cvs commit: src/contrib/ipfilter HISTORY Makefile bpf_filter.c ip_fil.c ipf.h ipmon.h ipt.h kmem.h mlfk_rule.c opts.h radix.c radix_ipf.h snoop.h src/contrib/ipfilter/BSD Makefile Makefile.ipsend kupgrade src/contrib/ipfilter/iplang iplang_l.l ...

Scott Long scottl at
Thu Aug 24 21:20:48 UTC 2006

Guido van Rooij wrote:
> On Thu, Aug 24, 2006 at 03:22:42PM -0400, John Baldwin wrote:
>>>>>>> Log:
>>>>>>> MFC: ipfilter 4.1.13
>>>>>>This begs for a question: if ipfilter can not be updated without
>>>>>>breaking the tree anymore, does it belong there in the first place?
>>>>>We could of course also get rid of the tree ;)
>>>>Tree without ipfilter or ipfilter without tree - this is a
>>>>Seriously though, I wonder how ipfilter is being tested before being
>>>>checked in? Past two commits by you suggested that no testing was done
>>>>at all. 
>>>Missing an added file when doing a check-in is a forgivable mistake, 
>>>especially since Guido isn't the author and likely doesn't have as deep
>>>of an understanding of the changes as the author does.  And, I do highly
>>>appreciate that Guido is responding quickly to the problems.  The 
>>>process with ipfilter still isn't perfect, but it's definitely an
>>>improvement over the past.
>>Yes, this is the first import Guido has done, and I think it has gone well 
>>relatively speaking.  Overall I think ipfilter imports will become less 
>>eventful in the future than they have been in the past.
> Thanks (also to Scott). It wasn't my first though :-(
> To all others: it was just a stupid mistake and I already apologized.
> -Guido

The key is to start with low expectations.  Now that you've made the 
mistake of doing a good job, expectations will raise accordingly =-)


More information about the cvs-src mailing list