Re: Plans to split Tracker 1.12 (Was: Re: [Tracker] The Big Rip)

[Sam Thursfield <ssssam gmail com>]

On Fri, Nov 4, 2016 at 1:58 PM, Bastien Nocera <bnocera redhat com> wrote:
> Just a quick reply to say that where the code lives, in this case, is
> probably of less importance than how applications can use it.
>
> We need a way to use Tracker within sandboxes, which means getting rid of the
> miners running in the session, outside the sandbox, and start having application
> controlled mining.

My thoughts here are that if the mining moves inside the sandbox, the
store should do as well, and if the data is needed outside the sandbox
it should be exported in some form (gnome-shell search providers are
one example of exporting the data, an app could also offer a raw dump
of the data as Turtle or JSON if folk want to do clever SPARQL queries
across it). This should be possible now, if the sandbox opens a new
session bus for its Tracker processes to use. Doing this is hacky at
present due to assumptions in the store and miner, but I think all the
bits needed are there.

Sam