In message <200005130353.UAA18516@champagne.dsg.stanford.edu>, Jonathan Stone w
rites:
>Its not the *work* I see people complaining about, its the
>breakage done during the work.  Seems to me that IPv6 changes
>weren't being tested without IPv6 until after they were committed,
>at which point other people noticed things had broken.

Testing is another part of the work.

>If so, then *that*, is, as they say, a fair cop.  Let me say that
>Itojun is being very, very proactive about fixing the breakage; but
>its still closing the stable door after the horse has, uh, broken.

As long as no *release* version has this problem, I'm not entirely offended.
This may sound crazy, especially given that my Mom's laptop runs -current
as of Tuesday or so, but I generally *expect* -current to include code that
needs more testing, or partially-completed work.

I think we're all spoiled; we spend a few years using -current and get used
to it being more stable than the major releases of some systems, and we forget
that it *is* an experimental tree.

-s