I've got a DCT-6200 with yet another vendor ID. Here's a trivial patch to add
support for it.
--
Jarod Wilson
[EMAIL PROTECTED]
--- 6200ch.c.orig 2005-02-24 23:35:06.339455213 -0800
+++ 6200ch.c 2005-02-24 23:33:54.524717285 -0800
@@ -31,6 +31,7 @@
#define DCT6200_VENDOR_ID1 0x0ce5
#define
mythfilldatabase is crashing when its calling the ~MDBManager destructor.
here is a backtrace.
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 16384 (LWP 23953)]
0x0f574ab0 in ~MDBManager (this=0x10061e50) at mythdbcon.cpp:111
111 delete m_sem;
(gdb) thread app
On Fri, Feb 25, 2005 at 02:05:10PM +1000, David Whyte wrote:
> I think the major issue is having two copies of the queries. All
> clients will need to be updated if a performance enhancement or table
> change were made.
>
> IMO it is much easier to ask for the data, and know just the format of
>
Here you go
$ cat /proc/cpuinfo | grep "model name" | head -n 1
model name : VIA Samuel 2
right now I just use -march = i586 but i suppose there are some other
flags that would work better?
note, pentiumpro or i686 causes illegal instruction segfaults even
though this is technically an i686
I think the major issue is having two copies of the queries. All
clients will need to be updated if a performance enhancement or table
change were made.
IMO it is much easier to ask for the data, and know just the format of
the data that is being returned.
Whytey
On Thu, 24 Feb 2005 19:58:41 -
On Thu, Feb 24, 2005 at 09:01:36PM -0500, Anduin Withers wrote:
> > However, is there any interest out there in writing a new protocol that is
> > able to give a remote frontend absolutely everything it needs to deal with
> > myth?
>
> I know it is heresy but I'd like the same thing, if people wa
On Thu, Feb 24, 2005 at 03:57:47PM -0500, Daniel Thor Kristjansson wrote:
> X-Spam-Level:
>
> ]On Thursday 24 February 2005 10:00, Daniel Thor Kristjansson wrote:
> ]> please send the output of `uname -p` for other CPUs so I can
> ]> add these to the configure script.
> ]Interesting... my FC1 &
> However, is there any interest out there in writing a new protocol that is
> able to give a remote frontend absolutely everything it needs to deal with
> myth?
I know it is heresy but I'd like the same thing, if people want/need direct
access to the DB let them do it, it doesn't preclude a good
> run the query: 3019 rows in set (15.95 sec)
>
> [
> CREATE INDEX title ON recorded (title);
> ]
>
> run the query: 3019 rows in set (4.78 sec)
I just committed code to CVS to add an index on the recorded.title
column. It took my times for that query down from about 3.1 down
to 1.8 seconds so
On Thursday 24 February 2005 07:24 pm, Paul Volkaerts wrote:
> A common set of clases to handle drawing on top of TV or on QT is
> definately a good idea.
>
> But how to handle multiple plugins all wanting to access the display is the
> problem as I see it; and I can think of two options:
>
> (2)Or
> Heh, cool - the UI part of this is what mythui is partially supposed to
> accomplish. I had intended to convert over the OSD code so it
> uses the same
> base classes/etc for drawing as everything else will. This way,
> anything can
> draw on the OSD (or anywhere else) without being written
> s
So I've seen a lot of talk in the past week about the myth protocol. Most
of it seems to either be centered around encapsilating it within some other
interface (ie, SOAP, java, etc), or in writing a new protocol to do what
the current one does.
However, is there any interest out there in writing
What would be the ultimate goal of this java project ? just curious
On Fri, 25 Feb 2005 07:02:37 +1000, David Whyte <[EMAIL PROTECTED]> wrote:
> Hey Yan,
>
> Yeah, I actually have this working now. The major problem *I* found
> was the use of the DataOutputStream instead of a PrintStream. Thi
> > 89%/ 46fps[mpeg2video @ 0xb78b8854]invalid mb type in B Frame at 66 22
> > 90%/ 46fps
> > 1
> >
> > Finished commercial break flagging at Thu Feb 24 02:00:08 2005
> Flagging just "plays" the recording using the normal playback code,
> so if mythcommflag gets an error like this, then odds are
Hi
24 hours after trying to subscribe to myth-users there is no ack.
Maybe someone here can help:
In Perth I have no grabber. I have no xml listings.
