Update?
|
Or, if you want to wait 24 hours, I'll have fixed the db packages to not
require java and make the java bindings a completely separate build
(then pure darwin users can build bdb too, and it will not require most
users to get the java sdk).
Peter
- --
Peter O'Gorman - http://www.pogma.com
don't believe that we have come near the point of
running out of space yet.
So, forget about it :)
Peter
- --
Peter O'Gorman - http://www.pogma.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.3 (Darwin)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQE
with this system, and fink would take care of the rest?
I'd like to see/try the fink patch :)
Peter
- --
Peter O'Gorman - http://www.pogma.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.3 (Darwin)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQE
. It seems likely that the output of the prebinding
tool is being set to the install_name. Especially as diff/cmp tell me
that the 2 files differ.
Ben Reed has done all the fink-prebinding work, so we'll have to ask him
when he gets up.
Peter
- --
Peter O'Gorman - http://www.pogma.com
-BEGIN PGP
- --
Peter O'Gorman - http://www.pogma.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.3 (Darwin)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQE/wLG+hJEqt987YAQRAsxAAJ96A7EjeTaNzDPUNQsuzTKq+sahJwCeM2M0
IlrjT3S4oz8Luj8idsSo22U=
=H/6Z
-END PGP SIGNATURE
and ask users to set DYLD_IMAGE_SUFFIX to get a decent backtrace.
Anyway, I'm full of bright ideas, but no numbers, so I'll go get some
numbers later today.
Thanks,
Peter
- --
Peter O'Gorman - http://www.pogma.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.3 (Darwin)
Comment: Using GnuPG
.
Peter
- --
Peter O'Gorman - http://www.pogma.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.3 (Darwin)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQE/v/mthJEqt987YAQRAnCWAKCp3Xj3GlTz1WUtIUdtcU+fh8TCXgCfatkr
mfcJ+W2dCkDTmt46IDnp80w=
=4N1/
-END PGP SIGNATURE
at compile time so this:
if( ... ) {
use Foo;
}
is really this:
use Foo;
if( ... ) {
}
This means POSIX was always being loaded and $basepath/lib/perl5 always
being added to @INC.
Thanks for this.
Peter
--
Peter O'Gorman - http://www.pogma.com
Benjamin Reed wrote:
Uninstalling the original version in 5.6.0 is against Fink policy
though. Shouldn't it still be finding ours first if we're setting
PERL5LIB?
I don't know the exact problem here, but from a clean bootstrap we do
not set PERL5LIB. The failures weren't from bootstrap
this?
Peter
- --
Peter O'Gorman - http://www.pogma.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.3 (Darwin)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQE/u4PBhJEqt987YAQRAgFkAJ9PWOfaiukscqHsWqaq+ycr/FG6wACgg9VW
sCZb+IU6+1TnsCDz9qZl2Dk=
=8eU0
-END PGP SIGNATURE
TheSin wrote:
further more, on the fink site!
http://fink.sourceforge.net/doc/porting/libtool.php#dylibversionfix
Can you point me at the libtool docs/code which say -release is
deprecated please?
Just because it is on the fink web site doesn't necessarily mean it is
correct :-p
Peter
- --
Peter O'Gorman - http://www.pogma.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.3 (Darwin)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQE/uhwxhJEqt987YAQRAkVqAKCmHc0Ne1VgUgP/zlz4/41oOqqUewCgnzXP
yVofU6SwYvjqftiHowICAyQ=
=tVtx
-END PGP SIGNATURE
And when you run fink list system-xfree86 after that it should properly
detect the one you've got.
Peter
- --
Peter O'Gorman - http://www.pogma.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.3 (Darwin)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQE
may or may not break backward/forward binary compatibility
by arbitrarily changing packages like this.
Peter
--
Peter O'Gorman - http://www.pogma.com
---
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you
, except his rename(2) call failed wile mine worked.
I'm pretty confused too.
Peter
--
Peter O'Gorman - http://www.pogma.com
---
This SF.Net email sponsored by: ApacheCon 2003,
16-19 November in Las Vegas. Learn firsthand the latest
developments
Peter O'Gorman wrote:
I could make a hand-written .la and put it into the 10.3 freetype2 with
a libfreetype.dylib symlink to /usr/X11R6/lib/libfreetype.6.dylib that
would likely solve the problem actually. Hmm. I'll have to check into
that this weekend.
