<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<br>
<br>
<div class="moz-cite-prefix">On 8/7/2015 11:44 AM, Kirk Bocek wrote:<br>
</div>
<blockquote cite="mid:55C4FC75.3080201@kbocek.com" type="cite">
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
<br>
<br>
<div class="moz-cite-prefix">On 8/7/2015 11:24 AM, Frank Phillips
wrote:<br>
</div>
<blockquote
cite="mid:CAADtG4j23cETTSNtctXeurn0aGD2u9FZWbS3xbRbB+ys45m=SQ@mail.gmail.com"
type="cite">
<div dir="ltr">
<div>On Thu, Aug 6, 2015 at 10:20 AM, Kirk Bocek <span
dir="ltr"><<a moz-do-not-send="true"
href="mailto:t004@kbocek.com" target="_blank">t004@kbocek.com</a>></span>
wrote:<br>
</div>
<br>
<snip><br>
<br>
<div class="gmail_extra">
<div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex"><span
class=""><br>
<br>
On 8/6/2015 5:30 AM, Michael T. Dean wrote:<br>
</span><br>
2. /var/lib/mythtv/.bash_profile looks like:<br>
<br>
#See nvidia readme:<br>
#export __GL_SYNC_TO_VBLANK=1<br>
/usr/bin/setterm -blank 0 -powerdown 0<br>
<br>
if [ -z "$DISPLAY" ] && [ $(tty) == /dev/tty8 ];
then<br>
while [ 1 == 1 ]<br>
do<br>
startx<br>
sleep 4<br>
/usr/bin/xset -display :0 s noblank<br>
/usr/bin/xset -display :0 s off<br>
/usr/bin/xset -display :0 -dpms<br>
done<br>
fi<br>
<br>
One host is Nvidia, this one is the Liva so the Nvida
line is commented out.<br>
<br>
3. /var/lib/mythtv/.xinitrc looks like:<br>
<br>
xset -dpms s off<br>
xsetroot -solid black<br>
xhost +<br>
#See nvidia readme:<br>
export __GL_SYNC_TO_VBLANK=1<br>
#/usr/bin/pkill irexec<br>
#/usr/bin/irexec -d /var/lib/mythtv/.lircrc<br>
/usr/bin/mythfrontend --logpath=/var/log/mythtv<br>
#mythtv-setup<br>
#/usr/bin/fvwm &<br>
#xterm<br>
<br>
So, if I want a simple xterm window for testing, I
uncomment those lines and comment out the mythfrontend
line and pkill X. My loop in .bash_profile means if
mythfrontend crashes or exits it will restart after four
seconds. I actually find it pretty useful.<br>
<br>
Does anyone see any problems with this startup process?
<div class="HOEnZb">
<div class="h5"><br>
</div>
</div>
</blockquote>
<div><br>
</div>
<div> With the blue background covering a responsive
frontend, for some reason the screen stack is out of
order the first time you start. I don't know if it will
help, but I have sometimes used a sleep command before
mythfrontend.<br>
<br>
xset -dpms s off<br>
ratpoison &<br>
sleep 3 && mythwelcome --logpath /tmp --loglevel
crit<br>
<br>
</div>
<div>You could try starting a window manager before
mythfrontend (the & on it is important).<br>
</div>
</div>
<br>
</div>
</div>
<br>
</blockquote>
<br>
Now, in my other VAAPI thread I was told to turn on OpenGL. Guess
what? That fixed *this* problem on my ECS Liva. I restarted the
host 8-10 times and it worked every time.<br>
<br>
Now I hadn't thought of using the window manager. I know the WM
itself is pointless since Myth won't use it. I'll try the sleep
too to see if X needs to settle down.<br>
<br>
The other host having this problem is quite a bit slower starting
up than the ECS Liva so this might be applicable.<br>
<br>
mythwelcome? You meant mythfrontend right?<br>
<br>
</blockquote>
<br>
Well thank you Frank. The window manager idea solved the blank
screen on my other, very slow booting Nvidia host. It had always
shown the blue background and then briefly flashed a white screen.
But now it goes to the main menu after the flash instead of staying
on the blue background. So *does* mythfrontend use a window manager
if available?<br>
<br>
I installed twm which is easily available on CentOS 6. I'm sure I
could have found others.<br>
<br>
</body>
</html>