tdl subdirectory created by "libtoolize --ltdl".
Blowing away autom4te.cache does not help.
Is there any more known about this issue? I can provide more details if
necessary.
--
Braden McDaniel e-mail: <[EMAIL PROT
On Sun, 2008-10-05 at 12:21 +0200, Ralf Wildenhues wrote:
> Hello Braden,
>
> thanks for providing feedback!
>
> * Braden McDaniel wrote on Fri, Oct 03, 2008 at 09:17:42AM CEST:
> > On Tue, 2008-01-29 at 07:54 +0100, Ralf Wildenhues wrote:
> > > * Paulo J. Matos wr
On Sun, 2008-10-05 at 19:36 +0200, Ralf Wildenhues wrote:
> * Braden McDaniel wrote on Sun, Oct 05, 2008 at 07:23:40PM CEST:
> > On Sun, 2008-10-05 at 12:21 +0200, Ralf Wildenhues wrote:
> > > * Braden McDaniel wrote on Fri, Oct 03, 2008 at 09:17:42AM CEST:
> >
t
has been working fine. However, the error message will point to it.
Only if I remove the first occurrence does the error message point at
the second invocation (the one mentioned listed above).
I am using Autoconf 2.63. (Specifically, the one in Fedora 10.)
--
Braden McD
On Sat, 2009-03-28 at 07:27 -0600, Eric Blake wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> According to Braden McDaniel on 3/27/2009 11:02 PM:
> > When I try to use AC_LANG_PROGRAM in conjunction with AC_EGREP_CPP, I'm
> > seeing this error:
> >
)
AT_CLEANUP
What causes this? And how do I fix it?
--
Braden McDaniel
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
than your
configure script is likely ever to be.
So do a check if it's easy. But if it adds a bunch of complexity that
will significantly affect the cost of configure.ac's upkeep, don't
bother.
--
Braden McDaniel
___
Autoconf mailing
On Sun, 2009-04-19 at 21:50 -0600, Eric Blake wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> According to Braden McDaniel on 4/19/2009 8:31 PM:
> > I'm seeing an error processing a testsuite.at that was working at some
> > point; that point was probabl
On 8/6/09 4:36 PM, Murray S. Kucherawy wrote:
What is the general wisdom when deciding whether to add support for a new
"--with" vs. just encouraging use of CFLAGS/LDFLAGS on the command line?
In particular I'm talking about openssl here. I have a package that needs it.
Several systems come
It looks like the abs_top_srcdir shell variable is no longer set for
testsuite in Autoconf 2.63. What can I do instead (to refer to a file
in the source tree)?
--
Braden McDaniel
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org
On Mon, 2009-11-23 at 07:33 +0100, Ralf Wildenhues wrote:
> Hello Braden,
>
> * Braden McDaniel wrote on Mon, Nov 23, 2009 at 06:20:16AM CET:
> > It looks like the abs_top_srcdir shell variable is no longer set for
> > testsuite in Autoconf 2.63. What can I do instead (to r
o, the test not to be done.
Why do you want the test to be conditional?
My experience is that, in general, conditional tests add complexity with
very little return. Why not just let the test fail where the requisite
bits are missing and decide if that actually matters later in the
script?
On Fri, 2010-02-12 at 16:34 +0100, Vincent Torri wrote:
>
> On Fri, 12 Feb 2010, Braden McDaniel wrote:
>
> > On Fri, 2010-02-12 at 11:04 +0100, Vincent Torri wrote:
> >> Hey,
> >>
> >> I restart this thread.
> >>
> >> So, what I
rwise
violates the principle of least surprise.
--
Braden McDaniel
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
geably. (This, of course, doesn't
fix all of your POSIX interoperability problems; but it helps.)
--
Braden McDaniel
___
Autoconf mailing list
Autoconf@gnu.org
https://lists.gnu.org/mailman/listinfo/autoconf
r, I think there are some macros in
the macro archive that can help.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
___
Autoconf mailing
Harald Dunkel wrote:
I am new to autoconf and automake, so please excuse if you
heard this too often.
