* install error on Mac OS X
@ 2006-11-01 0:00 Edward O'Connor
2006-11-01 0:02 ` Kazu Yamamoto
0 siblings, 1 reply; 26+ messages in thread
From: Edward O'Connor @ 2006-11-01 0:00 UTC (permalink / raw)
Hi,
'make install' on OS X isn't copying much of anything into
/usr/local/share/emacs/22.0.90:
,----[ ls -lR /usr/local/share/emacs/22.0.90 ]
| total 0
| drwxr-xr-x 2 root wheel 68 Oct 31 10:55 etc/
| drwxr-xr-x 2 root wheel 68 Oct 31 10:55 leim/
| drwxr-xr-x 2 root wheel 68 Oct 31 10:55 lisp/
| drwxr-xr-x 3 root wheel 102 Oct 31 10:55 site-lisp/
|
| /usr/local/share/emacs/22.0.90/etc:
|
| /usr/local/share/emacs/22.0.90/leim:
|
| /usr/local/share/emacs/22.0.90/lisp:
|
| /usr/local/share/emacs/22.0.90/site-lisp:
| total 8
| -rw-r--r-- 1 root wheel 106 Oct 31 10:55 subdirs.el
`----
This makes the installed Emacs binary somewhat less useful. :(
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-01 0:00 install error on Mac OS X Edward O'Connor
@ 2006-11-01 0:02 ` Kazu Yamamoto
2006-11-01 0:52 ` Edward O'Connor
2006-11-01 16:41 ` Richard Stallman
0 siblings, 2 replies; 26+ messages in thread
From: Kazu Yamamoto @ 2006-11-01 0:02 UTC (permalink / raw)
Cc: emacs-devel
> 'make install' on OS X isn't copying much of anything into
> /usr/local/share/emacs/22.0.90:
First of all, Emacs should be installed in /Application in MacOS.
The following sequence works with CVS Head.
% ./configure --enable-carbon-app=/Applications --without-x --prefix=/Applications/Emacs.app/Contents/Resources
% make
% sudo make install
--Kazu
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-01 0:02 ` Kazu Yamamoto
@ 2006-11-01 0:52 ` Edward O'Connor
2006-11-04 3:01 ` YAMAMOTO Mitsuharu
2006-11-01 16:41 ` Richard Stallman
1 sibling, 1 reply; 26+ messages in thread
From: Edward O'Connor @ 2006-11-01 0:52 UTC (permalink / raw)
> First of all, Emacs should be installed in /Application in MacOS.
>
> The following sequence works with CVS Head.
>
> % ./configure --enable-carbon-app=/Applications --without-x
> --prefix=/Applications/Emacs.app/Contents/Resources
> % make
> % sudo make install
Ahh. Hmm. When did this change? I've been doing the following, and it's
worked for a long time:
% ./configure --enable-carbon-app --without-x
% make
% sudo make install
Ted
--
Edward O'Connor
hober0@gmail.com
Ense petit placidam sub libertate quietem.
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-01 0:52 ` Edward O'Connor
@ 2006-11-04 3:01 ` YAMAMOTO Mitsuharu
0 siblings, 0 replies; 26+ messages in thread
From: YAMAMOTO Mitsuharu @ 2006-11-04 3:01 UTC (permalink / raw)
Cc: emacs-devel
>>>>> On Tue, 31 Oct 2006 19:52:46 -0500, Edward O'Connor <hober0@gmail.com> said:
> I've been doing the following, and it's worked for a long time:
> % ./configure --enable-carbon-app --without-x
> % make
> % sudo make install
What did the last command say?
BTW, --without-x is no longer needed if either --enable-carbon-app or
--with-carbon is specified.
YAMAMOTO Mitsuharu
mituharu@math.s.chiba-u.ac.jp
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-01 0:02 ` Kazu Yamamoto
2006-11-01 0:52 ` Edward O'Connor
@ 2006-11-01 16:41 ` Richard Stallman
2006-11-02 13:09 ` Randal L. Schwartz
1 sibling, 1 reply; 26+ messages in thread
From: Richard Stallman @ 2006-11-01 16:41 UTC (permalink / raw)
Cc: hober0, emacs-devel
% ./configure --enable-carbon-app=/Applications --without-x --prefix=/Applications/Emacs.app/Contents/Resources
% make
% sudo make install
If those args to ./configure are necessary,
should Emacs be changed so that it works ok
without those arguments?