On second thought, that won't work either
in dependency_libs in the .la files).
Peter
--
Peter O'Gorman - http://www.pogma.com
---
This SF.Net email sponsored by: ApacheCon 2003,
16-19 November in Las Vegas. Learn firsthand the latest
developments in Apache, PHP, Perl, XML, Java, MySQL
or the stable
version changed.
Thanks,
Peter
--
Peter O'Gorman - http://www.pogma.com
---
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE
D. Höhn wrote:
If you would like to continue this important work, I would be more than
happy to setup a roadmap of your goals and sync them with the Fink team,
being your proxy. As Mister Reed said, we are very interested in
additional patches to further improve fink and I am sure that you
Koen van der Drift wrote:
Hi,
Two of the programs of the emboss package do not work on Mac OS X
apparently because of the limited stacksize on Mac OS X. For instance,
if I type just 'cirdna' at the prompt, I get an segmentation fault.
However if I type 'unlimit stack' first, the program works
, oddly enough). Mea culpa-- I probably should have looked
at the source code before trying selfupdate-rsync :-/
We simply never considered this possibility, sorry. I'll look at
changing it to leave non-fink distribution CVS dirs alone when I get a
chance.
Peter
--
Peter O'Gorman - http
David R. Morrison wrote:
The simplest thing to do is to simply add this as an alternative method.
But we were just discussing on #fink a more drastic change: we could
scrap anonymous cvs access altogether, and offer only two access methods,
rsync or ext cvs (the latter for fink developers only).
On Monday, September 8, 2003, at 9:46 PM, Martin Costabel wrote:
push(@provides, 'fontconfig1')if
(readlink('/usr/X11R6/lib/libfontconfig.dylib') =~ /libfontconfig\.1/
and
There is no libfontconfig in this version of X11.
Oops, might want to check for the file existing before doing a
On Monday, September 1, 2003, at 11:02 PM, jfm wrote:
I get _ with 'md5sum -c' :
Thank you,
Peter
---
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
___
Hi,
Okay, except for the checksums which failed to confirm for
Jean-Francois, I have added a whole bunch to unstable, some of them
unconfirmed, these may or may not fail, who knows, give me more
confirmations please :)
The following packages still have no checksums in unstable:
Chris
Hi,
The following source files I just fetched/refeched. These sources (some
of which went through SourceRename) are from packages in the unstable
tree which do not have a SourceN-Md5 field for the associated SourceN
field.
I am hoping for confirmation that these md5's are valid from people who
Hi,
I'm considering changing fink so that when the user does fink install
foo and foo has no Source-MD5 field, or the field is empty, it will
print the warning asking the user what to do (similar to the warning
currently printed when the md5 in the info file and the tarball differ).
Are we
On Friday, August 29, 2003, at 11:41 PM, David R. Morrison wrote:
Hi Peter. This one should be structured as a warning to developers,
not
to users (similar to the duplicate fields warning). In fact, there
is
already a warning if you do fink fetch foo and foo has no MD5-sum.
Okay, thought the
On Thursday, August 28, 2003, at 10:31 PM, Benjamin Reed wrote:
Peter O'Gorman wrote:
I don't see the need for this to be in a Recommends field, would it
not be better in a Desc* type field, or simply as a #comment in the
.info file?
It's not a matter of need, but it does make sense to me
On Wednesday, August 20, 2003, at 08:30 PM, Daniel Macks wrote:
'autoreconf -f' (from autoconf2.5 (2.57-1)). I see no warnings or
errors
during ./configure or make, until:
/bin/sh ../libtool --mode=link gcc -g -O2 -g -L/sw/lib -o gael
-export-dynamic main.o -L/sw/lib -lgnomevfs-2
gcc -g
On Tuesday, August 19, 2003, at 05:58 AM, Joe Block wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
I am creating a package for python23 docs, and have run into the
problem that the tar file is named html-2.3.tar.bz, but expands into a
directory tree named Python-Docs-2.3. It was easy
On Saturday, August 2, 2003, at 06:36 AM, Martin Costabel wrote:
Peter O'Gorman wrote:
I'd guess that there is a static constructor somewhere in one of the
other objects which gets called before main() and does something odd
You had the right idea: As it turns out, the crash does not happen
On Monday, June 30, 2003, at 01:23 AM, David R. Morrison wrote:
I'm still a bit unclear about what your change would accomplish, Peter.
