Re: [patches] small fixes to some messages and man pages, and some questions

[Karel Zak <kzak@redhat.com>]

On Mon, Aug 15, 2011 at 08:34:29PM +0200, Benno Schulenberg wrote:
> 
> > > Also I do not understand the descriptions of the new options --nested,
> > > --chained, and --onesector.  [...]
> > 
> > The original comment from code: 
> > 
> > /* [...]
> > NESTED: every partition is contained in the surrounding partitions
> >    and is disjoint from all others.
> 
> I still fail to understand. 

 The each logical partition is defined by extended partition table
 (EBR) http://en.wikipedia.org/wiki/Extended_boot_record

 EBR contains usually two partitions:
    1/ data partition (used as a logical partition)
    2/ extended partition (used as a pointer to the next EBR)

 so if you have 5 logical partitions then there is usually 5 small
 partition tables and 5 extended partitions :-) The partitions are
 usually NESTED. 
 
 The top-level (outermost) extended partition is the partition 
 visible for end-users in fdisk output.

> If "contained in the surrounding partitions"
> means "contained in the containing partitions", it says nothing.  If it
> means "contained in the partitions that come before and after", this
> seems impossible to me.  The only way I can make sense of the
> above comment is to read it as follows:
> 
> NESTED: every extended partition is contained within a surrounding
> extended partition [except of course the outermost one] and is disjoint
> from all other extended partitions at the same nesting level.
> 
> (Here I take "disjoint" to mean "does not overlap", but then I still fail
> to understand how "mutually disjoint" is different.)

 NESTED: every extended partition is contained within a parental 
         extended partition. 

 IMHO nothing other is usable in Linux.

> 
> > CHAINED: every data partition is contained in the surrounding partitions
> >    and disjoint from all others, but extended partitions may lie outside
> 
> Is a data partition the same as a logical partition?

Yes. 

IMHO the comment is mess as well as the implementation :-) Don't think
about it too much. We need to rewrite sfdisk and cfdisk....

> >    scriptreplay [options] [-t] timingfile [typescript [divisor]]

 Fixed in the usage().

    Karel

-- 
 Karel Zak  <kzak@redhat.com>
 http://karelzak.blogspot.com