[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Parted CHS patch
From: |
Patrick J. LoPresti |
Subject: |
Re: Parted CHS patch |
Date: |
17 Aug 2004 09:46:13 -0400 |
User-agent: |
Gnus/5.09 (Gnus v5.9.0) Emacs/21.3 |
I apologize for the long delay in responding.
Andrew Clausen <address@hidden> writes:
> Yes and yes.
>
> What's a good UI?
>
> # parted /dev/hda 23583/128/32
> (parted) select /dev/hda 23583/128/32
Whatever is simplest for you. I will be invoking it programmatically,
anyway.
> Should Parted override this geometry if it can infer it from the
> partition table?
Definitely not by default. I want this feature because the partition
table may be wrong (e.g., a drive which is moved between incompatible
systems).
> Should it ask?
>
> Eg:
>
> You told Parted that /dev/hda has geometry 23583/128/32.
> However, the partitions on /dev/hda say that /dev/hda has
> geometry 14556/255/63. If you select Fix, then Parted will
> tell the partition table that 23583/128/32 is the correct
> geometry. If you select Ignore, then Parted will assume that
> 14556/255/63 is correct.
>
> Fix/Ignore/Cancel?
This is fine, as long as Parted's "-s" switch defaults to "Fix". I
invoke Parted from a script which knows exactly what geometry it wants
to use, regardless of what is already on the drive.
> If the user selects "fix", should Parted also replace the Bios Parameter
> Blocks in FAT and NTFS file systems?
This does not matter for my application, which typically involves
blowing away the entire partition table anyway. So I have no strong
opinion either way. In general, though, if I specify a geometry and
ask for it to be "fixed", you can assume I know what I am doing. So
my call would be to fix everything you can to match what I ask for.
Thanks!
- Pat