BuildDepends is not passed to dpkg. We use that only within Fink.
As you suggest, when the shared libraries plan is fully implemented,
we'll
only have to
On Tuesday, July 1, 2003, at 03:07 AM, TheSin wrote:
this is indeed not true, I just tried it.
On Monday, June 30, 2003, at 11:18 AM, David R. Morrison wrote:
If you do fink build foo then indeed foo-shlibs gets installed.
In that case, please feel free to beat me around the head with a wet
On Sunday, June 29, 2003, at 03:18 PM, Max Horn wrote:
As for 2, we will need to otool everything we are putting into the
10.3 tree to check the depends, seems fair enough, I'd also like to
suggest changing the BuildDepends lines and fink so it does not add
Depends at build time. Everything
On Sunday, June 29, 2003, at 05:14 AM, Max Horn wrote:
--- NEW FILE: IDEAS ---
- decide on setting MACOSX_DEPLOYMENT_TARGET to 10.3 or not
- all pkgs that need a depend on essential pkgs should list that depend
- all pkgs that depend on ncurses will need a versioned dep as this
version
is bin
On Wednesday, June 25, 2003, at 09:10 AM, Thomas Kotzian wrote:
linc1 can't find a file during install phase.
the maintainer doesn't respond after a few weeks so i write to the
mailing-list.
here's the output:
Thomas,
Update your opensp3 and openjade packages, it should then work.
Peter
I am working on some gnu libtool notes, here is a first and very rough
draft.
Peter
Problem:
libtool always links against the installed library, not the one it just
built.
Solution:
On some older versions of libtool the var hardcode_direct is set to
yes. If you patch configure to change this
On Thursday, May 29, 2003, at 09:21 PM, David R. Morrison wrote:
What is the point of LibTool in Fink, Mac OS X include version 1.4.2
from glibtool --version:
ltmain.sh (GNU libtool) 1.4.2 (1.4 2001/11/19 00:06:02)
Sure, and now libtool is up to 1.5 . The newer versions have a lot of
Darwin/Mac
David,
You do not show the link command in this mail, but I am betting a
couple of libtool patches fromt he porting page would make them go away.
Looks like the convenience library bug to me :-)
Peter
On Sunday, April 6, 2003, at 11:27 PM, David wrote:
ld: multiple definitions of symbol
On Thursday, April 3, 2003, at 04:29 PM, Rohan Lloyd wrote:
It works pretty well for emacs, What do you think?
Although I quite like this idea, what about compiled C python modules,
the binaries are not compatible across python versions?
Peter
On Thursday, April 3, 2003, at 10:57 PM, Jeff Whitaker wrote:
So, I'll create numeric-py22 and pil-py22
packages ASAP and put them in experimental/jswhit.
If you were here, I'd kiss you :-)
Thanks, if you do not wish to maintain those packages, put my name on
them.
Peter
Hi,
Depending upon which version of python is installed, the modules will
go into a different place (/sw/lib/python-VERSION/site-packages).
Now, this is a problem for me, and I think I will have to make
numeric-py22 and pil-py22 packages in order for sketch to work without
having to answer a
On Thursday, April 3, 2003, at 03:09 AM, [EMAIL PROTECTED]
wrote:
I sure hope not. I don't have the time to maintain all those variants.
Yes, I know it is a pain, but I don't see a way around it, at least the
current version of sketch has no hope of working with python-2.3 due to
changes made
Here is the relevant discussion from #fink (edited, times in JST).
Apr 01 13:42:47 Clef anyone object to requiring md5s for all packages?
Apr 01 13:42:49 Clef no? good
Apr 01 14:07:37 pogma Clef: What do you mean requiring?
