<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
On 6/25/2009 4:36 AM, Alan Cox wrote:
<blockquote cite="mid:20090625123624.1a26e11f@lxorguk.ukuu.org.uk"
type="cite">
<pre wrap="">On Thu, 25 Jun 2009 04:07:07 -0700
Kevin Bowling <a class="moz-txt-link-rfc2396E" href="mailto:kevinb@analograils.com"><kevinb@analograils.com></a> wrote:
</pre>
<blockquote type="cite">
<pre wrap="">2 out of 5 failures. F10->F11 is completely unusable for any kind of
"uncommon" setup, i.e. LDAP login, Linux RAID, Xen DomU. Anything other
than one IDE hard disk with default layout really.
</pre>
</blockquote>
<pre wrap=""><!---->
2 out of 5 ? - I got 3 out of 4 fails including hitting error paths that
clearly were not tested because the error printing code didn't work but
spewed python traces - and yes its in bugzilla already.
</pre>
</blockquote>
One of the "success" required manual heavy lifting. F11 Xorg didn't
like the R100 radeon in my IBM x3650 server and absolutely crawled.
After two DAYS I interrupted and rebooted the box before it was done
doing package cleanups. Some careful work manually sorting out the
mess seemed to get it going.<br>
<br>
preupgrade-cli should not use a frame buffer of any type. Think about
TTY-only devices, or server BIOS that redirect console to
ethernet/serial...<br>
<br>
Anyways, be careful! Clearly not for the average user, and woefully
under-tested.<br>
</body>
</html>