cvs commit: src/sbin/bsdlabel bsdlabel.c

[Poul-Henning Kamp]

In message <20030828185445.1A96C2A7EA at canning.wemm.org>, Peter Wemm writes:
>"Poul-Henning Kamp" wrote:
>> In message <200308280815.12638.wes at softweyr.com>, Wes Peters writes:
>> 
>> >If I understand this correctly, this is a nice little change.  You're 
>> >trying to discourage people from just using the 'c' partition *as a 
>> >filesystem*, right?  
>> 
>> yes.
>
>I'd like to see this go further and have the magic 'c' partition go away.
>eg: instead of using ad0c or ad0s1c, use ad0 or ad0s1 as the 'whole
>partition'.    Having ad0c and ad0 as pseudo aliases for each other is a
>waste of a disklabel partition IMHO.

I agree, but could you just sign on this form for me:


  I Peter Wemm, promises to resolve any and all backwards compatibility
  issues in software and wetware that might arise from removing the 'c'
  partitions magicness in BSDlabels:

	

					____________________________
					(sign P. Wemm)

:-)

The bogus offset is currently recorded in the offset of the c partition
and changing the interpretation of that field will break some but
not by definition all BSD labeled disks out there.

If it the same to you, I would far rather work on discontinuing the
BSDlabel format going forward, and concentrate on a format with
fever restrictions of all sorts.


-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.