linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Andries.Brouwer@cwi.nl
To: bcrl@redhat.com, cw@f00f.org
Cc: esr@thyrsus.com, garfield@irving.iisd.sra.com,
	linux-kernel@vger.kernel.org
Subject: Re: Configure.help editorial policy
Date: Sun, 23 Dec 2001 03:40:56 GMT	[thread overview]
Message-ID: <UTC200112230340.DAA66334.aeb@cwi.nl> (raw)

Benjamin LaHaise writes:

> If you think GB == 1000^3, then please go "correct" all the DRAM 
> manufacturers out in the world.  They just sent me 1GB of ram and
> it's coming up as 1073741824 bytes.

Oh, please - must we have this same discussion every year?


If someone uses an approximate value in some context then that
- is desirable if it is more convenient
- is admissable either (i) if greater precision is not of interest
  (or unavailable), or (ii) when the approximate value suffices to
  derive the exact value uniquely.


When things come in power-of-two sizes then no confusion is
possible if you use a decimal prefix: in reality the nearest
power of two is meant. Thus there is no urgent reason to be
precise when talking about the size of memory modules.
If you say 1GB, that is an American billion bytes, everybody
will assume that in fact the size was 1073741824 bytes.
Of course this will change when technology changes and memory
comes in arbitrarily sized units, like disks today.


These remarks taken together mean that it is not often necessary
to use these newfangled abbreviations like GiB or words like
gibibyte. Often precision is not required. Or precision is required
but the precise size is clear from the context.

But it is a serious mistake to doubt G = 1000^3.
k, M, G have one and only one meaning.
Just like "thousand" has only one meaning and still one can say
that this thing, that cost $1049, was bought for a thousand bucks.


Disks do not come in power-of-two sizes. So when talking about
disk sizes there is no "he says 1000000000 but it must be
a power of two so he must mean 1073741824". No, for disk sizes
it is just "he says 1000000000 so that is it".


In standard texts and other places where there must not be
any ambiguity one uses KiB, MiB, GiB. Also in a context where
binary and decimal units are both used it is clean to
separate them. When Linux boots, I see

hdf: 120064896 sectors (61473 MB) w/2048KiB Cache

and you see that the MB is decimal and the KiB binary. Good.


Concerning this Configure.help stuff, clearly it is not very
important what is written, but GiB is slightly better than GB,
and I doubt it would lead to any problems. People who have
never seen GiB will probably read it as gigabyte, and that is
approximately right.

Concerning the future of this standard for binary prefixes,
it looks like scientists and engineers like them and are
careful to distinguish G and Gi; in the open software world
programs are slowly adapted to correct usage; Microsoft has
not yet heard about the difference between GB and GiB.

Andries

             reply	other threads:[~2001-12-23  3:41 UTC|newest]

