[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Aristotle Pagaltzis wrote:
> * James Holderness <j4_james@xxxxxxxxxxx> [2007-12-07 03:55]:
> > Steven Lees wrote:
> >> On the initial sync of a feed, most sync-aware clients will
> >> still want to ask for the feed without tombstones, so that
> >> they can just get the non-deleted items.
> > I wouldn't think that's worth the effort. Tombstones should be
> > small enough and infrequent enough that I wouldn't care about
> > getting them on an initial sync.
> It’s pointless either way. You might avoid receiving a bunch of
> irrelevant tombstones on the first fetch – only to end up getting
> them all on your second fetch.
> You could invest somewhat significant extra complexity to make it
> actually work… all so you can avoid a handful of tombstones for a
> short period after subscribing to a feed. Ahem.
Maybe I've had too much of the ETag cool aid, but I would think that with a reasonably straightforward ETag implementation on the feed, then I would never need to see the tombstones in place before my first sync, regardless of how many times I resync.
I certainly agree that creates a little more work on both the client and server. Whether or not you think it's worth it depends on your scenario. For blog publishing, it's probably not that interesting. For two-way sync of a larger data set, it's a lot more useful.