SCIENTIFIC-LINUX-USERS Archives

June 2011

SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Lamar Owen <[log in to unmask]>
Reply To:
Lamar Owen <[log in to unmask]>
Date:
Fri, 24 Jun 2011 13:20:00 -0400
Content-Type:
Text/Plain
Parts/Attachments:
Text/Plain (28 lines)
On Friday, June 24, 2011 12:21:23 PM you wrote:
> Thanks. I wasn't aware of [the distro-sync full] option. In most cases, this will fix
> most problems. 

Yeah; of course, it will be those with the corner cases that might gripe and talk about there not being any 'QA' or 'testing' being done.....  Been there, griped that.... :-)

> The only place it will not is where an rpm pre(),
> post(), triggers(), etc was the area of the fix. 

Yeah, these would be the hard cases.

> The key issue is how long it takes to find this problem versus doing a
> clean install. When I was younger I had no problems with spending 2
> weeks to find that a beta package removed an obscure symlink but the
> fix doesn't put it back. 

Indeed.  I'm still running essentially the same home directory was I did in pre-RH7 days.  As I was an early adopter of bero's vision of KDE, and thus went from Mandrake 5.3 up through the versions until today on Fedora 14.  The transition hasn't always been easy, and I have had to blow out .kde a few times, but otherwise still the same /home....  There's probably more cruft in there.....

So there's always going to be issues, and, as jbj told me years ago, RPM only deals in packages, not systems.

It would even be possible, though, for two system to update differently; say, I'm on SL6.0, and miss a few updates along the way to 6.1.  It's possible for two systems sitting at the same aggregate update level across all packages to be different if any interim packages did things differently; say I got foo-1.2.3-56.el6.x86_64 as part of 6.0, and an update to foo-1.3.2-2.el6.x86_64 was taken by one machine but not the other, and then both machines get foo-1.3.3-6.el6.x86_64 as part of 6.1, they might be different even if everything else is the same.  That goes to QA of every package's scriptlets, really, but it is hard and time-consuming work (and I'm not telling you something you don't already know!).  

This is why an early upgrader from, say, F14 to F15 may end up with a different system (and different problems) from one who waited three months to go F15.

Now, the upstream EL beta cycle and the SL alpha/beta cycle are somewhat different, since one would hope no wierd cases would arise between rebuilds of the packages between SL betas, but it obviously can't be ruled out.

Or to put it succinctly, RPM and yum are not ACID-compliant, since RPM really only cares about the packages, not about the system.

ATOM RSS1 RSS2