FCIP ConCall minutes 3/27/02

["Andy Helland" <andyh@lightsand.com>]

Following are the minutes of the FCIP Conference Call 3/27/02

--------------------------------------

LightSand's turn as the call host.

Participants included:

Ralph Weber, Jim Nelson, Bill Krieg, Murali Rajagopal, Elizabeth Rodriguez,
Bob Snively, Don Fraser, Brian Forbes, Raj Bhagwat, David Peterson, Venkat
Rangan, Neil Wanamaker, Ken Hirata, and Andy Helland (designated scribe).

Primary topic of discussion was comment resolution for the 136(!) comments
received as result of "last call".  Conventional wisdom says that we will
need a "last, last call" after we get done incorporating these comments.

First, Kudos to Ralph for his ever so thorough, professional and timely
summary of the comments received.  (oops, the scribe waxes editorial).
Anyway, thanks Ralph.

Rather than review all of the comments individually, we focused on the
problems areas that Ralph had been having that he believed required author
consensus/discussion.

Comment numbers below refer to <draft-ietf-ips-fcip-wglc-00.txt>

Item 1 (Comment 34)
Team agreed that correct wording should be "Protocols and version number
fields MUST be checked against expected values and 1's complement of
Protocols and Version number SHOULD be checked against expected fields"

Item 2 (Comment 38)
Team agreed that neither "MUST implement" nor "MUST implement and use" were
appropriate for the discussion of FC Frame transit time.  Instead, we will
remove "implement" and punt this to FC-BB/FC-BB-2.

Item 3 (Comment 47)
ACCEPT editorial comment to include SHOULD or RECOMMENDED.
REJECT rest of the comment.  We think it's OK to establish links from both
sides.  Refer to the answer for Comment 124.

Item 4 (Comment 53)
General consensus is that this is not a security hole per se but rather a
security policy issue.  Comment REJECTED.  Text remains as original.

Item 5 (Comment 57)
Refer to resolution of Comment 124.

Item 6 (Comment 60)
"SHALL" should be "should"  [How much should could a should-chuck chuck if a
should-chuck SHALL chuck should?  ;-) ]
The context is "TCP_NODELAY option *should* be set to one.
Comment was ACCEPTED.

Item 7 (Comment 81)
Normative vs informative references.
AGREED that normative references [11, 21, 22, 23, and 24] should all be
informative references.  Also, we need to add an informative reference to
MPLS.  Action Andy.  << Ralph:  Please add RFC 3031 after MPLS in Section
11. -Andy>>

Item 8 (Comment 82)
AGREED to add David Black and Mallikarjun to the acknowledgements list.

Item 9 (Comment 71)
Tentative agreement to rephrase as "when pre-shared keys are used, IKE
agressive MUST be used."  Venkat will take this issue off-line.

Item 10 (Comment 103)
Change terminology to read "FC/FCIP Entity pair".  Ralph will rototill
document to make global changes.

Item 11 (Comment 123)
REJECTED.  Rationale is to leave open our options for future use of the
Special Frame (SF).

Item 12 (Comment 128)
REJECTED.  New wording to be something like "reject the TCP connect request
using appropriate TCP means".  This allows us to not be too specific about
the mechanisms of TCP inside this document.

Item 13 (Comment 80)
IP QoS and the DSCP bits.  Action Andy:  Take this offline with David and
bring back answer (and commentary).

Next ConCall to be hosted by Venkat/Rhapsody on 4/6/02.  Details to follow
from Venkat.

Constructive comments may be sent by return mail.  Non constructive comments
should be dropped off at http://www.shibumi.org/eoti.htm     :-)

-Andy