Re: [Tracker] track only postscript and pdfs?

[Jamie McCracken <jamiemcc blueyonder co uk>]

On Wed, 2008-04-16 at 23:44 +0200, Mikkel Kamstrup Erlandsen wrote:
> On 16/04/2008, Michael Biebl <mbiebl gmail com> wrote:
>         2008/4/15, Mikkel Kamstrup Erlandsen
>         <mikkel kamstrup gmail com>:
>         
>         > On 15/04/2008, Michael Biebl <mbiebl gmail com> wrote:
>         > > 008/4/11, Parker Jones <zoubidoo hotmail com>:
>         > >
>         > > >
>         > > >  Hello,
>         > > >
>         > > >  I'd like to use tracker to monitor only pdf's and
>         postscript files.  Is
>         > this possible?
>         > > >
>         > > >  At the moment, I ignore a number of extensions *.doc
>         *.html *.png  etc,
>         >  but the ignore list is getting too long and still too many
>         files are
>         > getting tracked.
>         > >
>         > >
>         > > I think this should be  doable:
>         > > We should not only have a blacklist of file extenstions,
>         but also a
>         > whitelist.
>         > > Then add * in the blacklist and *.pdf and *.ps to the
>         whitelist. The
>         > > whitelist should have a higher priority then the
>         blacklist. I'd say
>         > > (without looking at the code), that this could be fixed
>         easily.
>         >
>         >  Or go old-school and just make a list of regexps..? Then
>         one could simply
>         > write a not-pdf-or-ps blacklist regexp.
>         >
>         
>         
>         Would be an option. But do you think this would be intuitive
>         for the
>         average user?
>
> Was that a rhetorical question? Heck, no! I bet that less than 0.1% of
> the Tracker users know regexps :-)
>
> Anyways, Beagle does in fact seem to support it. Or maybe it is just
> globbing.

Im not sure we need white listing as such

what would be sensible is to enable/disable indexing by service category
(Ie turn on or off music/video/Documents/text/dev file indexing)

hopefully that coarse grain approach will suffice and still be easy for
users

jamie