Actually I think our biggest problems stem from a deeply entrenched institutional culture that believes that endless unstructured discussion among a group of self-selected narrowly focused participants, without significant external supervision, and typically without any of the other steps associated with other disciplines of engineering, is a good way to do protocol design. To the extent it worked in the pre-Kobe days, it might have been more because of favorable circumstances than because of the superiority of that way of doing things:
- the Internet was small and relatively uniform and easy to understand;
- the IETF was small, with less turnover, and most of us knew each other;
- if you were talented and wanted to do Internet work, IETF participation was almost a no-brainer - the barriers to entry (financial and otherwise) were low and the working atmosphere was friendly;
- working groups were fewer and focused on less ambitious tasks;
- there was far less time-to-market pressure;
- the target environment was much less demanding and hostile, making the design job much easier;
- there was less pressure to bend/subvert the architecture; and
- there were a lot more academics and generalists involved in IETF who were naturally inclined to think about cross-area issues.
All of this discussion reminds me of the Nasrudin story where he's looking for his keys on the ground outside his house, because the light was better there, even though they were inside his house when he last saw them. People keep focusing on IESG when the biggest problems are at the WG level.
Keith
_______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf