[mythtv] [mythtv-commits] mythtv commit: r10651 by danielk

[Daniel Kristjansson]

On Wed, 2006-07-26 at 20:05 +0200, Janne Grunau wrote:
> > Fixes #2101. Allow EIT data to be collected when the section reader
> > is in use by the DVB signal monitor.
> My patch missed a break for case 0x1ffb. It has no effect though since 
> the default sets the same parameters.
Ack, I should have caught that. I'll fix it soon.

> > I'm not going to apply the other patch to use the section reader
> > exclusively. The section reader code can not fully support both ATSC
> > and DVB without being made significantly more complicated,
> It can, just set filter and mask for all PIDs to 0x00. This is a kind of 
> silly except that it makes the transport stream table monitor obsolete.
> Updated filters and mask attached.
Right. You could just make the mask and filters 0x0 for every pid, but
I was afraid doing this could break the very USB DVB devices we're
trying help. I figured this was safe for the EIT pids, since we've
already tuned by the time this gets enabled, and it is safe for 0x1ffb
since none of these devices are ATSC devices. I'm actually a little
concerned about the default, since this would be used for the PMT,
but if someone complaints we can change it back to filter and mask
of 0x02 and 0xff, respectively.

> > and does not support "tuned" DVB channel scanning.
> I don't see that. Can you please explain?
Just because it doesn't pick up the "other" tables. Tuned scanning is
uses these to find new transports to scan (or should be).

> > It also does not work with 
> > some of the DVB-S and DVB-C broadcasters which improperly label their
> > NIT and SDT tables as "other NIT" and "other SDT" tables.
> I forget these too often, fixed. But I suspect that DVBStreamData can't 
> handle only NIT_ or SDT_other.
Which is why I didn't just add them. I think we need to wait for 0.21,
in case this breaks something.

> PMTs on these PIDs also fixed but I'm afraid other ATSC can also be on 
> these low PIDs. If this is the case I should stop to change the filter 
> parameters.
In theory ATSC could put other ATSC tables on the NIT and SDT pids, but
in the wild all I've seen is a PMT on pid 0x10.

> > The section reader as written is a functionally limited fallback for
> > problematic USB devices, extending it to be fully functional would
> > need to be a 0.21 project.
> Removing the transport stream table monitor should not happen before 
> 0.20 release. But further improvements to the section reader should be 
> possible if they don't touch anythong other.
As long as they section monitor doesn't give the DVBStreamData data
which the TS monitor doesn't give it, the changes should be safe. I'm
out of town all of next week so I won't be around to apply emergency
fix patches in that time...

I'm all for using the section monitor in the DVBSignalMonitor, it's just
too late for 0.20. We may need special broken DVB device support for the
USB devices, but the only way to know for sure is to send a series of
patches to someone using one of these devices who is willing to test
the patches...

-- Daniel