Ira,
First off, it would have been nice to toss each of the 16 issues
on the mailing list as they were identified, rather than waiting
until the very last minute and throwing the whole lot onto the list
as people are getting ready to fly to the PWG meetings next week.
I don't think that's too much to ask, do you? Moreover, I *expressly*
made such a comment to Tom during the JMP meeting in San Diego last
month, strongly encouraging him to not repeat this same scenario as
has been done so often in the past.
And, by posting issues as they arise to the mailing list, you not
only allow for interleaving of human effort, but you also satisfy
the IETF working group rules you have referenced.
Second, I would have preferred a MINIMALLY USEFUL IMPLEMENTATION
out in the marketplace over a year ago. This certainly would have
been possible if some folks didn't constantly insist on tossing in
the kitchen sink into the effort. (Believe me, I'm not the only
one who has made this statement over the last two years.)
Of course I didn't mean to offend the Xerox implementors. However,
I think it's high time Xerox takes a few moments and considers the
rest of the JMP group's desire to FINALIZE the FIRST IMPLEMENTATION
of this Job Monitoring MIB.
There's a difference between "Good" and "Good Enough".
There is always Version 2 (and 3 and 4...). If NASA took the same
approach to designing rockets as was done in designing the Job MIB,
then we never would have made it to the moon back in 1969...since NASA
would still be trying to figure out how to design that same rocket
to get to Pluto.
Steve Waldbusser kept telling us (in so many words, and on so many
occasions back in 1993-1994):
Start off small and minimal, with (ideally) NO optional objects
whatsover, to get some real market experience with a wide base
of interoperable products. Then, go back and do Version 2, taking
into account what you learned in Version 1.
This is all I (AND OTHERS) have been asking all this time, to start
off small and get product in the field ASAP.
You make a reference to a "half viable standard". When the dust settles
on the latest round of mapping Job MIB objects to TODAY'S EXISTING
PRINTING SYSTEMS, I'll bet you will find that such systems can barely
handle the majority of defined objects.
The bigger the design, the more work to develop. The bigger the
product, the more likely you will lose on robustness and interoperability.
Please extend my apologies to the Xerox implementors. I'll need to
partition quite a bit of time to review the 16 issues now presented
on the list. Chances are, now that the PWG meetings are upon us (and
we're hosting them), I won't get such time until the JMP meeting next
Friday...if I'm lucky.
...jay
----- Begin Included Message -----
Date: Fri, 20 Jun 1997 15:12:58 PDT
From: imcdonal at eso.mc.xerox.com (Ira Mcdonald x10962)
To: hastings at cp10.es.xerox.com, jkm at underscore.com
Subject: Re: JMP> Comments on JMP V0.82 from Xerox developers
Cc: jmp at pwg.org
Hi Jay,
Would you prefer half a viable standard now to an actually
usable one a little later. You certainly managed to offend
the Xerox SNMP implementors who took the time to review the
latest draft in detail and post their comments (thanks to
Tom's editorial cleanups) BEFORE the PWG meeting.
The IETF strongly urges all sanctioned working groups to
make ALL substantial decisions solely via mailing list.
Face to face meetings may suit some of the PWG members,
but they certainly penalize the distant folks (in other
countries) and those who speak English as a second
language.
How about constructive comment on the input from Xerox
implementors?
Disgruntled,
- Ira McDonald (outside consultant at Xerox)
High North Inc
----- End Included Message -----