This math/fftw3 change is turning into a PITA.
First that thread last month, and now again.
Maybe more people should just use packages...
Clint Olsen wrote:
>I got this message:
>
>===>Verifying install for gfortran43 in
>/usr/ports/lang/gcc43
>Making GCC 4.3.4.20090524 for FreeBSD 6.4
>
--- On Thu, 6/4/09, Leslie Jensen wrote:
> From: Leslie Jensen
> Subject: Re: Trouble upgrading math/fftw3 on 6.4-STABLE
> To: "bf"
> Cc: freebsd-ports@FreeBSD.org, clint.ol...@gmail.com
> Date: Thursday, June 4, 2009, 12:40 AM
> >
> > to /etc/make.co
I'd agree with Thierry. Ideally, you wouldn't want to
even build a piece of software that didn't work right,
let alone install it. But once you admit the
possibility that mistakes may have been made, and
recognize that tests are desirable, then
*conceptually* you want to test a port in
circumsta
At the time of writing the md5 checksums of cups-1.3.8-source.tar.bz2 from the
easysw (California) and funet.fi (Espoo, Fi) master sites agree with the
reference md5 checksum listed at www.cups.org, while the port distinfo and the
md5 checksum of the same file on the rz.tu-bs.de (Braunschweig, D
If you are reasonably confident that the port will work, and that there are
no glaring errors, why not file a PR now, and then set to work checking
things? It will probably take a few days to process the PR anyway, because
there is a backlog, and because committers are cautious about making change
Yes, the gcc ports are used partly because the Fortran compiler was removed
from the base system. (They might have
been used anyway, to take advantage of
the improvements in later versions of
gcc. ) bsd.gcc.mk fixes the specific
compiler used via the value of
the WITH_FORTRAN variable in a port
Ma
--- On Tue, 1/13/09, Glen Barber wrote:
> From: Glen Barber
> Subject: Re: math/R fails to build
> To: "Philipp Ost"
> Cc: bf20...@yahoo.com, po...@freebsd.org
> Date: Tuesday, January 13, 2009, 8:03 PM
> On Tue, Jan 13, 2009 at 5:10 PM, Philipp Ost
> wrote:
> > Hi,
> >
> > math/R fails to bui
Did you read the thread entitled "Perl upgrade question" in which a
method for doing this was discussed, in the past few days?
Alternatively, you could run something like:
#!/bin/sh
for pdir in `pkg_info -aqo`
do
[ -z `make -C /usr/ports/$pdir -V USE_PERL5 -V USE_PERL5_RUN` ] || \
echo "$pdir"
#o
Yes, I understood your question. I don't know of a pkg_info switch that
will do what you want. Although my example was for perl, I was
attempting to show you a more general approach to getting the information
that you need. A moment's thought would show you that you could adapt my
example, and u
--- On Fri, 1/16/09, bf wrote:
...
> example, and use something like:
>
> #!/bin/sh
> PORTSDIR=${PORTSDIR:-/usr/ports}
> if [ "`pkg_info -E $1`" ]
> then
> dir1="`pkg_info -qo $1`"
> for pdir in `pkg_info -aoq`
> do
> pdepends="`
> Hello
> Will Scilab 5 be ported to FreeBSD?
> Thanks
> Otacílio
Yes, I hope to soon, but I am busy at the moment. I'd be interested
in hearing whether users of the existing scilab 4.x port would like the
new version, which is Java-based, to replace the old port -- or just be
added in addi
A simple mistake (repeated) in the port Makefile. Known? -- well, thanks to
your message, it is now:
http://www.freebsd.org/cgi/query-pr.cgi?pr=130965
b.
___
freebsd-ports@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/free
Just a note to those of you who helped to integrate the Gnome, X11, and
Fortran updates in Ports: thank you very much for the work that you have
put in, and for your efforts to find and solve any remaining problems. I'm
looking forward to using the new software, and I'm sure that many others in
t
>Sevan / Venture37 wrote:
>> Hi
>> Whilst attempting to create a port for OpenNMS I've hit a wall, one of
>> the dependencies is net/jicmp, the problem I'm having is when attempting
>> to build OpenNMS, jicmp is not detected as being installed & a
>> installation is always attempted & the proces
I was experimenting with cparser and libfirm, and I decided, as a first
step, to build and install both from their FreeBSD Ports on 8-CURRENT i386
using the system C compiler (CFLAGS=-O2 -fno-strict-aliasing -fstack
protector-all -march=athlon-tbird) , and then use the two to rebuild
libfirm. But w
I don't understand: instead of "dropping support" for 6.x, why don't
you just USE_GCC=4.x for 6.x if the problems are primarily due to
the older compiler?
Regards,
b.
P.S. Aren't you supposed to be on a long vacation? ;)
___
freebs
Alex:
I can understand your frustration. The Xorg update, although it helps
a lot of people, is inevitably going to cause problems for some, because
it is run by so many people in different ways with a wide variety of
hardware. It's comparable in some ways to updating the OS, and despite
the har
>Pedro F. Giffuni wrote:
>> if FreeBSD moves to a GPL3'd toolchain without an extremely compelling
>> >reason then I would consider a move to another OS.