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-01 16:41 ` Richard Stallman
@ 2006-11-02 13:09 ` Randal L. Schwartz
2006-11-03 7:07 ` Richard Stallman
0 siblings, 1 reply; 26+ messages in thread
From: Randal L. Schwartz @ 2006-11-02 13:09 UTC (permalink / raw)
>>>>> "Richard" == Richard Stallman <rms@gnu.org> writes:
Richard> % ./configure --enable-carbon-app=/Applications --without-x --prefix=/Applications/Emacs.app/Contents/Resources
Richard> % make
Richard> % sudo make install
Richard> If those args to ./configure are necessary,
Richard> should Emacs be changed so that it works ok
Richard> without those arguments?
My configure line is merely:
./configure --prefix=/opt/emacs --without-x
and it builds the Carbon version just fine, installing it into /opt/emacs.
It's not wise to overwrite the one in /Applications.
--
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Perl/Unix/security consulting, Technical writing, Comedy, etc. etc.
See PerlTraining.Stonehenge.com for onsite and open-enrollment Perl training!
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-02 13:09 ` Randal L. Schwartz
@ 2006-11-03 7:07 ` Richard Stallman
2006-11-03 19:08 ` Randal L. Schwartz
0 siblings, 1 reply; 26+ messages in thread
From: Richard Stallman @ 2006-11-03 7:07 UTC (permalink / raw)
Cc: emacs-devel
My configure line is merely:
./configure --prefix=/opt/emacs --without-x
Why is --without-x needed?
Should we make that automatic, for MacOS builds?
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-03 7:07 ` Richard Stallman
@ 2006-11-03 19:08 ` Randal L. Schwartz
2006-11-04 3:11 ` YAMAMOTO Mitsuharu
2006-11-05 7:07 ` install error on Mac OS X Richard Stallman
0 siblings, 2 replies; 26+ messages in thread
From: Randal L. Schwartz @ 2006-11-03 19:08 UTC (permalink / raw)
Cc: emacs-devel
>>>>> "Richard" == Richard Stallman <rms@gnu.org> writes:
Richard> My configure line is merely:
Richard> ./configure --prefix=/opt/emacs --without-x
Richard> Why is --without-x needed?
Richard> Should we make that automatic, for MacOS builds?
It may be leftover legacy. I know I needed it early on to build a carbon app
instead of an x app.
It's probably good to bias the default as a carbon app, and then requrie an
explicit switch to make an X app.
--
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Perl/Unix/security consulting, Technical writing, Comedy, etc. etc.
See PerlTraining.Stonehenge.com for onsite and open-enrollment Perl training!
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-03 19:08 ` Randal L. Schwartz
@ 2006-11-04 3:11 ` YAMAMOTO Mitsuharu
2006-11-04 8:00 ` Randal L. Schwartz
2006-11-05 7:07 ` Richard Stallman
2006-11-05 7:07 ` install error on Mac OS X Richard Stallman
1 sibling, 2 replies; 26+ messages in thread
From: YAMAMOTO Mitsuharu @ 2006-11-04 3:11 UTC (permalink / raw)
Cc: rms, emacs-devel
>>>>> On 03 Nov 2006 11:08:37 -0800, merlyn@stonehenge.com (Randal L. Schwartz) said:
>>>>> "Richard" == Richard Stallman <rms@gnu.org> writes:
Richard> My configure line is merely: ./configure --prefix=/opt/emacs
Richard> --without-x
Richard> Why is --without-x needed? Should we make that automatic,
Richard> for MacOS builds?
> It may be leftover legacy. I know I needed it early on to build a
> carbon app instead of an x app.
> It's probably good to bias the default as a carbon app, and then
> requrie an explicit switch to make an X app.
Mac OS X does not install X11 by default. And --without-x is not
necessary to build the Carbon version unless X11 was additionally
installed to the system.
YAMAMOTO Mitsuharu
mituharu@math.s.chiba-u.ac.jp
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-04 3:11 ` YAMAMOTO Mitsuharu
@ 2006-11-04 8:00 ` Randal L. Schwartz
2006-11-05 7:07 ` Richard Stallman
1 sibling, 0 replies; 26+ messages in thread
From: Randal L. Schwartz @ 2006-11-04 8:00 UTC (permalink / raw)
Cc: rms, emacs-devel
>>>>> "YAMAMOTO" == YAMAMOTO Mitsuharu <mituharu@math.s.chiba-u.ac.jp> writes:
>> It's probably good to bias the default as a carbon app, and then
>> requrie an explicit switch to make an X app.
YAMAMOTO> Mac OS X does not install X11 by default. And --without-x is not
YAMAMOTO> necessary to build the Carbon version unless X11 was additionally
YAMAMOTO> installed to the system.
Well, that's the problem then. Most systems *have* X11 installed,
in my experience. But most systems will want Carbon, not X11 Emacs,
in my opinion.
--
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Perl/Unix/security consulting, Technical writing, Comedy, etc. etc.
See PerlTraining.Stonehenge.com for onsite and open-enrollment Perl training!
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-04 3:11 ` YAMAMOTO Mitsuharu
2006-11-04 8:00 ` Randal L. Schwartz
@ 2006-11-05 7:07 ` Richard Stallman
2006-11-06 9:14 ` YAMAMOTO Mitsuharu
1 sibling, 1 reply; 26+ messages in thread
From: Richard Stallman @ 2006-11-05 7:07 UTC (permalink / raw)
Cc: emacs-devel, merlyn
Mac OS X does not install X11 by default. And --without-x is not
necessary to build the Carbon version unless X11 was additionally
installed to the system.
If the usual case on the Mac is to use Carbon and not X,
it seems to me that Emacs should ALWAYS use Carbon when built
in the default way. To use X should require a specific option.
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-05 7:07 ` Richard Stallman
@ 2006-11-06 9:14 ` YAMAMOTO Mitsuharu
2006-11-06 21:00 ` Richard Stallman
0 siblings, 1 reply; 26+ messages in thread
From: YAMAMOTO Mitsuharu @ 2006-11-06 9:14 UTC (permalink / raw)
Cc: emacs-devel, merlyn
>>>>> On Sun, 05 Nov 2006 02:07:43 -0500, Richard Stallman <rms@gnu.org> said:
> If the usual case on the Mac is to use Carbon and not X, it seems to
> me that Emacs should ALWAYS use Carbon when built in the default
> way.
I'm not sure which is the usual case. Users of precompiled binary
packages might be dominant among the Carbon-version users.
> To use X should require a specific option.
What should we do for the case that --with-x-toolkit (or other
X-specific options) is specified without explicit --with-x or
--without-carbon?
YAMAMOTO Mitsuharu
mituharu@math.s.chiba-u.ac.jp
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-06 9:14 ` YAMAMOTO Mitsuharu
@ 2006-11-06 21:00 ` Richard Stallman
2006-11-07 2:05 ` YAMAMOTO Mitsuharu
0 siblings, 1 reply; 26+ messages in thread
From: Richard Stallman @ 2006-11-06 21:00 UTC (permalink / raw)
Cc: merlyn, emacs-devel
> To use X should require a specific option.
What should we do for the case that --with-x-toolkit (or other
X-specific options) is specified without explicit --with-x or
--without-carbon?
It would make sense for --with-x-toolkit to imply --with-x.
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-06 21:00 ` Richard Stallman
@ 2006-11-07 2:05 ` YAMAMOTO Mitsuharu
2006-11-07 7:26 ` Jan Djärv
2006-11-20 22:28 ` Mac OS X, 22.0.91, configure without --enable-carbon-app Benjamin Riefenstahl
0 siblings, 2 replies; 26+ messages in thread
From: YAMAMOTO Mitsuharu @ 2006-11-07 2:05 UTC (permalink / raw)
Cc: merlyn, emacs-devel
>>>>> On Mon, 06 Nov 2006 16:00:25 -0500, Richard Stallman <rms@gnu.org> said:
>> To use X should require a specific option.
> What should we do for the case that --with-x-toolkit (or other
> X-specific options) is specified without explicit --with-x or
> --without-carbon?
> It would make sense for --with-x-toolkit to imply --with-x.
Could anyone who is familiar with autoconf check the following patch?
