Date:        Sat, 13 May 2000 13:21:19 -0700
    From:        Jonathan Stone <jonathan@DSG.Stanford.EDU>
    Message-ID:  <200005132021.NAA21902@champagne.dsg.stanford.edu>

  | Or, to put it another way, who do we collectively think is
  | supposed to test them, and who *is* acutally testing them?

It seems to me that you (and some others) are testing them.
Certainly the sendmail problem got found, and now fixed.

What more can you possibly ask for?

  | Until now I'd assumed that was part and parcel of "IPv6 integration".
  | Is that wrong?  If it *is* wrong, who is actually be testing IPv6
  | changes on V4-only systems?

You are.  And others too I think.

The IPv6 system wouring without IPv6 in the kernel is your desire.
Mine is that the IPv6 code work without IPSEC in the kernel (because
I'm too lazy to fetch the crypto-intl code and install it, let alone
learn how to use it).  Someone else probably wants IPSEC in, but IPv6 out.
Someone else probably wants it all to work on a sparc64.   And most
probably someone else wants ...

There's a limit to how much any single developer can test.  To each of
us with wants, our wants are the most vital factor in the wole thing.
so they get tested when we make use of them - that way, just about everything
is going to be tested, by someone.

kre