Last Call for publication of RIPE-501bis (reply by 20120426)
In the IPv6 working session at the RIPE meeting in Slovenia, we discussed: http://go6.si/docs/Requirements-for-IPv6-in-ICT-equipment-v.9.pdf The meeting participants felt that the document is now ready for publication. Therefore, I would like to give everybody a final chance to bring up major issues that must be resolved before publication. Please comment by Thu 20120426 if you have major issues that require a delay in publication. We will ask the RIPE NCC to publish the document after the Last Call ends if no serious issues are uncovered by that date. The RIPE NCC will do an editorial pass for spelling, grammar and minor issues before it will publish the document. Feel free to contact the working group chairs directly <ipv6-wg-chairs@ripe.net> if you find anything minor that doesn't need working group review and we will make sure that it gets fixed in the final document. Thanks, David, Shane & Marco ---
http://go6.si/docs/Requirements-for-IPv6-in-ICT-equipment-v.9.pdf
really good and comprehensive list. Two remarks, the weight of which to be determined by the authors: 1) I'm missing RFCs 2671 and 3226 ("Do support EDNS0 and do it right") in the list of requirements for mobile nodes. RFC 3596 is present, though. Usually the former are mandatory but I have no idea what special size considerations might have applied here. 2) Since all other devices are required to support EDNS0 with "large enough" payloads, the requirements list for CPEs should probably contain RFC 5625, so the CPEs do not get in the way of all those compliant gear on both sides. -Peter
Peter, On Thursday, 2012-04-26 15:04:40 +0200, Peter Koch <pk@DENIC.DE> wrote:
http://go6.si/docs/Requirements-for-IPv6-in-ICT-equipment-v.9.pdf
really good and comprehensive list. Two remarks, the weight of which to be determined by the authors:
1) I'm missing RFCs 2671 and 3226 ("Do support EDNS0 and do it right") in the list of requirements for mobile nodes. RFC 3596 is present, though. Usually the former are mandatory but I have no idea what special size considerations might have applied here.
2) Since all other devices are required to support EDNS0 with "large enough" payloads, the requirements list for CPEs should probably contain RFC 5625, so the CPEs do not get in the way of all those compliant gear on both sides.
I think we need to publish the document without these changes. This is a case where "the perfect is the enemy of the good"(*). Regarding future versions or an errata list... my own advice would be to leave EDNS0 RFCs out of this document, since it is intended for IPv6 recommendations not as a general "here are good things to have" document. I admit I'm not sure where such a list would reside. :( -- Shane (*) http://en.wikipedia.org/wiki/Perfect_is_the_enemy_of_good
Shane,
1) I'm missing RFCs 2671 and 3226 ("Do support EDNS0 and do it right") in the list of requirements for mobile nodes. RFC 3596 is present, though. Usually the former are mandatory but I have no idea what special size considerations might have applied here.
2) Since all other devices are required to support EDNS0 with "large enough" payloads, the requirements list for CPEs should probably contain RFC 5625, so the CPEs do not get in the way of all those compliant gear on both sides.
I think we need to publish the document without these changes. This is a case where "the perfect is the enemy of the good"(*).
believe it or not, that phrase even exists in German ;-)
Regarding future versions or an errata list... my own advice would be to leave EDNS0 RFCs out of this document, since it is intended for IPv6 recommendations not as a general "here are good things to have" document. I admit I'm not sure where such a list would reside. :(
I think consistency is key here, next time. -Peter
participants (3)
-
David Kessens -
Peter Koch -
Shane Kerr