YAMAMOTO Mitsuharu
mituharu@math.s.chiba-u.ac.jp
Index: configure.in
===================================================================
RCS file: /cvsroot/emacs/emacs/configure.in,v
retrieving revision 1.416
diff -c -r1.416 configure.in
*** configure.in 4 Nov 2006 03:28:27 -0000 1.416
--- configure.in 7 Nov 2006 01:58:43 -0000
***************
*** 1779,1785 ****
if test "${window_system}" = x11 && test "${HAVE_CARBON}" = yes; then
if test "${with_carbon+set}" != set \
&& test "${carbon_appdir_x+set}" != set; then
! HAVE_CARBON=no
fi
fi
if test "${HAVE_CARBON}" = yes; then
--- 1779,1791 ----
if test "${window_system}" = x11 && test "${HAVE_CARBON}" = yes; then
if test "${with_carbon+set}" != set \
&& test "${carbon_appdir_x+set}" != set; then
! for var in with_x with_x_toolkit with_gtk with_xim \
! with_xpm with_jpeg with_tiff with_gif with_png; do
! if test "`eval 'echo \"${'\"$var\"'+set}\"'`" = set; then
! HAVE_CARBON=no
! break
! fi
! done
fi
fi
if test "${HAVE_CARBON}" = yes; then
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-07 2:05 ` YAMAMOTO Mitsuharu
@ 2006-11-07 7:26 ` Jan Djärv
2006-11-07 11:55 ` Andreas Schwab
2006-11-20 22:28 ` Mac OS X, 22.0.91, configure without --enable-carbon-app Benjamin Riefenstahl
1 sibling, 1 reply; 26+ messages in thread
From: Jan Djärv @ 2006-11-07 7:26 UTC (permalink / raw)
Cc: emacs-devel, rms, merlyn
YAMAMOTO Mitsuharu skrev:
>>>>>> On Mon, 06 Nov 2006 16:00:25 -0500, Richard Stallman <rms@gnu.org> said:
>
>>> To use X should require a specific option.
>> What should we do for the case that --with-x-toolkit (or other
>> X-specific options) is specified without explicit --with-x or
>> --without-carbon?
>
>> It would make sense for --with-x-toolkit to imply --with-x.
>
> Could anyone who is familiar with autoconf check the following patch?
>
> YAMAMOTO Mitsuharu
> mituharu@math.s.chiba-u.ac.jp
>
> Index: configure.in
> ===================================================================
> RCS file: /cvsroot/emacs/emacs/configure.in,v
> retrieving revision 1.416
> diff -c -r1.416 configure.in
> *** configure.in 4 Nov 2006 03:28:27 -0000 1.416
> --- configure.in 7 Nov 2006 01:58:43 -0000
> ***************
> *** 1779,1785 ****
> if test "${window_system}" = x11 && test "${HAVE_CARBON}" = yes; then
> if test "${with_carbon+set}" != set \
> && test "${carbon_appdir_x+set}" != set; then
> ! HAVE_CARBON=no
> fi
> fi
> if test "${HAVE_CARBON}" = yes; then
> --- 1779,1791 ----
> if test "${window_system}" = x11 && test "${HAVE_CARBON}" = yes; then
> if test "${with_carbon+set}" != set \
> && test "${carbon_appdir_x+set}" != set; then
> ! for var in with_x with_x_toolkit with_gtk with_xim \
> ! with_xpm with_jpeg with_tiff with_gif with_png; do
> ! if test "`eval 'echo \"${'\"$var\"'+set}\"'`" = set; then
> ! HAVE_CARBON=no
> ! break
> ! fi
> ! done
> fi
> fi
> if test "${HAVE_CARBON}" = yes; then
For clarity, and to get rid of the backticks, I would write the for loop like
this:
for var in with_x with_x_toolkit with_gtk with_xim \
with_xpm with_jpeg with_tiff with_gif with_png; do
eval "x=\${$var+set}";
if test "$x" = set; then
...
Other than that, it looks fine.
Jan D.
