Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802SEC] FW: Just when you thought you had all this figured out


Please recall that we (802) sent a letter to the IEEE-SA
Standards Board asking that P1541 (standard for prefixes
for binary multiples) be approved as a 2 year, Trial Use
standard, rather than a 5 year standard.

During the trial period, you have the opportunity to
comment on the contents of the standard. Your comments
must be addressed before the standard is elevated to
full use status. Note that while the trial period ends
two years after the standard was approved, the cutoff
date for comments is set in advance of the end of the
trial period. If substantive comments are submitted
that require changes to the draft, the sponsor must
seek a PAR to revise the draft, whereupon it will be
subjected to the normal development and balloting
process. Furthermore, since the balloting on SCC14
standards is now "open" to materially interested
parties, you will have the opportunity to join the
balloting group on any revision of P1541.

I can't seem to find my latest greatest copy of P1541,
which contains the instructions for submitting comments.
If any one has a copy of the published standard, please
post the instructions. For now, I have attached a
copy of the draft as it was submitted to RevCom.

On a technical note, I totally agree with Pat's position
that it is better to use a precise representation of
a binary number (such as 65 535, or 2^16 - 1 bytes) rather
than an imprecise and unfamiliar abbreviation (such as
64 kibibytes), and that this representation must be
included in the normative portion of the standard. As
far as I am concerned, the kibis, mebis and gibis should
be placed in an informative annex, for the amusement
(or bemusement) of the reader.

Let me also emphasize that inaction will result in
P1541 being elevated to a 5 year, Full Use standard,
whereupon we will simply have to live with it.

Howard Frazier

Stevenson, Carl R (Carl) wrote:

> Beats me ... as the original usenet author said,
> "Just when you though you had this all figured out." :-)
> Frankly, I thought the whole "ibi" thing was rather
> stupid ... but that's just my personal view.
> Carl
>>-----Original Message-----
>>From: []
>>Sent: Monday, August 04, 2003 1:22 PM
>>Subject: RE: [802SEC] FW: Just when you thought you had all 
>>this figured
>>So, for the cases where megabyte was used to mean (10^3) * 
>>(2^10) bytes, will this now  be a kilokibibyte?
>>My objection to the IEEE standard on the same topic was that 
>>it appeared to require using the new units over other 
>>unambiguous options. In particular, when we use numbers such 
>>as 65,536 or (2^32 - 1) in a standard, we often mean exactly 
>>that number. Expressing it as 64 kibibytes wouldn't express 
>>the precision we need (as shown by the number often being a 
>>power of 2 minus 1 because it is represented by a bit field).
>>The confusion occurs more in the real world than in 
>>standards. I haven't read any standards that left themselves 
>>ambigous on number of bytes. They generally write the number 
>>out as above or as a power of 2.
>>-----Original Message-----
>>From: Stevenson, Carl R (Carl) []
>>Sent: Saturday, August 02, 2003 3:03 PM
>>To: ''
>>Subject: [802SEC] FW: Just when you thought you had all this 
>>figured out
>>FYI ... this was posted to one of the usenet ham
>>newsgroups with the above subject :-)