My question is: How can I use AC_CHECK_LIB without setting
a global LDFLAGS?
Yes; but of course, then the compiler's defaults will be used.
Presumably that's not what you want either.
Using autom
Harald Dunkel wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Braden McDaniel wrote:
|
| I don't fully understand the problem you're describing; but perhaps you
| want to do something like this:
|
| save_LDFLAGS="$LDFLAGS"
| LDFLAGS="$LDFLAGS -Lwhatever"
Harald Dunkel wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Braden McDaniel wrote:
| Harald Dunkel wrote:
|
|> -BEGIN PGP SIGNED MESSAGE-
|> Hash: SHA1
|>
|> Braden McDaniel wrote:
|> |
|> | I don't fully understand the problem you're describing; but
lib2...
>
> 2:) how can I do the same for headeer files?
You can look at the implementation of AC_PATH_X and AC_PATH_XTRA. The
machinations there are not pretty. Fortunately, there is usually a
Better Way.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://
know how well it's suited to non-Fink Mac OS X development, but
that is the kind of thing with which pkg-config is designed to assist.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]&g
Paul Eggert wrote:
Gregorio Guidi <[EMAIL PROTECTED]> writes:
[snip]
it is very important when creating a package to know in advance what
features will be enabled in the software at the end of the process,
without knowing the details of the host where the package is
building.
When you say "feature
Harry Mangalam wrote:
Hi All,
salient points:
Linux xxx.xxx.uci.edu 2.6.8.1-4-amd64-k8-smp #1 SMP Fri Jan 14
11:33:56 UTC 2005 x86_64 GNU/Linux
config.log pointed me to a number of erroneous conditions which I've
now corrected. However, I'm still experiencing the same problem.
That is, that
Patrick West wrote:
I am interested in doing something like this:
./configure --with-package=package1 --with-package=package2
But I am not sure how to do this. I try to use AC_ARG_WITH, but that
only takes package2 and ignores package1.
Any help would be greatly appreciated.
A single switch (in y
On Tue, 2005-03-15 at 16:25 -0500, Dan Manthey wrote:
>
>
> On Tue, 15 Mar 2005, Braden McDaniel wrote:
>
> > While you can use any value you want as the argument to a --with-*
> > option, anything other than "yes" or "no" usually complicates the UI a
which the specfied packages
> are _not_ mutually exclusive, so that doesn't pertain.
Then what, dependent? That is, if you use implementation X of A, you
must use implementation Y of B? I'm afraid I still don't totally
understand the scenario you're describing. (I tried to guess before, and
apparently guessed wrong.)
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
pkg-config may help you--particularly if you have some say over the
configuration toolset used by your dependencies.
Please be more judicious in your quoting.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber:
nts to do for you.
Your build directories are configuration-specific--which can include
platform-specific aspects. To configure for a platform (or even a
particular configuration of a platform): create a directory, change to
it, and run your configure script there.
--
Braden McDaniel
On Thu, 2005-12-15 at 06:26 -0800, BRM wrote:
> --- Braden McDaniel <[EMAIL PROTECTED]> wrote:
> > On Wed, 2005-12-14 at 08:31 -0800, BRM wrote:
> > > I originally sent this to the Automake list since
> I
> > > figured it was an issue with Automake, however, I
&
on is more
complicated than that, since a Mac OS X user might also have Mesa (or
similar) installed, in which case there will be OpenGL headers in
$includedir/GL, too. I came to the conclusion that the only recourse was
to insist that the user tell the package if the Apple OpenGL framework
should be
cation to said tools, someone is out of luck. That shouldn't be you.
In other words, users who put things in arbitrary places should expect
to have to set CPPFLAGS and LDFLAGS when running configure.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
n.. how do I include a .m4 file in my
> > configure.ac?
> >
> > thanks!
