ports/75155: [PATCH] audio/baudline: fetching does not require linprocfs

Ion-Mihai Tetcu itetcu at people.tecnik93.com
Tue Dec 21 16:20:46 UTC 2004


On Tue, 21 Dec 2004 16:42:29 +0100
Volker Stolz <stolz at i2.informatik.rwth-aachen.de> wrote:

> Am 21. Dec 2004 um 15:19 CET schrieb Ion-Mihai Tetcu:
> > > I used pre-install since I had to spend too much time
> > > thinking about why you'd go for pre-extract :) (including
> > > the roundtrip to pkg-message and then back to IGNORE)
> > 
> > The idea was to not spend time extracting, patching, ... if no linprocfs
> > present and the user could see this form the beginning. 
> 
> If you look closely at the port, you'll notice that virtually
> no time at all is involved since it's a binary port. Actually
> the right thing to do would be having linprocfs as a run-dependency
> which is not possible. Now, we get almost all the regression testing
> for this port the cluster can provide, which unluckily because of
> the above restriction means stopping short of generating a package.
> I think we can all live with the new situation.

Oh, I don't use the port itself; but because of linprocfs my
master_ports_blah_machine [1] was failing to fetch it and used to
require manual intervation. I should have looked closer at what the port
is / does.

[1] - that's a machine that fetches the distfiles for i386 ports by
parsing cvsup logs and it is MASTER_SITE for the rest, plus it builds
the ports we use on the other machines customized to our needs.


-- 
IOnut
Unregistered ;) FreeBSD "user"




More information about the freebsd-ports-bugs mailing list