svn commit: r187132 - head/usr.bin/make
M. Warner Losh
imp at bsdimp.com
Fri Jan 30 10:00:47 PST 2009
In message: <20090130173956.GA9119 at dragon.NUXI.org>
"David O'Brien" <obrien at FreeBSD.org> writes:
: On Fri, Jan 30, 2009 at 08:51:30AM -0700, M. Warner Losh wrote:
: > In message: <20090130015518.GA20404 at hades.panopticon>
: > Dmitry Marakasov <amdmi3 at amdmi3.ru> writes:
: > : * David E. O'Brien (obrien at FreeBSD.org) wrote:
: > : I think this should be backed out. Those verbose messages:
: > : - Are completely unexpected, this violates POLA.
: > : - Do break recognizeable make output people are used to
: > : - Really uglify make output for some custom makefiles (for example,
: > : generated by cmake:
: > : There's golden unix way rule: silence is golden. So please back this
: > : out, as this will really annoy many people.
: > This makes at least two requests... I hate them too. really really
: > really hate them. -Q is default in all my trees.
: > The real problem is that it exposes way too many internal target names
: > that are totally baffling, even to me who has a lot of build
: > experience. Also, it isn't clear how to use them.
: > O'Brien says they were disabled in 1994 for no good reason without
: > discussion, so he's turning them back on, without discussion. The
: > project is a very different place than it was then, and doing this
: > sort of thing is anti-social.
: 100,000 of things change within FreeBSD without discission that
: displeases some set of folks. That's nothing new, but I'm restoring
: compatibility and functionailty, not removing it.
Don't you have any ability to judge that this might be a bad thing?
So far the voting I solicited is 100% in favor of getting rid of what
you did... Usually that implies it is a bad commit...
: I found src/Makefile.inc1 r134903 / rev 1.444 very noisy, but lived with
: it. I don't care that -s now implies -Q, except that it still leaves so
: much "noise" like r134903 / rev 1.444 and other output.
: [Why the log message is about 'Unanimous Consent' and not verbosity?]
Because I messed up on the commit. I also talked to a lot of people
before making the commit, socialized the issue and got buy in from ru@
before making the change.
: I wonder what % build speed improvement quieting that behind 'make -s'
: would give?
I should measure it. I believe that it won't be that much since the
output is 1/10th the size of the latest commit...
More information about the svn-src-head