Thread overview: 115+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-12-23  3:40 Andries.Brouwer [this message]
  -- strict thread matches above, loose matches on Subject: below --
2001-12-30 12:06 Configure.help editorial policy Martin Knoblauch
2001-12-30  4:10 Timothy A. Seufert
2001-12-27 20:20 Andries.Brouwer
2001-12-27 15:42 Andries.Brouwer
2001-12-27 17:33 ` Timo Jantunen
2001-12-25 20:54 Andries.Brouwer
2001-12-24 11:08 Matt Reuther
2001-12-23 23:44 Andries.Brouwer
2001-12-23  4:06 Andries.Brouwer
2001-12-25 11:40 ` Pavel Machek
2001-12-22 20:32 Timothy A. Seufert
2001-12-27 12:26 ` Kai Henningsen
2001-12-27 14:55   ` Davidovac Zoran
2001-12-22 13:30 Per Jessen
2001-12-22  2:42 Thomas Hood
2001-12-21 19:34 Timothy Covell
2001-12-21 20:24 ` Chris Ricker
2001-12-22  2:01   ` Timothy Covell
     [not found] <Pine.LNX.4.33.0112201605310.9934-100000@coffee.psychology.mcmaster.ca>
2001-12-20 21:28 ` Reid Hekman
2001-12-28 13:10   ` David Woodhouse
2001-12-27 12:10 ` Kai Henningsen
2001-12-20 21:17 Petr Vandrovec
2001-12-20 19:32 Eric S. Raymond
2001-12-20 20:27 ` Reid Hekman
2001-12-20 20:23   ` Eric S. Raymond
2001-12-22  0:04     ` Rob Landley
2001-12-20 22:41   ` Mike Eldridge
2001-12-21 15:47   ` Matthias Andree
2001-12-20 22:49 ` Vojtech Pavlik
2001-12-20 23:31 ` David Garfield
2001-12-20 23:52   ` Eric S. Raymond
2001-12-21  0:20     ` Tom Rini
2001-12-21 18:43   ` David Garfield
2001-12-21 18:40     ` Eric S. Raymond
2001-12-21 19:18       ` Benjamin LaHaise
2001-12-21 20:09         ` Oliver Xymoron
2001-12-21 23:03           ` Jeff Mcadams
2001-12-27 10:35             ` Kai Henningsen
2001-12-21 20:10         ` Chris Wedgwood
2001-12-21 20:31           ` Benjamin LaHaise
2001-12-21 20:36             ` Rik van Riel
2001-12-21 20:47               ` Benjamin LaHaise
2001-12-21 21:00                 ` Chris Wedgwood
2001-12-21 21:06                   ` Rik van Riel
2001-12-21 21:10                   ` Benjamin LaHaise
2001-12-21 21:33                     ` Thomas Dodd
2001-12-21 22:05                       ` Nicholas Knight
2001-12-22  0:07                     ` Oliver Xymoron
2001-12-22  0:27                       ` Benjamin LaHaise
2001-12-23 22:46                     ` Cameron Simpson
2001-12-21 20:57             ` Chris Wedgwood
2001-12-21 21:03               ` Benjamin LaHaise
2001-12-21 21:09                 ` Rik van Riel
2001-12-21 21:28             ` Oliver Xymoron
2001-12-27 11:12             ` Kai Henningsen
2001-12-27 11:22             ` Kai Henningsen
2001-12-27 11:18           ` Kai Henningsen
2001-12-21 20:11         ` Timo Jantunen
2001-12-22  0:32         ` Alan Cox
2001-12-22 16:14           ` Vojtech Pavlik
2001-12-27 19:44             ` Allan Sandfeld
2001-12-27 20:18               ` Acrimon Beet
2001-12-27 10:45         ` Kai Henningsen
2001-12-27 14:19           ` Vojtech Pavlik
2001-12-21 21:17       ` Rik van Riel
2001-12-23 22:42         ` Cameron Simpson
2001-12-23 22:53           ` Rik van Riel
2001-12-23 22:46             ` Eric S. Raymond
2001-12-26 17:44               ` Riley Williams
2001-12-26 22:17                 ` Cameron Simpson
2001-12-26 23:34                   ` Dominik Mierzejewski
2001-12-27  0:08                     ` Riley Williams
2001-12-27  0:52                       ` Dominik Mierzejewski
2001-12-27 21:34                         ` Riley Williams
2001-12-27 23:27                         ` Pavel Machek
2001-12-27  6:02                     ` Daniel Phillips
2001-12-27 11:24                       ` Dominik Mierzejewski
2001-12-27 15:09                         ` Martin Mares
2001-12-28  3:23                         ` Daniel Phillips
2001-12-28  9:58                           ` Matthias Andree
2001-12-27  0:09                 ` Eric S. Raymond
2001-12-27  0:39                   ` Dominik Mierzejewski
2001-12-27 19:46                     ` Riley Williams
2001-12-27 11:46               ` Kai Henningsen
2001-12-27 21:42                 ` Riley Williams
2001-12-23 23:00             ` Cameron Simpson
2001-12-23 23:10               ` Rik van Riel
2001-12-23 23:12                 ` Eric S. Raymond
2001-12-24  6:25           ` Mike Galbraith
2001-12-27 15:15             ` Gábor Lénárt
2001-12-21 22:53       ` Stephen Satchell
2001-12-21 22:55         ` Eric S. Raymond
2001-12-21 23:07         ` Nicholas Knight
2001-12-22 19:40         ` T. A.
2001-12-22  0:12       ` Rob Landley
2001-12-22  9:58         ` Eric S. Raymond
2001-12-23  9:40           ` Rob Landley
2001-12-23 19:11             ` Eric S. Raymond
2001-12-27 11:35         ` Kai Henningsen
2001-12-23  9:47       ` David Woodhouse
2001-12-27 11:41       ` Kai Henningsen
2001-12-21 19:12     ` David Weinehall
2001-12-22  4:49       ` Keith Owens
2001-12-21 20:23     ` David Garfield
2001-12-21 20:56       ` Eric S. Raymond
2001-12-21 10:36 ` Mike Jagdis
2001-12-22  0:23   ` Rob Landley
2001-12-22  8:33     ` Eric Windisch
2001-12-22  8:53     ` Alan Cox
2001-12-23  9:17 ` David Woodhouse
     [not found] ` <esr@thyrsus.com>
2001-12-27 11:57   ` Kai Henningsen
2001-12-27 14:22     ` Vojtech Pavlik
2001-12-27 16:42       ` Alan Cox
2001-12-29 17:22         ` Dan Hopper

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=UTC200112230340.DAA66334.aeb@cwi.nl \
    --to=andries.brouwer@cwi.nl \
    --cc=bcrl@redhat.com \
    --cc=cw@f00f.org \
    --cc=esr@thyrsus.com \
    --cc=garfield@irving.iisd.sra.com \
    --cc=linux-kernel@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).