From: pmacdona@sanjuan (Peter MacDonald) Subject: Re: SLS package installation Date: Tue, 8 Sep 1992 22:08:49 GMT
In article <trussell.715939665@cwis> trussell@cwis.unomaha.edu (Tim Russell) writes:
>bugs@cs.montana.edu (Dave Seelmeyer) writes:
>
>>their job, and all is ok so far. The problem is when I want to do a full
>>installation (full base and X11). I have tried sysinstall -all and also
>>doinstall /dev/hda2. With the first, it asks to insert disk #3 in the
>>floppy drive and press return. Upon doing this, it repeats the message.
>>It will do this 3 times and return to the prompt. With the second, it
>>starts with the same disk #3 message and then responds with "magic match
>>failed" or somesuch and comes back to the prompt.
>
> Peter doesn't want to admit it, but apparently even his own kernel
>didn't get his patches right. According to him, his kernel is supposed
>to be patched so it automatically figures out what kind of a filesystem
>it's mounting, but it doesn't work.
Well, begining to wonder if I screwed up, I just downloaded the SLS
1.Z and tried it. It worked as advertised, that is, it mounts dos
disks correctly. Since many others manage the install without problem
something you are doing is different than them. Not knowing what else
to do, I reiterate my warnings.
1) The "magic match failed" message is always seen for DOS disks because
the kernel is trying "minix fs" first, then "ext fs" then "dos".
2) IF YOU DON'T DOWNLOAD *ALL* FILES (including the files "disk3"...)
THE INSTALL PROGRAM FAILS, BECAUSE THEY VERIFY THE CORRECT ORDER.
That is, disk 3 must contain the following:
BASE.TAZ
DISK3
...
The file DISK3 is not just a description file that you can omit.
BTW:
The next release of SLS will be .98 and will be around the end of
September (earlier if possible) and will be entirely jump table,
and thus will be upgradable with just kernel/shlibs, plus a few
fixed binaries. It will also have improved documentation, as
well as better warning messages about things like above.
I would like to see Interviews included as well (idraw et al).
Peter
pmacdona@sanjuan.uvic.ca