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-07 7:26 ` Jan Djärv
@ 2006-11-07 11:55 ` Andreas Schwab
2006-11-08 8:13 ` YAMAMOTO Mitsuharu
0 siblings, 1 reply; 26+ messages in thread
From: Andreas Schwab @ 2006-11-07 11:55 UTC (permalink / raw)
Cc: merlyn, rms, YAMAMOTO Mitsuharu, emacs-devel
Jan Djärv <jan.h.d@swipnet.se> writes:
> For clarity, and to get rid of the backticks, I would write the for loop
> like this:
>
> for var in with_x with_x_toolkit with_gtk with_xim \
> with_xpm with_jpeg with_tiff with_gif with_png; do
> eval "x=\${$var+set}";
> if test "$x" = set; then
Or even this:
if eval test \"\${$var+set}\" = set; then
Andreas.
--
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756 01D3 44D5 214B 8276 4ED5
"And now for something completely different."
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-07 11:55 ` Andreas Schwab
@ 2006-11-08 8:13 ` YAMAMOTO Mitsuharu
0 siblings, 0 replies; 26+ messages in thread
From: YAMAMOTO Mitsuharu @ 2006-11-08 8:13 UTC (permalink / raw)
Cc: merlyn, Jan Djärv, rms, emacs-devel
>>>>> On Tue, 07 Nov 2006 12:55:49 +0100, Andreas Schwab <schwab@suse.de> said:
> Jan Djärv <jan.h.d@swipnet.se> writes:
>> For clarity, and to get rid of the backticks, I would write the for loop
>> like this:
>>
>> for var in with_x with_x_toolkit with_gtk with_xim \
>> with_xpm with_jpeg with_tiff with_gif with_png; do
>> eval "x=\${$var+set}";
>> if test "$x" = set; then
> Or even this:
> if eval test \"\${$var+set}\" = set; then
Thanks for the comments. They are much cleaner.
YAMAMOTO Mitsuharu
mituharu@math.s.chiba-u.ac.jp
^ permalink raw reply [flat|nested] 26+ messages in thread
* Mac OS X, 22.0.91, configure without --enable-carbon-app
2006-11-07 2:05 ` YAMAMOTO Mitsuharu
2006-11-07 7:26 ` Jan Djärv
@ 2006-11-20 22:28 ` Benjamin Riefenstahl
2006-11-22 8:27 ` YAMAMOTO Mitsuharu
1 sibling, 1 reply; 26+ messages in thread
From: Benjamin Riefenstahl @ 2006-11-20 22:28 UTC (permalink / raw)
Hi all,
I just compiled yesterday's pretest on Mac OS X. I tried a simple
./configure --prefix=$HOME/emacs-pretest
This detects Carbon and also enables Emacs.app, but "make install"
tries to install Emacs.app in "/". The reason is probably that
configure sets the variable "carbon_appdir" to the empty string in
Makefile. When I give --enable-carbon-app explicitly (but without an
explicit path), "carbon_appdir" is set to "/Applications", which is
the documented default.
Ideally of course, I would like to have --enable-carbon-app follow
--prefix, when I give --prefix.
benny
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: Mac OS X, 22.0.91, configure without --enable-carbon-app
2006-11-20 22:28 ` Mac OS X, 22.0.91, configure without --enable-carbon-app Benjamin Riefenstahl
@ 2006-11-22 8:27 ` YAMAMOTO Mitsuharu
2006-11-24 11:50 ` Benjamin Riefenstahl
0 siblings, 1 reply; 26+ messages in thread
From: YAMAMOTO Mitsuharu @ 2006-11-22 8:27 UTC (permalink / raw)
Cc: emacs-devel
>>>>> On Mon, 20 Nov 2006 23:28:22 +0100, Benjamin Riefenstahl <b.riefenstahl@turtle-trading.net> said:
> I just compiled yesterday's pretest on Mac OS X. I tried a simple
> ./configure --prefix=$HOME/emacs-pretest
> This detects Carbon and also enables Emacs.app, but "make install"
> tries to install Emacs.app in "/".
I couldn't reproduced it.
> The reason is probably that configure sets the variable
> "carbon_appdir" to the empty string in Makefile.
In that case, the bundle should not be not installed.
