bug-parted
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Disk addresses in MB


From: Andrew Clausen
Subject: Re: Disk addresses in MB
Date: Tue, 9 Jul 2002 11:50:19 +1000
User-agent: Mutt/1.3.28i

On Tue, Jul 09, 2002 at 09:41:53AM +0800, Mathias Koerber wrote:
> but just before. For switching them while parted is running
> a simple command ´units xxx´ may suffice.

Yeah, that's what I'm thinking.  Just, maybe there will be lots
of other settings?  Can you think of any?  I can't.

> I thought you were looking fora *default* setting...

Default as in "input" (if you leave out a unit specifier)

> >Why is this important?
> 
> because when I twiddle with my partition table, I want to be
>       a) in control

You aren't in control.  Sorry.  The values you type in are only
approximations and *will* be changed by parted, to something that will
work.  If the only posssibilities are very different from what you typed,
Parted will warn you.

>       b) at least in the know exactly what parted is doing

Giving units won't tell you that.  Perhaps really verbose messages
might.  I doubt they would be helpful though.

>       c) able to understand what it´s going to do

You don't want to understand, unless you have a particular interest
in partition tables.  It's **REALLY** complicated.  Seriously.
(sfdisk's expert mode will show you what's really going on)

> Let's just say an interface that deas with discrete units is much more
> confidence inspiring than something that converts such addresses into
> a fractional amount, potentially rounding or truncating.

I guess UI's that inspire confidence are nice.  Just, I don't think
it's possible to give an illusion that everything is "what you see
is what you get".  Like, it should feel like everything is an
approximation (and that approximations are fine), because that's
the reality behind it.

> The highest precision I have seen parted use is 3 digits,
> which means 1KB of space. MOst disks use sectors of
> 512Bytes and partition in clusters of x of them.
> As shown in my earlier example, sometimes parted
> shows 0.31MB (31KB) between partitions, sometimes
> 0.32. To me that looks like a rounding error, which
> I´d rather not have to worry about when dealing with
> addresses which are to all intents and purposes integers
> internally.

But, the integers are unintuitive, IMHO.  And they aren't particularly
useful.

> I *like* parted and what it can do, I just think it should
> use more practical units..

Well I think they are *practical*, although perhaps not confidence
inspiring.

Cheers,
Andrew




reply via email to

[Prev in Thread] Current Thread [Next in Thread]