RE: IETF & this list

[Andy Bierman <abierman@cisco.com>]

At 06:29 AM 1/11/2003 -0800, Randy Bush wrote:
><first cuppa>
>
>> I think it's the OPS ADs responsibility to do something about the
>> 'NM problem'.
>
>it would seem hard to explain this position in light of the
>discussions on problem-statement, where i interpret non-trivial
>sentiment against AD-driven architecture.

I didn't say the ADs should write the RFCs.  The ADs should
put a process in motion (e.g., charter a WG) to do that.


>> A lot of good input has been collected from the operator
>> community over the last 1+ years
>
>i would have a hard time couching it so positively.  it struck me
>as a combination of ear-wax/denial removal and rudely silencing
>some of the ops who want to talk about trivia.  but this is my
>first cuppa.

I think a lot of specific issues have been identified as a result
of the OPS roadshow and IAB NM workshop meetings.


>to be personally honest, i have had some ideas and cravings about
>where we should go and how we might start to get there.  but i have
>been unable to communicate them well enough to entice anybody, or
>maybe they're just wrong.  
>
>i also don't think it's a quick fix, especially one of the style
>that some folk on the problem-statement list think should rule the
>ietf.  to be really pessimistic, i think we lack the computer
>science needed to *really* manage networks, see tim griffin and my
>broadside at <http://psg.com/~randy/020301.core-studies.ps>.

I'm not talking about quick fixes, but I do think the problem
can be modularized (e.g., protocol vs. data model).  As a first
step, why not replace proprietary screen-scraping CLI scripts
with a protocol providing a better programmatic API?


>otoh, i do see some intital steps.  but i seem to lack the ability
>to communicate them to the ietf NM community.  i would feel really
>guilty except for some personal circumstances and that no one else
>seems to be able to communicate much of a vision either.
>
>talk is cheap, send vision.

Some of us are working on just that, but the I-D isn't done yet.


>randy

Andy


--
to unsubscribe send a message to xmlconf-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/xmlconf/>