Makefile.in:
if test "${carbon_appdir}" != ""; then \
umask 022; mkdir -p $(DESTDIR)${carbon_appdir}/Emacs.app; \
(cd mac/Emacs.app; (tar -chf - . | \
(cd $(DESTDIR)${carbon_appdir}/Emacs.app; umask 022; tar -xvf - \
&& cat > /dev/null))) || exit 1; \
fi
YAMAMOTO Mitsuharu
mituharu@math.s.chiba-u.ac.jp
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: Mac OS X, 22.0.91, configure without --enable-carbon-app
2006-11-22 8:27 ` YAMAMOTO Mitsuharu
@ 2006-11-24 11:50 ` Benjamin Riefenstahl
2006-11-25 2:11 ` YAMAMOTO Mitsuharu
0 siblings, 1 reply; 26+ messages in thread
From: Benjamin Riefenstahl @ 2006-11-24 11:50 UTC (permalink / raw)
Hi Yamamoto,
YAMAMOTO Mitsuharu writes:
> I couldn't reproduced it.
Sorry, that was a misinterpretation on my part.
I just looked at the output of "make -n install" and didn't actually
look hard enough at the exact code generated. I saw
umask 022; mkdir -p /Emacs.app; \
and thought it was broken, because I didn't see the surrounding "if"
statement.
So --enable-carbon-app is not supposed to be the default on Mac? I
would have thought that would be the natural setting.
benny
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: Mac OS X, 22.0.91, configure without --enable-carbon-app
2006-11-24 11:50 ` Benjamin Riefenstahl
@ 2006-11-25 2:11 ` YAMAMOTO Mitsuharu
2006-11-26 18:46 ` Benjamin Riefenstahl
0 siblings, 1 reply; 26+ messages in thread
From: YAMAMOTO Mitsuharu @ 2006-11-25 2:11 UTC (permalink / raw)
Cc: emacs-devel
>>>>> On Mon, 20 Nov 2006 23:28:22 +0100, Benjamin Riefenstahl <b.riefenstahl@turtle-trading.net> said:
> Ideally of course, I would like to have --enable-carbon-app follow
> --prefix, when I give --prefix.
>>>>> On Fri, 24 Nov 2006 12:50:09 +0100, Benjamin Riefenstahl <b.riefenstahl@turtle-trading.net> said:
> So --enable-carbon-app is not supposed to be the default on Mac? I
> would have thought that would be the natural setting.
I'm confused because the above two look inconsistent. Which
destination of Emacs.app do you think is natural when no options are
provided, /usr/local/Applications or /Applications?
YAMAMOTO Mitsuharu
mituharu@math.s.chiba-u.ac.jp
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: Mac OS X, 22.0.91, configure without --enable-carbon-app
2006-11-25 2:11 ` YAMAMOTO Mitsuharu
@ 2006-11-26 18:46 ` Benjamin Riefenstahl
2006-11-27 3:31 ` YAMAMOTO Mitsuharu
0 siblings, 1 reply; 26+ messages in thread
From: Benjamin Riefenstahl @ 2006-11-26 18:46 UTC (permalink / raw)
Cc: emacs-devel
Hi Yamamoto,
>>>>>> On Mon, 20 Nov 2006 23:28:22 +0100, Benjamin Riefenstahl
> <b.riefenstahl@turtle-trading.net> said:
>
>> Ideally of course, I would like to have --enable-carbon-app follow
>> --prefix, when I give --prefix.
>
>>>>>> On Fri, 24 Nov 2006 12:50:09 +0100, Benjamin Riefenstahl
> <b.riefenstahl@turtle-trading.net> said:
>
>> So --enable-carbon-app is not supposed to be the default on Mac? I
>> would have thought that would be the natural setting.
YAMAMOTO Mitsuharu writes:
> I'm confused because the above two look inconsistent.
That's because they are :-(. The Mac OS X way of handling
installations of applications and the traditional Unix way do not
agree.
Let me give my expectations in order of priority:
a) When I compile Emacs on Mac OS X using the defaults, I expect to
get an Emacs according to the customs of that OS. That means I
expect that "make install" puts an Emacs.app into /Applications. I
would think this also is the expectation of other users, especially
those that do not have prior experience in Unix.
b) When I give a --prefix arg to ./configure, I expect that everything
is put there that "make install" creates, including Emacs.app
(probably in a subdirectory Applications).
c) I don't have any preconceived expectations about how
--enable-carbon-app behaves, with or without an argument.
