[ckan-discuss] Odd error message from API

[William Waites]

* [2011-04-01 19:10:56 +0200] David Read <david.read at okfn.org> écrit:

] So for that use case, a cache/aggregator on CKANs is one way. Another
] is federated log-in. Another might be dcat (I don't know much about
] it). Another is package syncing. I'm merely suggesting we schedule a
] meeting to discuss our needs and compare the solutions before
] investing any more in any of them.

Sure, we should have a design session, and collect together the
various alternatives and look at them.

DCat doesn't help here, it is just a data model that sort of
maps closely enough to CKAN's data model and is emerging as
a standard metadata interchange format - it is more of an
alternative to JSON than anything else, it doesn't have anything
to say about aggregation or syndication or whatever.

Federated log-in will be very hard once you start thinking of
permissions. It's not only federated authentication which is
easy but federated authorisation which is hard - how should 
some municipality know if someone from somewhere else should
be allowed to change a record? This is a rabbit hole.

Package syncing, I'm not sure exactly what this entails. I
suspect it might be a special case of the aggrecache or maybe
the other way around.

] I'm still missing the reason that you're worried about detecting edit
] conflicts, because I'm surprised (yet delighted) if a package on
] ckan.net is updated more than once in 24hrs. There may be other
] reasons to consider completely changing our API, but this one doesn't
] seem very pressing (yet!).

When we have harvesting, it depends on the harvest window. So
for example if you edit a package immediately after a harvest
job and then someone edits within the harvest window (which 
might reasonably be set to some large value) from upstream,
you have a conflict even though the edits were actually quite
far apart in time. It is pretty easy to handle this situation
but you have to be able to detect it.

Cheers,
-w
-- 
William Waites                <mailto:ww at styx.org>
http://river.styx.org/ww/        <sip:ww at styx.org>
F4B3 39BF E775 CF42 0BAB  3DF0 BE40 A6DF B06F FD45