* James Holderness <[EMAIL PROTECTED]> [2005-12-11 12:30]:
> Those documents are technically related. But when you think
> about the practical implications, how the client needs to deal
> with it, how it affects the end user, I don't think it's a good
> idea.

Well, the idea for this, if I remember right, was actually mine;
and was based mainly on the thought you’re calling “technically.”
The idea was to give threading-unaware clients a chance to do
something useful.

Based on your argument I would agree that it would be best if the
related link pointed to the permalink of the entry being replied
to, rather than the feed. In fact that is a very good idea in any
case.

But in that case you run into the problem that there is no clear
way to associate the atom:[EMAIL PROTECTED]'related'] to a particular
permalink with its thr:in-reply-to pointer, ie. the information
is imprecise and  threading-aware clients have to guess.

A complication here is that we found in previous discussion that
nesting makes things more complex; ideally, therefore, any
approach would only add direct child elements to the atom:entry
element.

Given this constraint, do you have a better idea how do address
the following concerns?

• Threading-unaware clients should get at least some information
  that allows the user to notice that there’s more to the entry,
  even if the user agent remains blissfully unaware of the thread
  as such.
  
• The available information precise enough that threading-aware
  can correlate all the related bits without guessing.

The existing draft avoids problems with #2 by simply duplicating
URIs, so it’s always obvious which related links and which
threading elements belong together. If the URIs diverge, you need
some other mechanism of crossreferencing from one to the other.

> Do you know of any client implementations that do that sort of
> thing?

Honestly, I haven’t seen any yet that do anything interesting at
all with extra links (be they related, via, or something else)…
(Which rather irks me.)

Regards,
-- 
Aristotle Pagaltzis // <http://plasmasturm.org/>

Reply via email to