Apr 01 14:08:32 pogma Clef: If you mean that the package will not
build
Hi,
I just built gtk+2-2.0.9 with xfree86-4.3 with no errors.
The problem seems to have been that xfree86-4.3 does not have
Xft/XftFreetype.h, but when I grep for this
[pogma:src/gtk+2-2.0.9-3/gtk+-2.0.9] peter% grep -rl
Xft\/XftFreetype.h .
./config.log
./configure
./configure.in
Simply
On Saturday, March 29, 2003, at 09:50 PM, Chris Zubrzycki wrote:
If i remember correctly, it does not build the xft backend libs then,
maybe compare the filelisting of the 2 debs? one with xfree 4.2 and
one with the check removed and against xfree 4.3
I have no idea, and am too lazy to check,
On Thursday, March 27, 2003, at 10:43 PM, David R. Morrison wrote:
This is a plea to fink developers to make sure that you include MD5
fields
for all of the sourcefiles your package downloads.
When you run fink validate on your .info file, it will warn you if
the
MD5 field is missing.
Here
On Thursday, March 27, 2003, at 11:21 PM, Peter O'Gorman wrote:
Here is a list of stable packages missing the MD5, there are many
other validate warnings
I missed the crypto tree, here are those too:
Warning: No MD5 checksum specified for source.
(dcmtk-ssl-3.5.1-14.info)
Warning: No MD5
Hi,
From a report on -users, I realized that g-wrap needs to have it's
.la's around so that the modules may be loaded into the guile
interpreter. Currently there are 3 packages g-wrap, g-wrap-dev and
g-wrap-shlibs.
Now I could rewrite the splitoffs and move the .la's into
g-wrap-shlibs, or I
On Monday, March 10, 2003, at 12:03 AM, Benjamin Reed wrote:
sane-backends
Are there any objections to any of these?
Yes. sane-backends includes a whole bunch of scanner drivers which are
either known to not work or not known to work. I don't want to see this
in stable.
Thanks,
Peter
On Monday, March 10, 2003, at 06:16 AM, Ben Hines wrote:
Yes. sane-backends includes a whole bunch of scanner drivers which
are either known to not work or not known to work. I don't want to
see this in stable.
.. so document that. A few parts of sane-backends not working is not
worth holding
Hi,
Well, I made myself a thread safe dlcompat, but that was not the
problem with scigraphica at all. Turns out it was the same problem as
the guy on opendarwin-hackers had, the python modules were built
-twolevel_namespace -bundle_loader.
I did fink rebuild numeric pygtk and it now works. The
Whereever it is getting the -arch i386 it isn't going to work on Mac OS
X as there are no fat libraries or compiler toolchain installed.
When building darwin there is an env var/make option. make RC_ARCHS=ppc
, what is required for this package I do not know, but you'll have to
change it.
Well, that one is easy, either define environ to *_NSGetEnviron() with
a patch in the right source files, or remove the libtool -no-undefined
flag, the former is preferable.
If someone does add a new guile to fink, please let me know in advance
as gnucash will have to be revision upped and
I'd suggest a PostRmScript with an ln -sf to restore the symlink.
Peter
On Friday, February 7, 2003, at 10:59 PM, Jeff Whitaker wrote:
Folks: I have come across a potential problem while putting together
the
xfree86-4.2.99.901 package. Xfree86 4.3 installs a font database for
fontconfig
On Thursday, February 6, 2003, at 04:19 PM, Ben Hines wrote:
btw, i think the bind_at_load two-step stuff is only need for C++. It
shouldn't be in the C part of libtool. pogma?
Yes, I posted to the libtool list recently (last week?) asking why it
is done two step and -bind at load for all
On Wednesday, February 5, 2003, at 07:45 PM, Ben Hines wrote:
I don't know how to make fink packages, but could really need
Scigraphica, http://scigraphica.sourceforge.net.
Is this the same python dlopening crash that some other poster had?
