svn commit: r194210 - head/sys/sys
sam at freebsd.org
Mon Jun 15 14:39:16 UTC 2009
M. Warner Losh wrote:
> In message: <1245014030.12125.34.camel at neo.cse.buffalo.edu>
> Ken Smith <kensmith at cse.Buffalo.EDU> writes:
> : On Sun, 2009-06-14 at 19:53 +0000, Simon L. Nielsen wrote:
> : > Author: simon
> : > Date: Sun Jun 14 19:53:52 2009
> : > New Revision: 194210
> : > URL: http://svn.freebsd.org/changeset/base/194210
> : >
> : > Log:
> : > Bump __FreeBSD_version for OpenSSL 0.9.8k import.
> : >
> : > Modified:
> : > head/sys/sys/param.h
> : >
> : > Modified: head/sys/sys/param.h
> : > ==============================================================================
> : > --- head/sys/sys/param.h Sun Jun 14 19:51:43 2009 (r194209)
> : > +++ head/sys/sys/param.h Sun Jun 14 19:53:52 2009 (r194210)
> : > @@ -57,7 +57,7 @@
> : > * is created, otherwise 1.
> : > */
> : > #undef __FreeBSD_version
> : > -#define __FreeBSD_version 800097 /* Master, propagated to newvers */
> : > +#define __FreeBSD_version 800098 /* Master, propagated to newvers */
> : >
> : > #ifndef LOCORE
> : > #include <sys/types.h>
> : It's been suggested we change the heuristic to be stable branches start
> : with 200. I think we're going to do that so don't panic about there
> : only being one number left for changes between now and the time we
> : branch.
> Why not 500?
> : At first it surprised me we ate through almost 100 version numbers in
> : head, which is a place traditionally abi changes are *expected* and the
> : attitude has more or less been head is a free-for-all with all bets off.
> : But some IRC chatter enlightened me to there actually being people who
> : do advanced testing of ports and whatnot in head under the encouragement
> : of portmgr@ despite its free-for-all nature to try and cut down on work
> : during later stages of the release. That seems reasonable so I think in
> : general starting to give more than 100 bumps to work with in head is
> : worthwhile.
> I think so too. If we're going to change, we might as well change big
> and infrequently.
netbsd reserves 4 digits. I've never bought the idea that people can
ignore ABI changes (even in HEAD). But not recording an incompatibility
is just stupid.
More information about the svn-src-head