If you stop o think about it, the use of the prefixes for decimal numbers predated the use for binaries by a long time. I doubt that you will find the usage for 1024^n until the second half of the 20th century. It is the 1024^n that is the latecomer to the party. In that sense, you are correct in stating that it is "right".
 
The SI folks have made a concession by saying that it is OK to use the prefixes for 1024^n numbers other than computer memory and  clock speeds as long as you state your usage ahead of time. I hope that ends the silly insertion of BI into the prefixes.
 

Regards,
Richard Schuh

-----Original Message-----
From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED]On Behalf Of Miguel Delapaz
Sent: Friday, October 06, 2006 2:26 PM
To: [email protected]
Subject: Re: I know it's dumb, but.......


Yes, terribly silly.  Also, OSes generally express file size in 1024^n which can be confusing when determining how many files you can cram on a disk with capacity 1000^n.  Have fun trying to get everyone to change though :-)  1000^n is obviously "right"...but who wants to go and make up different terms when we're "only" off by 24? :-)

More wikipedia: http://en.wikipedia.org/wiki/Megabyte

Regards,
Miguel Delapaz
z/VM TCP/IP Development


The IBM z/VM Operating System <[email protected]> wrote on 10/06/2006 02:09:41 PM:

> :soapbox.
> I am a little concerned that there still is the discrepancy between
> 1000^n and 1024^n depending on the context. While we could get away
> with this when n was 1, now that we're at n=3 the error is significant
> when you confuse them.
> At one point I believe people declared that in disk storage context a
> GB would mean 1000^3 where in memory it is 1024^3. I think that's
> silly. So how about virtual memory and paging? When we get to Exabytes
> the difference is as much as between an 3390 on MVS and on VM.
>
> Rob

Reply via email to