I don't know, does scigraphica expect a
On Tuesday, February 4, 2003, at 12:01 AM, Benjamin Reed wrote:
cc -multiply_defined suppress -prebind blah || cc -flat_namespace
-undefined suppress blah
1. libkdeui's LIBADD is -lkdecore
2. the first half of the link complains that -lqt-mt is indirectly
referenced
3. it builds the library
Your head is broken, I assume you used CPAN or something to
update/install a perl module and you got a /usr/bin/HEAD which
overwrote your original /usr/bin/head.
Get your old head back and all will be fine.
Peter
On Saturday, February 1, 2003, at 11:18 PM, Mark Knipfer wrote:
In Mac OS
On Thursday, January 23, 2003, at 11:59 PM, Christian Schaffner wrote:
-g -O2 -DHAVE_CONFIG_H -DDARWIN -DSIGPROCMASK_SETS_THREAD_MASK
-no-cpp-precomp -I/sw/include -I../include -I../include/arch/unix -c
apr_hash.c touch apr_hash.lo
apr_hash.c:373: conflicting types for `apr_hash_count'
On Thursday, January 16, 2003, at 01:47 AM, Max Horn wrote:
So what do you think? Is this still your stance:
At 23:28 Uhr +0900 15.01.2003, Peter O'Gorman wrote:
You are, of course correct with the compatibility versions, and I
think it is worth getting this upstream even if it breaks
On Wednesday, January 15, 2003, at 11:02 PM, Max Horn wrote:
Maybe one of you can tell me how to fix this properly: I am working on
evolution 1.2.1, and it gets to the build phase just fine (which does
a make install DESTDIR, by the way). But it dies trying to relink
one of its libs because
On Wednesday, January 15, 2003, at 12:53 AM, Max Horn wrote:
# Like Linux, but with the current version available in
# verstring for coding it into the library header
major=.`expr $current - $age`
versuffix=$major.$age.$revision
minor_current=`expr $current + 1`
On Monday, January 13, 2003, at 11:03 PM, Jared wrote:
One more question. I am having difficulty finding good information
about PackageMaker. I can't figure out how to install more than one
file to different locations. For example, my libxml2 package will have
to install some files into
On Tuesday, January 14, 2003, at 08:23 AM, Charles McCathieNevile
wrote:
I have installed sodipodi on a couple of 10.2.3 systems, one running X
installed by fink and the other running it from the apple X11 with
system-free86.
In both cases it (and as far as it needed them all its
In order for kdemultimedia to work I had to modify dlcompat in cvs so
that it binds all libraries fully when bind_at_load is specified (dyld
should be doing that but isn't), the dlcompat which does this is
currently in opendarwin cvs.
Now the problem, although this dlcompat fixes
On Monday, December 16, 2002, at 02:34 PM, Patrick Sodré wrote:
/daemon
/service
There is no close to the here document here...
Put a
in and you should be good to go.
PostInstScript:
if [ x$1 = xinstall ]; then
daemonic enable openafs-client;
fi
PreRmScript:
if [ x$1 = xremove ]; then
On Sunday, December 1, 2002, at 11:23 PM, Randal L. Schwartz wrote:
Ben Er, we already do have a fink cleanup command. :)
No we don't:
Common commmands:
[snip list without cleanup]
This is from fink --help.
If it's not there, it DOESN'T EXIST.
Hmm, look at the header for that
On Wednesday, November 27, 2002, at 07:27 PM, David wrote:
The main reason for having such an account still holds, namely, that I
do not have to turn people away who WANT to donate. I am not going to
actively ask for it.
Collecting money from people on the basis that it might come in
On Sunday, November 17, 2002, at 10:16 PM, Carsten Klapp wrote:
I received positive reports for a few of my unstable packages. How
many positive reports are generally needed before making a binary is
considered?
file 3.39-1, qtplay 1.1-1 Willy Traver [EMAIL PROTECTED]
cmatrix
I tried to mail the maintainer [EMAIL PROTECTED] about a bug in his
package (well they both have bugs, one hardcoded /sw in the patch, the
other has missing source), but he obviously never paid for mac.com (I
don't blame him for this, I never paid either).
I am going to update one of them,
On Monday, November 4, 2002, at 01:11 PM, Ben Hines wrote:
Bingo. definitely a gcc bug, already filed:
http://gcc.gnu.org/cgi-bin/gnatsweb.pl and search for 7327.