To implement these, I guess I would do away with --disable-carbon-app
(we still do have --without-carbon), just call it --carbon-appdir and
document it as:
--carbon-appdir=DIR specify install directory for Emacs.app on Mac
OS X [/Applications or, iff --prefix is given,
PREFIX/Applications]
Examples:
$ ./configure
=> carbon_appdir=/Applications
$ ./configure --prefix=path
=> carbon_appdir=$prefix/Applications
$ ./configure [--prefix=path1] --carbon-appdir=path2
=> carbon_appdir=path2
$ ./configure [--prefix=path1] --without-carbon
=> carbon_appdir=
If you think this is too complicated to explain, one could drop the
interaction with --prefix.
> Which destination of Emacs.app do you think is natural when no
> options are provided, /usr/local/Applications or /Applications?
/usr/local/Applications has no significance in Mac OS X AFAIK, so I
think nobody looks there for installed applications. Users are used
to manipulate stuff in /Applications manually, it's not considered
much of a system directory, so installing there should be ok.
benny
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: Mac OS X, 22.0.91, configure without --enable-carbon-app
2006-11-26 18:46 ` Benjamin Riefenstahl
@ 2006-11-27 3:31 ` YAMAMOTO Mitsuharu
2006-11-28 16:08 ` Steven Tamm
0 siblings, 1 reply; 26+ messages in thread
From: YAMAMOTO Mitsuharu @ 2006-11-27 3:31 UTC (permalink / raw)
Cc: emacs-devel
>>>>> On Sun, 26 Nov 2006 19:46:16 +0100, Benjamin Riefenstahl <b.riefenstahl@turtle-trading.net> said:
> Let me give my expectations in order of priority:
> a) When I compile Emacs on Mac OS X using the defaults, I expect to
> get an Emacs according to the customs of that OS. That means I
> expect that "make install" puts an Emacs.app into /Applications. I
> would think this also is the expectation of other users, especially
> those that do not have prior experience in Unix.
This breaks the following message shown by ./configure --help.
By default, `make install' will install all the files in
`/usr/local/bin', `/usr/local/lib' etc. You can specify
an installation prefix other than `/usr/local' using `--prefix',
for instance `--prefix=$HOME'.
This originally comes from /usr/share/autoconf/autoconf/general.m4.
Anyway, according to ChangeLog, --enable-carbon-app was added by
Steven Tamm. He is also the current maintainer of the Mac Carbon port
and the author of mac/make-package. So I think he is the best person
to ask.
YAMAMOTO Mitsuharu
mituharu@math.s.chiba-u.ac.jp
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: Mac OS X, 22.0.91, configure without --enable-carbon-app
2006-11-27 3:31 ` YAMAMOTO Mitsuharu
@ 2006-11-28 16:08 ` Steven Tamm
0 siblings, 0 replies; 26+ messages in thread
From: Steven Tamm @ 2006-11-28 16:08 UTC (permalink / raw)
Cc: Benjamin Riefenstahl, emacs-devel
I tried to avoid mucking around with the make file that much because adding in all my DMG creation seemed like unnecessary bloat (like adding a --mac-app-prefix). Since the /usr folder is hidden on OSX, it seemed that keeping it separate was a good idea.
I wouldn't be against moving this from make-package to the makefile, but I don't think it's that big a deal.
-Steven
On Sunday, November 26, 2006, at 07:32PM, "YAMAMOTO Mitsuharu" <mituharu@math.s.chiba-u.ac.jp> wrote:
>>>>>> On Sun, 26 Nov 2006 19:46:16 +0100, Benjamin Riefenstahl <b.riefenstahl@turtle-trading.net> said:
>
>> Let me give my expectations in order of priority:
>
>> a) When I compile Emacs on Mac OS X using the defaults, I expect to
>> get an Emacs according to the customs of that OS. That means I
>> expect that "make install" puts an Emacs.app into /Applications. I
>> would think this also is the expectation of other users, especially
>> those that do not have prior experience in Unix.
>
>This breaks the following message shown by ./configure --help.
>
> By default, `make install' will install all the files in
> `/usr/local/bin', `/usr/local/lib' etc. You can specify
> an installation prefix other than `/usr/local' using `--prefix',
> for instance `--prefix=$HOME'.
>
>This originally comes from /usr/share/autoconf/autoconf/general.m4.
>
>Anyway, according to ChangeLog, --enable-carbon-app was added by
>Steven Tamm. He is also the current maintainer of the Mac Carbon port
>and the author of mac/make-package. So I think he is the best person
>to ask.