>
> Oh, I just found out that putting them in acinclude.m4 does it
> automatically..
There is a better way.
Put the files with the macros in them in a directory called, for
instance, "m4".
Then
Daniel Pekelharing wrote:
How would I get autoconf to #define the install paths in my config.h ?
My app needs a shared data directory like: /usr/share/myapp,
how would I get that path?
Thanks!
I typically just do it on the command line rather than in config.h:
AM_CPPFLAGS = -DYOUR_DATADIR=\
se automake, autoconf,
> autoheader and aclocal. What's the correct order to run them in?
> I've never quite figured it out.
Run autoreconf and let it figure it out. :-)
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>
E. If the feature is enabled and the user does not
have GMP installed, fail with a message like "You must have GMP to
enable Feature X".
Otherwise, if you provide some fallback to whatever GMP provides such
that whether or not your compiled binaries use GMP is not obvious to
us
Chris Pickett wrote:
Hi,
I'd like to do unit testing for test-driven development of C programs.
I'd also like to use up-to-date autotools. I was wondering if anyone
had any experience with unit testing in their projects and had
recommendations, or even if autoconf was considering it for the
y get put into *_LDFLAGS as a unit. I do
not think I have observed any bad effects from this to date.
Is this advice really still relevant? Are all those packages that apply
-l and -L together as a unit Doing It Wrong?
--
Braden McDaniel <
On Wed, 2006-03-29 at 10:41 +0200, Ralf Wildenhues wrote:
> Hi Braden,
>
> * Braden McDaniel wrote on Wed, Mar 29, 2006 at 10:32:37AM CEST:
> > On Wed, 2006-03-29 at 10:02 +0200, Ralf Wildenhues wrote:
> >
> > > You definitely want to put dependent libraries (thos
hey probably do. Perhaps the libtool guys can chime in on
> whether the patch was a good idea.
(Not a libtool developer.)
This looks suspiciously like the -rpath that's been hosing me when I try to link
with a libGL outside the X11 directories (specifically, Mesa built in my home
directo
Matthew Woehlke wrote:
(If this isn't the place for 'how do I...?' questions, please re-direct
me, but I do see it isn't the -bugs list.)
First off, I have a very skeleton configure.in (see below). How do I add
The name "configure.ac" is preferred for autoconf >= 2.5x.
'--enable-debug' to t
user can supply the
requisite -I flags in the CPPFLAGS user variable. If either
AC_CHECK_HEADERS or even AC_CXX_HAVE_STL succeeds, you have the -I flags
you need.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
&
d tree ideas.
As I understand things, the general recommendation is to generate things
that will be included in the source package directly into the srcdir.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>
idate this better?
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
utoconf rather than
> configure.
Wildcards work in EXTRA_DIST. I've also used them successfully in my own
install-data-local rule.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
On Sun, 2007-10-14 at 21:03 -0600, Eric Blake wrote:
> According to Braden McDaniel on 10/14/2007 8:33 PM:
> > Is there something clever I could do to repeat a series of tests (in
> > testsuite.at) with different executables? The tests consist of a series
> > of input fi
the shell variable isn't expanded in the expected
output, so the check always fails.
How can I get this value into the expected output?
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>
in the manual yet - it's
> been there for some time, though).
Aha, very useful. Thanks!
I think a similar variant of AT_DATA would also be useful.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[E
sometimes
this level of resilience just isn't perceived as necessary. If you have
a good deal of faith in a library not to make subtractive API changes, a
version check is likely to be sufficient.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe
t;; I'm not convinced that's an adequate
acid test of what autoconf should support. It sounds to me as if you (or
the libraries you're using) are doing something foolish, and you deserve
to get burned. Your identifiers should be in a namespace.
--
Braden McDaniel
ew macros are likely to go to each of you in a
haphazard fashion. I cannot imagine that to be a desirable situtation.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
rg/doc/api-2.0
> Hopefully someone understands what I'm trying to do and can help.