My apologies for posting this and promptly going to bed. Thanks
Jean-Francois, Max and Ben for looking into it.
It kind of
Hi,
Well, I haven't investigated why, but, with -isystem, I have some
problems building pstoedit, I updated the revision in cvs so it works
with current fink-cvs, but it will be troublesome for other packages
also presumably...
With -isystem /sw/include :
configure:7476: g++ -c -g -O2
On Friday, October 25, 2002, at 01:17 AM, Max Horn wrote:
What do you people think? My suggestion would be to try that ASAP. I
could disable the default values of CPPFLAGS/LDFLAGS and instead give
default values to CPATH/LIBRARY_PATH. That's 5 minutes to change it
(and also 5 minutes to
On Monday, October 7, 2002, at 08:37 PM, Max Horn wrote:
Personally I cannot judge whether that is feasible or if it will cause
problems. The nice thing about it of course would be that e.g. libxml2
wouldn't anymore require people to build python or python-nox. But
maybe it has other
On Monday, October 7, 2002, at 06:07 PM, Sébastien Pierre wrote:
Le lundi, 7 oct 2002, à 07:41 Europe/Paris, Jeremy Erwin a écrit :
Many fink packages appear to be compiled with -O2 -g. While this
may be appropriate for debugging, the -g flag can increase the size
of binaries. A more
On Monday, September 30, 2002, at 12:40 AM, Martin Costabel wrote:
Why does db41-shlibs conflict with db4-shlibs? It wouldn't need a new
name if it did. And if it replaces db4-shlibs, gift needs new
dependencies.
I thought this error was fixed for gift in cvs. (and indeed when I look
Hi,
I would like to know if anyone has seen configure checking for
dladdr, if you have, could you please e-mail me the package, so
that I can check my implementation of that function.
In case you are interested you can check out the code at:
cvs -d :pserver:[EMAIL PROTECTED]:/cvs/od
You'll need to use strtoq instead.
You should just be ably to -Dstrtoll=strtoq on the cc line, but
that is ugly.
Peter
On Sunday, May 19, 2002, at 02:56 AM, Thomas Kotzian wrote:
i wanted to port graphviz - there's only 1 problem left (i hope so)
when linking, i cannot find the symbol:
On Wednesday, April 24, 2002, at 10:28 PM, Max Horn wrote:
* dangerous problem in passwd
* broken libmpeg
* broken r-base
* source bootstraping seems not possible w/o trick (trick being
that you download the debianutil source manually, and during
bootstrap tell Fink in which directory
Well, 1.0 has been released :-)
April 19th 2002 seems to have been the release date.
http://sourceforge.net/project/showfiles.php?group_id=15518release_id=25198
Peter
On Tuesday, April 23, 2002, at 04:12 AM, Jeff Whitaker wrote:
On Mon, 22 Apr 2002, Jin Zhao wrote:
Abiword is close to
On Monday, April 22, 2002, at 07:00 PM, Max Horn wrote:
I also would like to know if anybody here has built run any
of these so I can move:
pan-0.11.3-1
gaim-0.56-3
sdl-1.2.4-1
I have built, installed and minimally tested the latest gaim.
Also many others, is there a quick and easy
On Tuesday, April 16, 2002, at 10:45 PM, Ben Hines wrote:
At 3:34 PM +0200 4/16/02, Max Horn wrote:
Uhm, sorry, you are lagging behind a year or so ... :-)
dlcompat, which also provides these functions, has been
available for ages, from Apple's Darwin CVS. And we did a
complete
On Thursday, April 11, 2002, at 10:42 AM, Chris Zubrzycki wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
not giFTcurs. giFT-gtk. I cannot the gtk client to compile with
the - --no-dlopen option. Without it, it compiles fine, but It
cannot find the symbols in the .so file.
