At 4:51 PM -0500 1/12/00, Garrett Wollman wrote:
><<On Wed, 12 Jan 2000, Garance A Drosihn <[EMAIL PROTECTED]> said:
>
> > In 'ls' we are not talking about a block count, we are talking about
> > a byte-count.
>
>ls -s
Hmm, valid point. 'ls -l' is not using a block count though, and so
all of my previous comments still make sense for 'ls -l' and the new
option. 'ls -s' and the new option should follow a block count, but
then 'ls -s' is showing a block count for ALL files that it lists.
The examples I gave in my previous message don't end up as confusing
for 'ls -s' as they do for 'ls -l'. Maybe the new option should not
even apply to 'ls -s'? I have no preference for 'ls -s' behavior.
In any case, I'm still saying that in practice, people will find it
less confusing if 'ls -l' used 1000 as the divisor, not 1024. Or at
least, I found it less confusing, when I have done this same thing.
Yes, it may be "more pure" to use 1024 when comparing 'ls' listings
to block counts, but it is less confusing WITHIN a single 'ls -l'
listing if all the numbers are decimal, and not some combination of
base-10 and base-2. 'ls -l' listings (without the -s...) are already
a problem when comparing to 'du' or 'df' listings.
---
Garance Alistair Drosehn = [EMAIL PROTECTED]
Senior Systems Programmer or [EMAIL PROTECTED]
Rensselaer Polytechnic Institute
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
