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

Re: [802SEC] Kibis in Kauai


I looked through the emails you referenced in your message, and
noticed that in one of them

we have a fine example of what can go wrong in the application
of P1541.  In this message, we find the statements:

  Take, for example, the two data rates that the new "10 Gb" 
  Ethernet PAR is addressing. There is real potential for 
  confusion here.


  And don't let the pronunciation bother you; I just pronounce "10 
  Gibit/s" as if it were "10 Gibits". A lot easier than "10 Gigabits 
  per second."

  So I think we need to start thinking about migrating from Gb/s to 
  Gibit/s within 802.

In fact, in 802.3 standards, the operating speeds are decimal multiples, 
not binary
multiples. We make normative references to decimal values, with the 
degree of precision.  10 Gb/s equals 10 000 000 000 bits/second (plus or 
clock tolerance) or 10 Gigabits/s, definitely not 10 Gibibits/second. 
 Geoff and I
had to explain this to SCC14 for both Fast Ethernet and Gigabit Ethernet 
as well,
because SCC14 wanted us to use Mebibit and Gibibit in those standards.

Imagine the impact if a heavy handed change was made to IEEE Std 802.3 to
replace all instances of Mb/s or Gb/s with Mib/s or Gib/s. The result 
would be
mass confusion in the industry and the marketplace, a broken standard, 
implementations, and a disaster for end customers.

This is not a trivial issue, even if it is really hard to pronounce 
these prefixes
with a straight face.  There are millions of pages of documents, from 
to engineering specifications, to product documents, to text books and 
that would have to be carefully examined and re-written before there would
be anything like a consistent usage of these prefixes, to say nothing of the
tons of marketing materials that are seen every day by consumers. It's 
40 years of trying  with the metric system, and our speed limit signs in 
the US
are still posted in MPH, and we still buy milk by the gallon. Still, I 
think the
metric system is a wonderful thing, and I wish my country would move faster
to adopt it. I see no need to throw a new prefix challenge into the mix, 
without the consent of the materially interested parties who will have 
to sort
out the mess.


Roger B. Marks wrote:

> Howard,
> As background to my comments, I should say that, though I am not in 
> any way involved in this standards project, I am in favor of it.
> People may misunderstand the history of this project when they read 
> that the PAR was approved in May 2002. That was a revised PAR. The 
> original PAR was approved on 8 December 1998. (I just looked it up.) 
> The draft has been around for a very long time too. Perhaps that's why 
> the Sponsor Ballot proceeded so quickly.
> This topic has not been a secret. For example, Jim Carlo forwarded the 
> SEC a coordinating note on this issue almost three years ago:
> Attached to that note was P1541 Draft 2 (dated 1999-12-24) including 
> virtually the same language as your Slide 2.
> That coordinating note invited comment. I wrote to Bruce Barrow, and 
> he replied.
> Even earlier, and independently, I wrote to the SEC on this issue in 
> December 1999:
> with a followup by Vic:
> My note cited a few other sources, including an August 1999 IEEE 
> Spectrum article quoting Bruce Barrow:
> as saying "it is unlikely the standard would be adopted before 
> mid-2000." (Good prognostication!)
> Roger
> At 1:13 AM -0800 02/11/05, Howard Frazier wrote:
>> Dear IEEE 802 LMSC SEC Members,
>> I have asked Paul for a brief amount of time to
>> make a presentation at the opening plenary
>> meeting in Kauai on a subject that will come
>> before the IEEE-SA Standards Board in
>> December.  Attached please find a set of
>> slides that will explain the issue.
>> The topic is important, even though it is
>> sometimes hard to discuss it with a straight face.
>> I won't have enough time to present all of
>> the slides at the plenary, but I plan to cover at
>> least the 2nd and 3rd slides, and possibly
>> the 4th.  The remainder have been included
>> to help explain the issue, and hopefully
>> stimulate debate.
>> The Standards Board needs to hear from
>> people who have a stake in the subject.
>> For reasons outlined in the slide deck, I
>> believe that materially interested parties are
>> completely unaware of actions that have been
>> taken to date, and are not aware of the certain
>> impact of IEEE draft standard P1541.
>> I hope that you will see fit to share this material
>> with your working groups, and to elicit their
>> feedback.  I believe that if the LMSC takes a
>> position on this subject, it can have an effect
>> on the Standards Board's decision.
>> Here's the key to the whole thing: Look at
>> the "shall" statements on slide 2, and ask yourself
>> if all new and revised IEEE standards should have
>> to conform to them, because if P1541 becomes
>> an IEEE standard, that's what will happen.
>> Howard Frazier
>> Attachment converted: TiDrive:kibi.pdf (PDF /CARO) (00130AB2)