<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 12pt;
font-family:Calibri
}
--></style></head>
<body class='hmmessage'><div dir='ltr'><br><div>> > SUBSYSTEM=="dvb", \<br>> > ATTRS{device}=="0x8852", \<br>> > ATTRS{subsystem_vendor}=="0x6981", \<br>> > PROGRAM="/bin/sh -c ' K=%k; K=$${K#dvb}; N=$${K#*.}; \<br>> > if [ ! -e /dev/dvb/adapter_t1/$N ] ; \<br>> > then printf dvb/adapter_t1/%%s $${K#*.}; \<br>> > else printf dvb/adapter_t2/%%s $${K#*.}; \<br>> > fi ; \<br>> > exit 0'", SYMLINK+="%c"<br>> <br>> If I look at this rule (I made it a bit more readable) it is in<br>> essence 'first come first get'. The first demux to initialise gets<br>> placed in 'adapter_t1' the second in 'adapter_t2'. The same for the<br>> other devices. Obviously they initialise in parallel, creating the<br>> mixup.<br>> I assume that the distinction between dvb0 and dvb1 is made on the<br>> hardware order, if not, if this is also based on 'first come first<br>> get' there is, as I see it no way to distinguish the two adaptors.<br>> My next assumption is that the distinction between adaptor0 and<br>> adaptor1 is equal to the distinction between dvb0 and dvb1. So any<br>> udev-rule can not add anything in the distinction.<br>> <br>> Tot mails,<br>> Hika mailto:hikavdh@gmail.com<br><br>Thanks for your post, Hika<br><br>The parallel initialisation you mention really looks like the only explanation. I was hopefull, though...<br><br>To be honest, the pc is almost never off, so as long as the built-in udev rules make sure that adapter0 always points to satellite x and adapter1 to y, all is ok, even without fancy names.<br><br><br>Yianni<br></div> </div></body>
</html>