On Wed, Jul 20, 2016 at 9:43 AM, Neil Jenkins <[email protected]> wrote:
 > The reason we have position+limit (and the total is always returned) is so
> that a client can simulate having the whole mailbox loaded, without actually
> having to load the whole mailbox.

Thanks for the explanation. Make sense to me.

>  contacts will often be loaded in their
> entirety so the client can find the index, [...] the need to expand [calendar 
> event]
> recurrences client-side anyway makes the index in the list less useful in
> general).

This is why I fail to see the benefit of offset-based paging for
contacts and calendar events, Of course, being consistent across
object type search might be a good enough argument not to use token
style paging. Yet, from your experience having implemented a JMAP
client for all object types is there really a use-case for
server-side, offset-based paging for contacts or calendar events
(that's a genuine question, no subtlety here)? Token-based pagination
looks like a more lightweight scheme for both clients and servers, and
doesn't seem to dictate as many implementation details, but I might be
mistaken here.

Any way, I'd be happy to see it in the spec but am not pushing for it.
There's always the option to fall back to a x-nextPageToken property
for tightly bound client and server implementations that might benefit
from it.

Cheers,
Robert

-- 
You received this message because you are subscribed to the Google Groups 
"JMAP" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jmap-discuss/CABvA2aiSeSPN-CgivttYbYfGHuGMX7FaHbhx3gC_fJUV%2BFGVRw%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to