With 0.16 I set PIDs, serviceIDs and freq by hand - all works fine
With 0.17, CVS
SCAN works and finds channels.
SCAN again says 'already in data
Just to clarify, there is NO instention to create a *new* protocol.
The idea is to describe the current protocol so others can interface
with it.
Yes, I will be adding my Java layer to it for the things I wish to do,
but if done right, I think this should open up the possibilities to
being able t
On Thursday 24 February 2005 05:20 pm, Paul Volkaerts wrote:
> I'm looking for comments as to whether this is the right way to do this,
> and whether it is aligned with, or clashes with, things like MythUI. I
> don't want to check this in and find it breaks things people are working
> on; but I do
This patch removes stops the signal monitor thread from always running.
It now only runs for scanwizard. There are a few other fixes in there as
well, but hopefully this will cure Ed's segvs.
--
John Pullan <[EMAIL PROTECTED]>
Index: dvbchannel.cpp
===
> I have started to nut out the MythProtocol. There is a page on the
> Wiki at http://mythtv.info . If you guys want to add anything, go
> ahead.
The web page seems more directed at creating a Java interface layer than an
actual new mythtv protocol. (I'm interested in a Java layer as well, but I
As I mentioned a while back; I have prototyped a few changes to the OSD that
improves the integration of plugins. The code isn't that extensive but the
architectural implications are - it means in order to use this facility,
plugins need to be split into GUI and worker threads (plugins can contin
I have started to nut out the MythProtocol. There is a page on the
Wiki at http://mythtv.info . If you guys want to add anything, go
ahead.
Whytey
On Fri, 25 Feb 2005 09:00:11 +1300, Derek Watson <[EMAIL PROTECTED]> wrote:
> Surely building mythsoap as an optional integration abstraction layer
On Thursday 24 February 2005 16:09, Daniel Thor Kristjansson wrote:
> I'm assuming
> all Athlons have mmx support, is this incorrect?
I believe that is correct.
-JAC
___
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://mythtv.org/cgi-bin/mailman/lis
On Thu, 24 Feb 2005, Simon Kenyon wrote:
]Intel(R) Pentium(R) M processor 2.00GHz
]CFLAGS=-O2 -mcpu=pentium4 -O3 -pipe -fomit-frame-pointer
]AMD Athlon(tm) processor
]CFLAGS="-O3 -march=athlon-tbird -pipe -fomit-frame-pointer"
]VIA C3 Ezra
]CFLAGS="-Os -march=c3 -pipe -fomit-frame-pointer"
I've
Hey Yan,
Yeah, I actually have this working now. The major problem *I* found
was the use of the DataOutputStream instead of a PrintStream. This
cause problems because you would write to the Socket as bytes.
I just need to clean up the code, and then I will post it for others
to play with if the
]On Thursday 24 February 2005 10:00, Daniel Thor Kristjansson wrote:
]> please send the output of `uname -p` for other CPUs so I can
]> add these to the configure script.
]Interesting... my FC1 & FC3 boxes both report simply 'athlon' (not
]'athlon-xp').
"blammo.doh" had the same experience.
]W
Surely building mythsoap as an optional integration abstraction layer
has little or no impact upon the existing protocol. This is what I
thought was originally proposed by Adam, and I'm still interested in
helping make it happen.
Despite it's bulky nature, XML is superb to program against from a
p
On Thursday 24 February 2005 10:00, Daniel Thor Kristjansson wrote:
>
> please send the output of `uname -p` for other CPUs so I can
> add these to the configure script.
Interesting... my FC1 & FC3 boxes both report simply 'athlon' (not
'athlon-xp').
What's even weirder is my Debian Sid box do
On Thu, 24 Feb 2005 17:44:31 +, Simon Kenyon <[EMAIL PROTECTED]> wrote:
> On Thursday 24 February 2005 15:00, Daniel Thor Kristjansson wrote:
> > please send the output of `uname -p` for other CPUs so I can
> > add these to the configure script.
Athlon XP, Barton class,
[EMAIL PROTECTED] ~]#
> > This could have to do with the keyframe adjust map possibly. I'll try
> > to reproduce sometime.
