[development] Intelligent use of time (Re: Perm for setting RTBC

[andrew morton]

On 7/3/07, Earnie Boyd <earnie at users.sourceforge.net> wrote:
> Perhaps a voting system for RTBC where the reviewer/tester gives a vote
> and X votes moves the status to RTBC.  Once the status RTBC is given it
> is ready for the next release.  If the RTBC happens prior to code
> freeze then that patch should be considered as part of the upcoming
> release.

The problem with this is the vicious re-roll cycle dww pointed to
above. After any big patch hits the odds are high that existing RTBC
patches will need some work and then additional testing. I think that
setting up a formal voting system that requires multiple testers would
just add add overhead without a clear benefit. Knowing what a pain it
is to gather up reviewers for the votes would encourage people to be
less honest that a previously RTBC patch is now PNW or CNR.

andrew