On Wed, 10 Apr 2019, Ladislav Lhotka wrote:
Mahesh Jethanandani <[email protected]> writes:
For the record, I prefer Lada’s solution of having a union that takes
both a number and an identity/enum.
To be concrete, it means, for example (without descriptions and references)
typedef ike-integrity-algorithm-t {
type union {
type uint16;
type enumeration {
enum NONE {
value 0;
}
enum AUTH_HMAC_MD5_96 {
value 1;
}
...
}
}
}
The unassigned and private use values won't be present as enums but can
be used via the numeric value.
This IMO provides the benefits of both approaches and, moreover, the
enum values are documented in place.
But that is a bug, not a feature. The location of the enum values are in
the IANA registries which continiously are updated. Putting snapshots of
those in RFCs is a bad idea.
Paul
_______________________________________________
I2nsf mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/i2nsf
