Re: addition of TLV to locator ID or locator ID set

["Spencer Dawkins" <spencer@mcsr-labs.org>]

> None of this imply that the policy need to be load dependent, thus I think 
> we can avoid Spencer's concern about oscillation (that are valid concerns 
> any time the path selection is a function of some measured 
> characteristic.)

Good!

And just to be clear, my concern was only about load-dependent policy; I 
thought adding the TLV was a *good* idea, but apparently spent too many 
years in TSV and just focused on one small part of the posting :-)

Thanks,

Spencer