>
> YAMAMOTO Mitsuharu
> mituharu@math.s.chiba-u.ac.jp
>
>
>_______________________________________________
>Emacs-devel mailing list
>Emacs-devel@gnu.org
>http://lists.gnu.org/mailman/listinfo/emacs-devel
>
>
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
2006-11-03 19:08 ` Randal L. Schwartz
2006-11-04 3:11 ` YAMAMOTO Mitsuharu
@ 2006-11-05 7:07 ` Richard Stallman
1 sibling, 0 replies; 26+ messages in thread
From: Richard Stallman @ 2006-11-05 7:07 UTC (permalink / raw)
Cc: emacs-devel
It's probably good to bias the default as a carbon app, and then requrie an
explicit switch to make an X app.
I would tend to agree.
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: install error on Mac OS X
@ 2006-11-02 13:57 Seiji Zenitani
0 siblings, 0 replies; 26+ messages in thread
From: Seiji Zenitani @ 2006-11-02 13:57 UTC (permalink / raw)
Hello,
On 2006/11/02, at 8:37, Kazu Yamamoto wrote:
>
> % ./configure --enable-carbon-app=/Applications --without-x --
> prefix=/Applications/Emacs.app/Contents/Resources
> % make
> % sudo make install
>
> If those args to ./configure are necessary,
> should Emacs be changed so that it works ok
> without those arguments?
>
Yamamoto-san's way is favorable because everything is enclosed in
the /Application directory.
By the way, Mac users often move an Emacs icon (/Applications/
Emacs.app) to other place (/Somewhere/Else/Emacs.app). Therefore, it
is much more favorable if emacs is independent of its prefix path.
Take a look at the mac/make-package script in the emacs source code.
The line
./make-package --self-contained
will build a prefix-independent version of emacs, by re-arranging its
internal directory structure. I recommend this 'self-contained' way
rather than the above Yamamoto-san's way.
Also, for simplicity, I think 'mac/make-package' should be integrated
into the configure script. How about removing 'mac/make-package' and
then introducing a new configure option, something like './configure
--self-contained'?
best regards,
Seiji Zenitani
zenitani@gmail.com
^ permalink raw reply [flat|nested] 26+ messages in thread
end of thread, other threads:[~2006-11-28 16:08 UTC | newest]
Thread overview: 26+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-11-01 0:00 install error on Mac OS X Edward O'Connor
2006-11-01 0:02 ` Kazu Yamamoto
2006-11-01 0:52 ` Edward O'Connor
2006-11-04 3:01 ` YAMAMOTO Mitsuharu
2006-11-01 16:41 ` Richard Stallman
2006-11-02 13:09 ` Randal L. Schwartz
2006-11-03 7:07 ` Richard Stallman
2006-11-03 19:08 ` Randal L. Schwartz
2006-11-04 3:11 ` YAMAMOTO Mitsuharu
2006-11-04 8:00 ` Randal L. Schwartz
2006-11-05 7:07 ` Richard Stallman
2006-11-06 9:14 ` YAMAMOTO Mitsuharu
2006-11-06 21:00 ` Richard Stallman
2006-11-07 2:05 ` YAMAMOTO Mitsuharu
2006-11-07 7:26 ` Jan Djärv
2006-11-07 11:55 ` Andreas Schwab
2006-11-08 8:13 ` YAMAMOTO Mitsuharu
2006-11-20 22:28 ` Mac OS X, 22.0.91, configure without --enable-carbon-app Benjamin Riefenstahl
2006-11-22 8:27 ` YAMAMOTO Mitsuharu
2006-11-24 11:50 ` Benjamin Riefenstahl
2006-11-25 2:11 ` YAMAMOTO Mitsuharu
2006-11-26 18:46 ` Benjamin Riefenstahl
2006-11-27 3:31 ` YAMAMOTO Mitsuharu
2006-11-28 16:08 ` Steven Tamm
2006-11-05 7:07 ` install error on Mac OS X Richard Stallman
-- strict thread matches above, loose matches on Subject: below --
2006-11-02 13:57 Seiji Zenitani
Code repositories for project(s) associated with this external index
https://git.savannah.gnu.org/cgit/emacs.git
https://git.savannah.gnu.org/cgit/emacs/org-mode.git
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.