The IP over ATM Mailing List Archive by date

[Grenville Armitage <gja@bellcore.com>]

	[..]
>>> ipmc-12 hardly says all that much about MCSs anyway, punting
>>> that to a separate document on "What an MCS should or should not expect"
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

>>> (which is what I'd encouraged Rajesh to be developing).
>>>
>>> That's where you tighten up (or explicitly allow) certain MCSs behaviours
>>> (to fill in the gaps I _deliberately_ tried to leave in ipmc-12).
>>
>>The ipmc-12 version that I've got has a bunch of rules and regulations
>>about the MARS<->MCS interface:

Eh? One or two messages ago you where lecturing me on how the vagueness
in ipmc-12's  MARS<->MCS could let people build restricted MCSs. Now
you're claiming the restrictions are in ipmc-12. Its hard to clarify
anything for you Andrew when you change tunes.

>>what I hear you saying is that ipmc-13 
>>will remove these, correct?

No. And if you read what you'd quoted, you wouldn't have got that
impression. ipmc-12 has some not-clear-enough-for-Andrew guidlines
on certain issues (like MCSs) _because_ there's (a) another document
in the wings to cover these issues, and (b) we need implementation
experience to really know what is ultimately useful.

gja