Re: [Isis-wg] [rbridge] Why is MTU discovery important?

[Hannes Gredler <hannes at juniper.net>]

Robert Raszuk wrote:
> It seems to me just like section 4.2.4.3 Hello Contents of the above 
> draft says there is sufficiently large number of optional information 
> usable for TRILL which could be added to today's ISIS header that 
> perhaps it may make sense to just define it as separate TRILL-Hello msg.
> My point was that in the same time the padding issue would be solved yet 
> a lot of already available code for L3 ISIS could be reused.

so we have now separate L2-encaps, separate IIH, separate LSP and separate SNP PDUs
- all of that with changed semantics ...

sounds to me much more like a new protocol, than something vendors can integrate
in their existing IS-IS implementation.

if we want to further walk down that road, we should name the beast TRILL-something
and drop the isis-wg groups involvement, since this is not IS-IS any more.

/hannes