>
>This would be a big loss for the Project, I have no doubt about that! :)
>
>-Maxim
S. Don't antagonize who's willing to do as much work
Is there a reason why lang/tcl86, x11-toolkits/tk86, and their threaded
versions are not included in lang/Makefile and x11-toolkits/Makefile, even
though they've been hooked up in Mk/bsd.tcl.mk? If not, would someone
mind connecting them?
Regards,
b.
__
It seems that the latest changes to the lang/perl5.8 port (resulting
in CVS revision 1.100 of the Makefile, in response to PR ports/131434:
lang/perl5.8: suid fails) have broken the install target.
I had to change lines 145 and 146 from:
BSDPAN_VERSION=${PORTVERSION}_1
BSDPAN_WRKSRC= ${WR
Hang on guys, we're looking into it. The upstream maintainer
confirms that failures in command substitutions are _not_ the
expected behavior of bash 4.x, that they are not being observed on
many other systems, and that in the past similar problems like
this have been connected with problems in bi
--- On Thu, 3/12/09, bf wrote:
> From: bf
> Subject: Re: shells/bash-4.0 port horribly broken
> To: freebsd-ports@FreeBSD.org
> Cc: obr...@freebsd.org
> Date: Thursday, March 12, 2009, 4:48 PM
> Hang on guys, we're looking into it. The upstream
> maintainer
> conf
> By any chance, did you commit a Send-PR for this patch? I will
> try the patch later tonight anyway.
> Thanks!
I've submitted a slightly revised patch in a follow-up to an old
PR:
http://www.FreeBSD.org/cgi/query-pr.cgi?pr=ports/101230
It turns out that some of the parser problems with the
> After upgrading to latest lang/gcc42 I found that there is no gfortran
> any more.
>
> Comparing the Makefiles of the ports lang/gcc42 and lang/gcc43 shows
> that it had been removed. Was this intentional?
>
> I need version 4.2.xx of gfortran for programs like math/R etc. Is it
> possible
--- On Mon, 3/16/09, Rainer Hurling wrote:
> From: Rainer Hurling
> Subject: Re: lang/gcc42 without gfortran
> To: bf20...@yahoo.com
> Cc: ger...@freebsd.org, "Mark Linimon" ,
> freebsd-ports@FreeBSD.org
> Date: Monday, March 16, 2009, 5:48 AM
> Thank you for this information.
>
> As far a
--- On Fri, 3/27/09, bf wrote:
> From: bf
> Subject: Re: devel/icu: doesn't build at 8-CURRENT
> To: po...@freebsd.org
> Cc: r...@freebsd.org, b...@ipt.ru, m...@aldan.algebra.com
> Date: Friday, March 27, 2009, 6:17 PM
> My build on a recent snapshot of
> 8-CU
My build on a recent snapshot of 8-CURRENT succeeds without any
problems. There are different failures reported here, but some of
them appear to be related to the vfs problem found in PRs
kern/132620 and ports/132542, and solved a series of changes
ending in subversion r190145 on 20 Mar. 2009. S
1,1
WITH_SVGALIB=yes WITH_SHLIB=yes WITH_FT_BRIDGE=yes,
and I suspect the errors may be due to recent changes
in either the recently-patched WITH_FT_BRIDGE knob, or
recent changes in asymptote, because I have not
noticed any such probl
When attempting to view a variety of postscript files
with the recent ghostscript-gpl built from the ports
system WITH_SVGALIB=yes WITH_SHLIB=yes
WITH_FT_BRIDGE=yes on 6.2-Stable i386, gv fails with:
While reading gs_fapi.ps:
Error: /undefinedfilename in (FAPIconfig)
Operand stack:
(gs_fapi.ps
--- bf <[EMAIL PROTECTED]> wrote:
> When attempting to view a variety of postscript
> files
> with the recent ghostscript-gpl built from the ports
> system WITH_SVGALIB=yes WITH_SHLIB=yes
> WITH_FT_BRIDGE=yes on 6.2-Stable i386, gv fails
> with:
>
> Whi
I am running FreeBSD 6.1-RELEASE-p11 on an HP with an
Athlon T-Bird processor. Using the
compiler that shipped with this OS (gcc 3.4.4 FreeBSD
20050518), and the
gfortran component of the gcc 4.2.0 (20070110) port, I
tried to build taucs 2.2.4 from the ports
system with the makefile, via portmanag
OpenSSL 0.9.8e seems to cause libwww build failure on
6.2-STABLE. I didn't see a PR on it , so I decided to
give you a heads-up.
A partial build log output:
mkdir .libs
cc -O2 -fno-strict-aliasing -pipe -march=athlon-tbird
-Wall -o .libs/head head.o ../src/.libs/libwwwinit.so
-L/usr/lib ../src/
32 matches
Mail list logo
