|
This issue relates to the question of whether existing
RADIUS RFCs have defined the unsigned integer64 data type. The contention is that attributes
such as ARAP-Challenge-Response (RFC 2869) were in fact defined as 8 octet String attributes, not unsigned
integer64. Similarly, the contention is that the Interface-ID attribute was also not defined as an unsigned
integer64. While the bar for making changes to the Design Guidelines
document is very high, this issue is one which could potentially result in an Errata being filed and
approved. A proposed resolution to this issue is available below: http://ops.ietf.org/lists/radiusext/2009/msg00546.html The resolution of this question depends in part on an
understanding of the type of the Interface-ID attribute. Potential candidates
(integer64 seems to have been disqualified) include 8 octet String or Interface ID (variant of 8 octet String). Another sub-question that needs to be resolved is whether the
implementation advice given in the document (for RADIUS implementations to support the unsigned
integer 64 data type) is justified if it is determined that this data type is not in fact used
by any existing RADIUS attribute. If you have comments on the proposed resolution nor any of
the sub-questions, please post your opinions to the list. |