom (read: extension) relations or using extension elements. See also my comments below on 3).
2) Can/should the variant feeds use the same atom:ids for variants of the same entries?
Yes, IMO, as they identify the same entry, with different, more or less rich, representations.
However, does the current draft explicitly say that or can it be understood that different representations of the same entry must/should/may have different ids?
This reminds me of [EMAIL PROTECTED]"alternate" vs. [EMAIL PROTECTED]"alternate-of" discussion...
3) How does a client select a preferred variant?
Good question, but it seems to me that it's beyond the scope of the current (format) spec. It should IMO go with autodiscovery.
4) How does an aggregator discover the "complete" and/or "authoritative" variant of an entry?
Autodiscovery?
IMO, all this should be recorded and discussed later when the autodiscovery spec will be on air.
-- Thomas Broyer
