This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] Enforce GMP/MPFR version requires and fix fortran/28276,27021



On Sep 25, 2006, at 12:33 PM, Diego Novillo wrote:
Steve Kargl wrote on 09/25/06 14:46:
On Mon, Sep 25, 2006 at 02:21:39PM -0400, Daniel Berlin wrote:
On 9/25/06, Steve Kargl <sgk@troutmask.apl.washington.edu> wrote:
On Mon, Sep 25, 2006 at 08:06:21AM -0400, Diego Novillo wrote:
Steve Kargl wrote on 09/25/06 00:36:

If no one reviews the toplevel changes within the next few
days, I will commit the changes under the implicitly approved
rule.

The WHAT rule? No such thing exists.
How many other patches have simply whithered and died
while waiting for review?
If you can't get a review after pinging the maintainer, the proper
thing to do is to bring it up to the SC.
They would probably then appoint more maintainers.
Patch review is not a problem solved by simply not doing it.

The "implicitly approved rule" has two possible outcomes if an unreviewed patch is applied.

  1) The patch is applied.  It works as tested by the
  developer and as expected.  Life goes on.

2) The patch is applied. The bootstrap or the compiler is broken
for someone. That person can request the patch be reverted. At
which point, everyone will scream about the lack of a review and the
developer can show via the gcc-patches@ archive that the patch was
available for review for the previous XXX weeks but no one cared
enough to comment.


3) The patch is applied. It works, but it's ill-designed and merely
papers over the problem or introduces hidden side-effects on other parts
of the code. The maintainer for the code spends several days trying to
figure out changes he wasn't aware of and needs to re-write your patch.

4) Over time, reviewers get tired of running into #2 and #3, and start doing reviews in a more timely fashion.


If this actually happened it would be a big win for this approach.
I don't really believe it, though.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]