[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: How to configure konqueror to show KB and MB instead of KiB and
From:       Michael Pyne <mpyne () kde ! org>
Date:       2009-07-02 0:17:56
Message-ID: 200907012018.02248.mpyne () kde ! org
[Download RAW message or body]

[Attachment #2 (multipart/signed)]


On Wednesday 01 July 2009 12:36:52 Josh Berry wrote:
> However, I think it would be perfectly fine to have an option to
> switch units.  Show "KB" (= 1000 bytes), or show "KiB" (= 1024 bytes).
>  But I am opposed to any option that lets 1 KB = 1024 bytes.

Say again?

There's no way I'd support a patch that changes away the meaning of "KB" away 
from the commonly used 1024 bytes, at least not while "KB" is still in active 
use.

If we were to "change the meaning back" to kB it could only occur after 
deprecating KB in favor of KiB, getting buy-in to that from enough computing 
vendors (both software and hardware) to eliminate usage of 1024 byte KB, and 
only then taking the unit back to mean 1000 bytes.

And still no one has explained how using context to figure out KB == 1024 
bytes is hard for people, yet these same people understand what a kibibyte is.  
(I am completely uninterested in the case where people understand neither, 
especially given how the prevalence of Windows ensures that they will continue 
to learn KB == 1024 before they ever touch on kibibytes)

I understand that the error between SI prefixes and 2^(10n) gets most drastic 
as n goes up.  I sympathize, really I do.  But let's take this example, 
someone buys a 500 GB hard disk.  Hard disk manufacturer means 500 * 10^9 
bytes, not 500 * 2^30 bytes.

With KDE 3 style units, kdf would have shown approx a 465 GB hard disk.
With KDE 4 style units, kdf would have shown approx a 465 GiB hard disk.

It's the same damn number... it's still going to lead to confusion.

Regards,
 - Michael Pyne

["signature.asc" (application/pgp-signature)]

>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<


[prev in list] [next in list] [prev in thread] [next in thread]