ietf
[Top] [All Lists]

Re: The IETF Mission

2004-01-18 11:56:58
  *> 
  *> Lets take an example. I have been involved in QoS work, and there have been
  *> a number of specifications written on the subject; much of that started 
  *> with white papers, including especially
  *> 
  *> 0896 Congestion control in IP/TCP internetworks. J. Nagle.
  *>       Jan-06-1984. (Format: TXT=26782 bytes) (Status: UNKNOWN)
  *> 
  *> 0970 On packet switches with infinite storage. J. Nagle. Dec-01-1985.
  *>       (Format: TXT=35316 bytes) (Status: UNKNOWN)
  *> 

Fred,

Please note that these references are to published RFCs, which are
available in perpetuity in the official document archive of the
Internet community, the RFC series.  It is an illustration that
publishing significant white papers and discussion papers as RFCs has
real value, which is being lost lost in what you correctly characterize
as an over-emphasis on "standards".  We are letting the marketing types
rule.

  *> 
  *> To leave white papers and internet drafts, many of which are never 
  *> published as RFCs and a relatively small portion ever become standards, 
  *> out, and to leave the discussion part out is, I think, to leave out much of
  *> the real value of the IETF. Yes, both of those predate the IETF as we now 
  *> know it, but had the IETF existed then, they would have been very 
  *> appropriate in it. Today's counterparts include papers like some I 
  *> currently have posted (not intending to self-aggrandize, but they're the 
  *> one's I know most quickly). The posting of questions, problems, and ideas 
  *> is perhaps *the* key part; standards are from my perspective only one of 
  *> the products, and perhaps a byproduct. 


Yes. So let's consciously endeavor to ensure that sigificant
non-standards documents -- responsible position papers, white papers,
new ideas, etc.  -- become RFCs.  (Making Internet Drafts into an
archival series seems like a terrible idea to me, but that is a
different topic.)

Bob Braden



<Prev in Thread] Current Thread [Next in Thread>