>
> Just to be more precise: you get the shape of the correct frame but
> with the color of the frame shown when you hit the FF/REW button. I
> guess it misses the keyframe to produce the correct
> 87%/ 46fps[mpeg2video @ 0xb78b8854]ac-tex damaged at 116 61
> [mpeg2video @ 0xb78b8854]00 motion_type at 21 62
> [mpeg2video @ 0xb78b8854]00 motion_type at 1 21
> [mpeg2video @ 0xb78b8854]00 motion_type at 3 21
> [mpeg2video @ 0xb78b8854]invalid mb type in I Frame at 86 35
> 89%/ 46fps[mpeg2vid
On Thursday 24 February 2005 15:00, Daniel Thor Kristjansson wrote:
> please send the output of `uname -p` for other CPUs so I can
> add these to the configure script.
the script will have to be a lot more selective about setting the CFLAGS
there are a variety of athlons, which need different CFLA
On Wed, 23 Feb 2005, Chris Petersen wrote:
Ick, did I leave the 4-second timeout? It should be at least 20. I set
it to 4 while I was testing things, must have forgotten to set it back.
Just to confirm, this fix seems to work around the issue for me.
The backend probably still shouldn't hang bec
On Thu, Feb 24, 2005 at 11:19:32AM -0500, Dan Wilga wrote:
> Personally, I prefer it this way. I find that having it fade out
> results in too much screen clutter if you have browse mode for
> up/down arrows turned on. With fade working the way it used to, you
> could select a channel, wait for
On Thursday 24 February 2005 02:54, Bill Jackson wrote:
> I wonder - what do folks think of bitkeeper?
please no. i beg you no.
--
simon
___
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
> Anything more I can provide in the way of debugs, let me know. (no I
> have not tried the patch listed in this thread yet. Given the fact
> only one audio track is seen by both myth and mplayer, I'm not sure it
> would make any difference)
Applied the patch. Debug output is a little different, r
On Thu, 24 Feb 2005 10:09:36 -0700, Blammo <[EMAIL PROTECTED]> wrote:
> NYPD-Blue recorded last night. Play it back in mythfrontend, no audio.
> Hit +, the OSD shows "Audio Track 1" but never plays any sound.
> Flipped between AC-3 passthru/not, no diff. All other present
> recordings are fine. Her
> Hard to bump it up my own stack, since my own patch works for my
> own copy. The only thing likely to bump it up the stack is users
> regularly reporting how they are getting no sound on some recordings.
And to that extent here's my 2cents:
NYPD-Blue recorded last night. Play it back in mythfr
David Whyte wrote:
I have noticed that the proper way to close the socket is to call the
DONE command over the MythProtocol, but I was wandering what effects
this would have if it weren't done *everytime* (such as when the
frontend dies in a nasty way). Do the open connections get 'garbage
collect
At 9:44 PM -0600 2/23/05, David Engel wrote:
I recently noticed the OSD is no longer fades out on my pvr350 system.
It goes completely off after the normal display time. The OSD fades
just fine on my remote frontend with software decoding. I thought it
might have something to do with the BLT_FILL
On Thu, 24 Feb 2005, Jeremiah Morris wrote:
]On Feb 24, 2005, at 10:00 AM, Daniel Thor Kristjansson wrote:
]Let me draw your attention to my patch for doing audio detection in the ffmpeg
]configure script:
]
]http://www.gossamer-threads.com/lists/mythtv/dev/108824
Ok, I've added your audio suppor
Le 23.02.2005 21:23:03, Chris Pinkham a écrit :
> cutting commercials), though the FF shows some big blocky picture at
> the beginning (and the picture is really messed up) but it quickly
goes
This could have to do with the keyframe adjust map possibly. I'll try
to reproduce sometime.
Just to be m
David Whyte wrote:
I have attached the Java code incase anyone wanted to take a look, but
any other pointers would be great.
Cheers,
Whytey
Hi,
This might sound a bit stupid, but have you looked at the network output
using something like ethereal ? I have a sneaky suspicion that you
might be s
On Thu, 24 Feb 2005, Jeremiah Morris wrote:
]On Feb 24, 2005, at 10:00 AM, Daniel Thor Kristjansson wrote:
]
]Let me draw your attention to my patch for doing audio detection in the ffmpeg
]configure script:
]
]http://www.gossamer-threads.com/lists/mythtv/dev/108824
This almost works for me, I had
On Feb 24, 2005, at 10:00 AM, Daniel Thor Kristjansson wrote:
I've also put all the config options closer to the head of the file
with the dependent defines, libraries, etc below. The arts location
is found using the location of artsc.h; so if it isn't found on your
distribution, please send me th
[EMAIL PROTECTED] <> scribbled on Thursday, February 24,
2005 4:53 AM:
> Sitting here in my computer room, I realized that even with all the
> myth-dedicated hardware in the room, the machines with the most
> horsepower aren't even being used for myth. In the evening that's not
> a big deal, since
[EMAIL PROTECTED] <> scribbled on Thursday, February 24,
2005 7:37 AM:
> On Wednesday 23 February 2005 09:46 pm, Jim Westfall wrote:
>> I found some additional information on my test x86 backend box. It
>> appears that connectionClosed() signal is not always getting
>> triggered when a client dis
On Feb 24, 2005, at 4:24 AM, Johnny Strom wrote:
Yes this is good idea to build packages from cvs perhaps make a script
that dose it automatically every night.
