Re: [PATCH v2 3/3] count-objects: add human-readable size option
- From
Junio C Hamano <gitster@pobox.com>
- Date
- Aug 15, 2008, 00:10 UTC
- Message-ID
- <7vabffcf4e.fsf@gitster.siamese.dyndns.org>
- In-Reply-To
- <48A4C521.6090507@griep.us>
Marcus Griep <marcus@griep.us> writes:
Show 6 quoted lines
> Petr Baudis wrote: >>> +-H:: >> >> Can you guess what would I bug you about? ;-) > > If we get '-h', I'll submit another patch :-P
Wasn't the problem about documenting the addition of pack size, which is not about this human-readable option?
Show 8 quoted lines
>> If it's non-human-readable anyway, why are you dividing this by 1024? At >> any rate, it is not obvious at all that the size-pack is not actually >> size-pack but size-pack/1024. You should either add the (fixed) unit >> string behind or name it size-pack-kb - or just not divide it at all? > > I divide by 1024 here because the loose object size is reported in KiB. > The total that ends up in size_pack is in B, hence to be consistent, I > report the pack size in KiB as well.
I thought this part was ok without room for dispute.