On 06/12/11 10:11, John Pilkington wrote:
On 12/06/2011 08:30 AM, John Pilkington wrote:
On 06/12/11 00:04, Duncan Brown wrote:
Hi Guys
I've been running the media_build drivers on Centos 6 since it was
released. Using the pcTV nanostick in the UK with amazing results
The drivers have generally built fine for me, apart from maybe 2
occasions. Both times it worked again a day or two later (it's updated
almost continuously, and sometimes silly bugs slip in)
As the Centos kernel updates are so few and far between, its pretty much
been run and forget
As for the error above, are you sure you've got kernel-devel installed
John?
Dunc
That's a good point. It's a newish install. Probably not. Thanks. I'll
check.
That was, indeed, the problem. 480 modules built and installed. It took
about a quarter of an hour. Editing to select only relevant devices
might have made it quicker.
I see the new kmdls have landed too. Thanks very much, Axel.
The script looks as if it provides the firmware as well as the modules,
but the kmdls are simpler - for the user!
Now I have to find some unemployed receiver hardware :-)
I hope no-one will mind if I continue in this thread, although it's
getting slightly off-topic for ATrpms. I'm using the modules created by
the media_build script mentioned earlier in this thread.
I had a KWorld_UB499-2T usb device that I hadn't been able to make work
before; when I had tried it it was always 'busy' and AFAIR it interfered
with proper Myth operation.
Now it's currently recording two programmes on one mux and one an
another in MythTV 0.24 fixes under SL6, although there are some strange
features that need investigation and dmesg reports a significant
limitation - viz:
dev-usb: will use the device's hardware PID filter (table count: 31).
This causes backend error messages about once a second when recording
from some but maybe not all muxes. I had seen similar messages when
using a Freecom usb device that had IIRC had a PID table count limit of 15.
I had to get firmware for it from the KWorld Windows driver, using the
procedure set out in its linuxtv.org wiki page. /lib/firmware already
had lots of files in it from somewhere but that wasn't one. On
rebooting dmesg showed it had loaded ok and I tried running mythtvsetup.
That came up with the default theme, which I hadn't expected, and it
seemed to have no recollection of being run earlier: it's because the
hostname had been set by my ISP. I had had that before but forget
exactly how to resolve it.
I rebooted with no internet connection and defined two cards with four
virtual tunes each, which I expected to see as tuners 2,3,4,5 and
6,7,8,9, tuner 1 being a file-based DEMO.
Then I did a channel scan, starting with just the known frequency of the
BBC SD mux; all services; link to other muxes. This time that gave five
transports (as expected), the starting one still showing as Auto; so I
deleted that, and all dvb channels, and rescanned in the same way. 108
services found, although their numbering isn't as expected: some
channels, I think those in the Film4 mux, have numbers above 25000.
To get the recordings enabled it seems that I have to disable 'start on
demand' and 'active EIT scan', and system status now believes I have 22
tuners, many of which are unavailable. Maybe that's because I checked
on the settings of the two adapters after setting them up. I did delete
all tuners before I 'started over'; it seems to work, anyway, although
I haven't yet examined the recordings with Project-X.
Apologies if all this seems too MythTV oriented. I wouldn't have it
without ATrpms!
John P
_______________________________________________
atrpms-users mailing list
atrpms-users@atrpms.net
http://lists.atrpms.net/mailman/listinfo/atrpms-users
