Date: Wed, 18 Jul 2012 02:44:12 +0300 From: Alex Kozlov <spam@rm-rf.kiev.ua> To: Doug Barton <dougb@dougbarton.us>, Niclas Zeising <zeising@freebsd.org> Cc: svn-ports-head@freebsd.org, svn-ports-all@freebsd.org, ports-committers@freebsd.org Subject: Re: svn commit: r301039 - head/archivers/zlib Message-ID: <20120717234412.GA69885@ravenloft.kiev.ua> In-Reply-To: <5005F1E3.7080606@dougbarton.us> References: <201207172057.q6HKvEJc094784@svn.freebsd.org> <5005DB51.7050809@dougbarton.us> <5005E5C4.2030809@freebsd.org> <20120717230739.GA66332@ravenloft.kiev.ua> <5005F1E3.7080606@dougbarton.us>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Jul 17, 2012 at 04:14:43PM -0700, Doug Barton wrote: > On 07/17/2012 04:07 PM, Alex Kozlov wrote: > > On Wed, Jul 18, 2012 at 12:23:00AM +0200, Niclas Zeising wrote: > >> On 07/17/12 23:38, Doug Barton wrote: > >>> Was this causing you some sort of specific problem? AFAIK this happens > >>> often, possibly as a result of 'make makesum'. "Fixing" this everyplace > >>> that it occurs would almost certainly cause a lot of needless churn. > > > > Actually, it's pretty rare occurrence, > > Maybe I'm just lucky then, because I see it all the time. There were 7 cases in the last 3 months plus 2 pending: ports/audio/pianobar/distinfo: missing SIZE for pianobar-2012.06.24.tar.bz2 ports/mail/meta1/distinfo: missing SIZE for meta1-1.0.Alpha11.0.tar.gz > > so it does not add much churn. > > Also while harmless, this is still an error, like unused variable or typo. > No, it's functionally quite different from either of those, since lack How it's different from e.g. unused variable? > of a newline produces no actual problems. It works because sh assume EOF == EOL, it's would fall in more strict parser. > > I agree that ideally this type of change should be committed with > > other changes, but it's hard to achieve because usually successive commits > > are made by different committers. > I'll let others comment as well, but ideally this particular change > wouldn't be committed at all unless that particular file was being > changed otherwise. > > > Would be even better to find out why it happens at all, but i've done > > thousands of tests and still can't reproduce it with make makesum. > I agree that finding out why it happens would be interesting, but we've > got bigger fish to fry. -- Alex
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20120717234412.GA69885>