I do nightly CVS snapshots for OS X, the Perl script for building Myth
(and its dependencies) is available on my site. Others are welcome
This patch uses our ffmpeg configure script to setup the compile
flags. I've added support for adding pentiumpro, pentium 4, and x86_64
specific -march flags to configure based on the output of `uname -p`
please send the output of `uname -p` for other CPUs so I can
add these to the configure scrip
On Sun, 20 Feb 2005 23:13:52 -0500, Alan Gonzalez
<[EMAIL PROTECTED]> wrote:
> I've narrowed this down to changes from 11/29-11/30.
>
> Checkout of cvs on 11/29, no pause freezes.
> Then the problem started with a checkout of cvs on 11/30.
>
> Not sure exactly which changes did it, but I'm thinki
> I have noticed that the proper way to close the socket is to call the
> DONE command over the MythProtocol, but I was wandering what effects
> this would have if it weren't done *everytime* (such as when the
> frontend dies in a nasty way). Do the open connections get 'garbage
> collected' after
I have noticed that the proper way to close the socket is to call the
DONE command over the MythProtocol, but I was wandering what effects
this would have if it weren't done *everytime* (such as when the
frontend dies in a nasty way). Do the open connections get 'garbage
collected' after a certain
>Isaac Richards wrote:
>
>>On Tuesday 28 December 2004 01:37 am, tommy wrote:
>>
>>
>>>The pointer manipulation in the two lines of code are incorrect for
>>>k8 processors, they were causing the frontend to segfault on
>>>playback.
>>>This hasn't been tested on a 32 bit processor, so if someone c
On Wednesday 23 February 2005 20:41, tommy wrote:
> > Since 2 weeks (I think) I'm having this problem where the frontend
> > gives a segmentation fault when I select Watch TV and also when
> > I select a previously recorded file. Watching video's with mythvideo
> > (mplayer) works fine and also
> Yes this is good idea to build packages from cvs perhaps make a script
> that dose it automatically every night.
Yes, we're working on such a script for debian unstable here based on
mdz's packages. They should be ready soon.
___
mythtv-dev mailing lis
Adam Egger wrote:
On Wed, 23 Feb 2005 17:37:35 -0800, Nav Jagpal <[EMAIL PROTECTED]> wrote:
For example, Debian has three "branches" (Don't think that's the right
word for it).. Stable/testing/unstable. That may be overkill for this
project, but the idea is there.
So the easiest solution for the
Was running flagging on a PBS-HD special (no point, there are no
commercials, but mythcommflag isn't aware of "no commerical" flags at
runtime. It only appears to decide whether it gets jobqueued or not)
and mythcommflag dies, consistantly, at 90%, every time, with the
following error.
87%/ 46fp
On Wed, 23 Feb 2005 17:37:35 -0800, Nav Jagpal <[EMAIL PROTECTED]> wrote:
> For example, Debian has three "branches" (Don't think that's the right
> word for it).. Stable/testing/unstable. That may be overkill for this
> project, but the idea is there.
So the easiest solution for the packagers/di
> This hasn't been true for a while. That's the purpose of the
> mythjobqueue executable. It is a simple standlone program that
> scans the JobQueue for jobs and runs them (if you ran the backend setup
> program and turned on the settings to allow that host to run jobs).
I wasn't aware of it. I'
On Wed, 23 Feb 2005 18:14:37 -0600 (CST), Mark Matula
<[EMAIL PROTECTED]> wrote:
> On Wed, 23 Feb 2005, Isaac Richards wrote:
>
> > Ah, but you're missing bugzilla:
> > http://www.mythtv.org/bugs/show_bug.cgi?id=225
>
> Without meaning to sound at all like a smart-ass, a direct link on the
> myth
59 matches
Mail list logo