quoting Buchan Milne; Sunday 31 August 2003 11:29 am:
<snip>

> Well, you can check the majority of the data on the CDs, using 'rpm -K
> path/to/RPMS/*'. You can do this with the ISOs loop-mounted. Of course,
> this tells you nothing about DraxX, but if the RPMS all verify, chances
> are the 100MB of DrakX, stage2 etc should be ok.

I had already decided to go ahead anyway. What I find weird is I used the -c 
option ('always checksum') when I rsynced the ISOs in the first place; but I 
still see what I reported originally. I'm not certain as to what the hell I 
screwed up; but I'm positive it's me, not the system or tools at either end. 
:-)

> > I used -RWs so it
> > doesn't matter. Or should I just say screw it and do the usual hd.img
> > cooker install. That's fine for me; but a friend has an old Compaq he
> > wants to "sacrifice to the testing Gods" and I'd need disks for him
> > tomorrow.
>
> Give them a try and find out.

That's exactly what he said. "If we screw up it won't kill anything since the 
machine is a boat anchor without an OS. C'mon it _is_ a Compaq ya know!" 
Direct quote from the call I just made to him. It's his box.

> > I vaguely recall something similar in a previous release. Published
> > md5sums not matching the disks I mean. A couple of releases ago I think,
> > and I had already installed whichever version it was at the time without
> > trouble.
>
> The ISOs on ftp.sun.ac.za/iso-images/mandrake-beta match the md5sums, this
> mirror rsync's from sunet.se, so I don't think this is the case.
>
> Regards,
> Buchan

I'm sure you're 100% correct. I use the same mirror. Using the:

mkcd --checkmd5 <iso file>

method I get what I reported previously.

Using:
# md5sum -c 9.2rc1.md5sums.asc
MandrakeLinux-9.2rc1-CD1.i586.iso: OK
MandrakeLinux-9.2rc1-CD2.i586.iso: OK
MandrakeLinux-9.2rc1-CD3.i586.iso: OK

It's just another puzzle. I'm addicted to puzzles. For the mismatch I thought 
I recalled; my memory is to be trusted about as far as I can throw a fit. <g> 

Thanks for the response; and setting me straight. Again. 
You're good at that. :)

As for this lump; it gets fully cookered via a fresh install from those disks 
and a urpmi --auto-select in a few hours and we shall see what we see.

I've been suffering from test withdrawal for too long.

Best Regards;
Charlie
-- 
Edmonton,AB,Canada Registered user 244963 at http://counter.li.org


Reply via email to