What
You forgot to check out the fink module from cvs :-)
Fink will only update itself for point releases, in between
times you have to cvs update the fink module and run ./inject.pl
On Wednesday, April 3, 2002, at 11:42 PM, Chris Zubrzycki wrote:
I know it's in CVS, but I am confused. I run
On Monday, March 25, 2002, at 08:14 PM, Max Horn wrote:
At 13:35 Uhr +0900 25.03.2002, Peter O'Gorman wrote:
I have one suggestion for this. When you have built everything
with nothing installed, you should build again with everything
installed. Building twice like this will find both
I have one suggestion for this. When you have built everything
with nothing installed, you should build again with everything
installed. Building twice like this will find both missing
Depends and Conflicts...
I think it is a great idea.
Peter
On Monday, March 25, 2002, at 10:20 AM, Max
On Tuesday, March 19, 2002, at 11:32 PM, Bertrand Lupart wrote:
Correct me if i'm wrong, but if a user has a foo version of a package
that has obsolete download url, a fink update won't bring him back the
correct download url since the version hasn't been bump versionned.
So he might come
Might want to set it not to do expansion on files with the
.patch extension...
Peter
On Tuesday, March 12, 2002, at 09:27 PM, Max Horn wrote:
It is not confusing CVS, CVS is simply doing its job: key
word expansion. We use this for the documentation, for example.
On Wednesday, March 13, 2002, at 09:34 AM, Max Horn wrote:
Peter, the code should be changed to *not* read in the package
DB when fink list --help is issued =) Or we need that caching
code, it's no fun to wait 10-20 secs :(
I vote for the caching code, Finlay.. :)
Okay, I'll look at it
And why are you sending this here? Fink has no control over
which shell Apple decides to ship, although they will eventually
move to using bash as /bin/sh, probably not until 10.2 or even
later...
Peter
On Monday, March 4, 2002, at 11:51 PM, Eric Norum wrote:
Apple uses zsh as /bin/sh on
You'll have to play with libtool a little more than that. I
would suggest patching the ltmain.sh and ltconfig which come
with kde.
From what I understand kde wants to load dylibs at runtime
and/or link against loadable modules. Is that correct? In this
case you'll have to change the library
Okay, what exactly do you mean by better feedback?
Peter
On Monday, February 25, 2002, at 11:50 PM, Bill Bumgarner wrote:
http://aspn.activestate.com/ASPN/Mail/Message/pythonmac-sig/1042874
It seems Fink is being used by more and more folks that are not
necessarily familiar with Darwin or
Hi,
Your problems may be a conflict with gtk-doc which Masanori
hasn't noticed yet...
I am sure he'll fix it as and when he gets the time. It is,
after all the unstable tree we are talking about here :)
Am copying him.
Peter
On Tuesday, February 26, 2002, at 12:53 AM, Martin Costabel
I have one small problem with your proposal, it only lets us
make 2 debs from one build.
I am going to put my proposal in an example:
Package: mystuff
Version: 1.2.3
Revision: 1
Source: http://www.example.com/mystuff.tar.gz
Debs: shlibs, docs, bin
DebShlibs: %i/lib/libmylib-1.2.3.dylib
Fair enough, I don't really care how it is set up in the info
file, I just think that limiting the system to 2 debs is
inappropriate.
Peter
On Thursday, February 14, 2002, at 01:04 PM, El JoPe Magnifico wrote:
At this point, it'd be cleaner to do something like the debian/control
file,
doing any of this, please e-mail me your
sources, and I will put them on the submissions tracker.
Thanks,
Peter
On Wednesday, February 13, 2002, at 08:46 AM, Jorge Acereda
Maciá wrote:
On Wednesday, February 13, 2002, at 12:26 AM, Peter O'Gorman wrote:
Please put your dlcompat on package
Jorge Acereda Maciá has written a dlcompat with the ability to
dlopen dylibs.
I have put it on package submissions:
https://sourceforge.net/tracker/index.php?func=detailaid=516734group_id=
17203atid=414256
Please check it out and comment on it. It needs to not break anything :)
Thanks,
Peter
Hi,
I just did:
%cvs co fink
%cd fink
%./inject.pl
fink selfupdate only gets the latest released package manager.
Peter
On Sunday, January 27, 2002, at 11:58 AM, Jeremy Higgs wrote:
Is there a way to do this when using fink selfupdate-cvs? As
far as I can
tell, the changes that were made
301 - 400 of 402 matches
Mail list logo