I believe you need to use AC_TRY_LINK for C++ libraries.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
resumably it
ought to be written in sh).
--
Braden McDaniel <[EMAIL PROTECTED]>
Software Engineer, Object Sciences Corporation
"Paul Eggert" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> "Braden McDaniel" <[EMAIL PROTECTED]> writes:
>
> > The Coin <http://coin3d.org> folks have solved this by using a wrapper
that
> > translates the POSIX arguments
"Earnie Boyd" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> Braden McDaniel wrote:
> >
> > If sh is not available, presumably there would be substantial difficulty
in
> > using autoconf/configure at all.
> >
> > In practice,
"Earnie Boyd" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> Braden McDaniel wrote:
> >
> >>I know of no reason why you
> >>shouldn't be able to use cl with it. Let's discuss on
> >>[EMAIL PROTECTED] what your difficul
?
None.
configure CFLAGS="C compiler flags" CXXFLAGS="C++ compiler flags"
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
e ones they need (usually -D and -I,
and anything that affects binary compatibility).
I would suggest framing a solution in terms of pkgconfig.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
add this: view configure's ability to "fail
fast" when things aren't going to work as a side effect of its role. Its
role is to make different configurations work. If you can't make it do
that, adding logic that does nothing other than fail fast is a waste of
time.
--
`-g' flag in a way that doesn't *clobber* a user
> supplied CXXFLAGS?
You can't.
> (However, I assume resetting CXXFLAGS entirely is not wise, isn't it?)
That would blow away whatever the user had set CXXFLAGS to be.
--
Braden McDaniel
will also purge -g if it was added explicitly by the user.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
On Fri, 2003-08-01 at 17:40, Frank A. Uepping wrote:
> I want CXXFLAGS without the -g flag BY DEFAULT!
> Because the normal user doesn't want a bloat bin with debug information.
Normal users aren't going to care. And for users who do care, there is
the "install-strip"
ou're using automake, you'll have something like this in your
Makefile.am:
AM_CFLAGS = @MODULES_CFLAGS@
... since MODULES is what you called it when you invoked PKG_CHECK_MODULES.
You can call it anything you like. See pkg-config's man page for more
information.
--
Braden McDaniel
.org/software/ac-archive/htmldoc/ax_check_glut.html>
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
't know
> what to fill in the 'function' parameter to check. Nothing seems to
> work.
IIRC, AC_CHECK_LIB chokes on the C++ scope operator. Use AC_TRY_LINK
instead.
--
Braden McDaniel e-mail: <[EMAIL PROTECTED]>
<http://endoframe.com>Jabber: <[EMAIL PROTECTED]>
Bob Friesenhahn wrote:
Microsoft has now made its optimizing C/C++ compilers for Windows
available as a free download. See
"http://msdn.microsoft.com/visualc/vctoolkit2003/";. Do GNU Autotools
work with this compiler?
autoconf, basically. autoconf's check for whether it can use -g is
busted, bu
Is it possible to propagate the version (as given to AC_INIT in a
top-level package) to a configure script in a subpackage?
Braden
Paul Eggert wrote:
Braden McDaniel <[EMAIL PROTECTED]> writes:
autoconf, basically. autoconf's check for whether it can use -g is
busted, but the effect of this is just a benign warning from the
compiler. (I have a patch for this that I submitted to
autoconf-patches many months ago;
Richard Dawe wrote:
Hello.
Braden McDaniel wrote:
[snip]
Unfortunately I haven't been able to test this yet. CVS autoconf now
appears to want "help2man", which I have not yet succeeded in
building/installing on Cygwin. (The attempt sent me on quite a wild
goose chase, though:
PFLAGS='-I/home/thiago/include' LDFLAGS='-L/home/thiago/lib'
--
Braden McDaniel <[EMAIL PROTECTED]>
___
Autoconf mailing list
[EMAIL PROTECTED]
http://lists.gnu.org/mailman/listinfo/autoconf
69 matches
Mail list logo
