On Tue, 2011-03-08 at 08:18 +0100, Mathias Hasselmann wrote:
> Am Montag, den 07.03.2011, 14:56 -0800 schrieb Arjan van de Ven:
> > Oh and my exchange contact database contains between 80k and 120k
> > people. that'll take a while.....
..
> * Do we really want to fully sync such large datasets? Shall we
> sync only limited datasets, e.g. title and email address?
It seems clear that the ability to search a corporate contacts database
quickly and efficiently is rather a useful one. As I understand it (and
I'm not involved with this decision, but tend to think it is a sensible,
pragmatic one), Tracker necessarily must have all the 120k corporate
contacts in its local RDF store - to then be able to use sparql /
native-sql-faster-shortcut to query them. Perhaps not in memory per-se,
but on some nearby flash / spinning media, and of course synchronised
regularly vs. the authoritative 120k contacts in the corporate database.
For all the limits of the e-d-s searching language (and certainly it is
used for querying only contacts stores), it can interact with and cache
essentially infinite (LDAP, Exchange, Groupwise etc.) remote databases.
That contrasts with tracker's (much more powerful of course), wonderful
RDF magic - which AFAIR requires all the latest data to be local.
Of course, that is only one difference; but to me important. Clearly
tracker can be duct-taped / abstracted to support this in some way -
nothing is impossible with software, with the application of infinite
resources to a given problem.
Or perhaps, as normal I'm just mistaken ...
All the best,
Michael.
--
[email protected] <><, Pseudo Engineer, itinerant idiot
_______________________________________________
MeeGo-dev mailing list
[email protected]
http://lists.meego.com/listinfo/meego-dev
http://wiki.meego.com/Mailing_list_guidelines
