Hi,
I don't see a lot of support for Lou's suggestion of having more than 16
bits for the path key.
Lou points out that standards...
shouln't stop you from doing something intelligent/creative
While the idea would certainly leave future implementations plenty of room
to do clever things, we haven't had any indication that there is a pressing
requirement for this change.
Lou wrote...
a 32 bit key would allow for
potential future cases that we can already envision as well as
today's known cases. (BTW I know of one experimental usage of PCE
that I suspect will want this immediately.)
But this leaves us guessing, and without some further hints I don't think we
can make a change based on this level of evidence.
In conclusion, Lou also said...
On the other hand, if changing this now also impacts
implementations, I think we should only support longer keys once
there is real (not today's theoretical) demand for them.
Since I have been told of successful interop test of the current definition,
I suggest we leave things as they are.
If 32 bits (or more!) become a requirement in the future, we can develop a
new subobject.
Cheers,
Adrian
_______________________________________________
Pce mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/pce
