On Saturday 19 Nov 2005 17:21, Andrew de Quincey wrote:
> On Saturday 19 Nov 2005 13:49, w-thiel@xxxxxxx wrote:
> > On Sat, Nov 19, 2005 at 12:13:59AM +0000, Andrew de Quincey wrote:
> > > OK, restarting the discussion on the scan API. The last proposal was
> > > from Holger, and I like it for the most part. Here is a combination of
> > > that proposal with my thoughts:
> > >
> > > struct dvb_fe_scan_spec {
> > > unsigned freq_start;
> > > unsigned freq_end;
> > > unsigned srate_start;
> > > unsigned srate_end;
> > > };
> >
> > For me - being a feedhunter when I have the time - I would like the
> > symbolrates in a different way: not only srate_start and srate_end, but
> > also the possibility to provide a list of symbolrates (or even a list of
> > symbolrates with low/high for each). There are 'typical' symbolrates
> > used on a specific sat.
> > Of course, this can be done in a userland program as well with your
> > specification.
> >
> > That could be something like:
> > struct dvb_fe_scan_spec {
> > unsigned freq_start;
> > unsigned freq_end;
> > int fe_scan_n_ranges;
> > struct dvb_scan_sr_range *scan_range;
> > };
> > with:
> > struct dvb_scan_sr_range {
> > unsigned int srate_start;
> > unsigned srate_end;
> > };
> >
> > A dvb_scan_sr_range could be for example 5630-5632, a list could be:
> > { 5630, 5632 }
> > { 6320, 6322 }
> > {13260,13260 }
>
> Interesting - I was actually going to ask if anyone had any ideas on the
> best way to scan symbol rates, because I'm not 100% sure how to go about
> it.
Yeah, I was trying to think of another way - e.g. some kind of bitmask (e.g.
0x02 == 2000-3000), but that probably wouldn't be optimal if you know the
_exact_ scan rates as you do in your example... so a list of ranges is the
most compatable with both hardware and application use... Your suggestion
looks good to me.
