Re: [Evolution-hackers] ESourceExtension and backends... "claimed" generic property ?

[Tristan Van Berkom <tristanvb openismus com>]

On 02/05/2013 11:55 PM, Matthew Barnes wrote:
> On Tue, 2013-02-05 at 23:00 +0900, Tristan Van Berkom wrote:
> > I know, we've discussed this already a few months ago, it just
> > seems like something essential to the extension API, you
> > tell the backend to do "foo" and you just don't know if that
> > backend knows about "foo" (yet) or not.
>
> Seems like that could be mitigated by the backend capabilities property,
> which -- at least for static capabilities -- should be reachable through
> the factory APIs and not require an actual backend instance.

Ah, good point.

We should consider adding capability identifiers for recent additions
such a ESourceRevisionGuards & ESourceBackendSummarySetup.

Thanks for pointing that out as it's been in the back of my mind,
lingering as an unfinished detail ;-)

Cheers,
    -Tristan

> The static capabilities of a given backend type are the same for all
> instances of that type, if I'm not mistaken.  So it would make sense to
> be able to query the factory for available backend types and their
> respective static capabilities prior to instantiating one.
>
> ESource and its extensions are meant to be just a dumb data container.
> It's used for more than just backend configuration.
>
> Mail signatures nowadays are ESources with a [Mail Signature] extension,
> for example, and I'm also considering porting our filter rules and saved
> searches to use ESource in a similar manner.
>
> Matthew Barnes