FreeBSD flood of 8 breakage announcements in 3 mins.

Julian H. Stacey jhs at berklix.com
Wed May 15 15:51:20 UTC 2019


Hi, Reference:
> From:		Kurt Jaeger <pi at freebsd.org>
> Date:		Wed, 15 May 2019 17:38:36 +0200

Kurt Jaeger wrote:
> Hi!
> 
> > > > PR headline : "FreeBSD flood of 8 breakage announcements in 3 mins."
> > > > 
> > > > https://lists.freebsd.org/pipermail/freebsd-announce/2019-May/date.html
> > > > 
> > > > Volunteers who contribute actual fixes are very much appreciated;
> > > > But those styled as 'management' who delay announcements to batch floods
> > > > damage us.
> > > 
> > > 8 announcements and one freebsd-update is easier on the
> > > admin and the re-team than 8 announcements and 8 freebsd-update runs.
> > > 
> > > That's probably why they are batched. Because all of the fixes
> > > are bundled in one update.
> > > 
> > > If the re-team-capacity is limited, what would be the alternative?
> 
> > Alternative is to for announcers to do Less work: 
> > Send each announcement when ready.
> 
> The problem is not the announcement, the problem is providing
> the freebsd-update.
> 
> If announcements are send when ready, and the freebsd-update is
> not ready, therefore, the timeframes to attack systems with unpatched
> problems are much longer.

True as far as that goes for binary users, but often source patches
are available faster, which begs the question: when to announce ?
When there's diffs ? When diffs are commited to src/ (used to be the norm *) ?
When there's some binary update ? 
Whne a whole bunch of 8 arrive in 3 minutes ? Gasp !

* I happen to use src/ never use freebsd-update.
Equally bound to be some who use binary updates & not source

Cheers,
Julian
-- 
Julian Stacey, Consultant Systems Engineer, BSD Linux Unix, Munich Aachen Kent
 http://stolenvotes.uk  Brexit ref. stole votes from 700,000 Brits in EU.
 Lies bought; Groups fined; 1.9 M young had no vote, 1.3 M old leavers died.


More information about the freebsd-stable mailing list