<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 TRANSITIONAL//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; CHARSET=UTF-8">
<META NAME="GENERATOR" CONTENT="GtkHTML/3.18.2">
</HEAD>
<BODY>
On Wed, 2008-06-11 at 15:44 -0700, Cliff Wells wrote:<BR>
<BLOCKQUOTE TYPE=CITE>
<TT>It really is a problem with packages (there's clearly no reason</TT><BR>
<TT>package-1.1 couldn't use dependency-1.0, other than the packager</TT><BR>
<TT>explicitly requiring 0.9), so I'm not sure this within Smart's ability</TT><BR>
<TT>or scope to address. </TT><BR>
</BLOCKQUOTE>
<BR>
The one thing I wish Smart allowed was to set a channel to disabled only for the session when using CLI. I run smart 'smart upgrade --update' a LOT. I do that in SSH terminal sessions. The most problematic (I'm a Fedora 9 user but this goes back many iterations of Fedora) is ATRPMS. As you say, probably because I also include some Livna packages. But the most direct solution is to temporarily disable livna and atrpms so the CORE packages can be updated. Also ran into this a lot with Mono repos of various stripes (why is it so hard to keep monodevelop current on fedora?!?!?!? - but I digress).<BR>
<BR>
The fact that I get these loops is actually a feature. If Smart were as incapable as any other PM available (not including Apt, btw) then I'd not have loops. I'd have instead, what I had before Smart, unresolvable updates that could only be resolved by downloading specific packages, issuing rpm -e (sometimes even with nodeps), etc.<BR>
<BR>
very very glad to have it, no matter what the toolkit ;)<BR>
<TABLE CELLSPACING="0" CELLPADDING="0" WIDTH="100%">
<TR>
<TD>
-- <BR>
Richard Hendershot <<A HREF="mailto:rshendershot@mchsi.com">rshendershot@mchsi.com</A>>
</TD>
</TR>
</TABLE>
</BODY>
</HTML>