> Perhaps we could guarantee that the shuffle results in a different
> order? I think that would resolve any objection to making shuffling the
> default.
.. at the cost of a rather bizarre "shuffle … but won't produce the original"
behaviour and code :)
Anyone objection to simply printing what we are doing…? I'd be very happy with
that and would also benefit anyone looking over build logs, etc. and regardless
of whatever shuffling or sorting was in play.
Regards,
--
,''`.
: :' : Chris Lamb
`. `'` la...@debian.org / chris-lamb.co.uk
`-
