GNU bug report logs - #1151
Building Emacs on Mac OS 10.3.9

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 1151 in the body.
You can then email your comments to 1151 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Alex <kensanata <at> gmail.com>
To: emacs-pretest-bug <at> gnu.org
Subject: Building Emacs on Mac OS 10.3.9
Date: Sun, 12 Oct 2008 22:27:41 +0200

I'm trying to build Emacs from CVS on my old iBook running Mac OS 10.3.9.

The last Emacs I built on this machine was GNU Emacs 22.1.1
(powerpc-apple-darwin7.9.0, Carbon Version 1.6.0) of 2007-06-05 on
Alpinobombus.local

I also tried to build Emacs 22.3 on this iBook a while ago and it
ended with the following error:
ld: /usr/lib/gcc/darwin/3.3/libgcc.a(_fixunssfdi.o) illegal reference
to symbol: ___cmpdi2 defined in indirectly referenced dynamic library
/usr/lib/libgcc_s.1.dylib

So now I'm trying Emacs from CVS:

./configure --with-ns
make bootstrap

This is where it ends:

In file included from nsterm.m:49:
keyboard.h:324: warning: redefinition of `Boolean'
/System/Library/Frameworks/CoreServices.framework/Headers/../Frameworks/CarbonCore.framework/Headers/MacTypes.h:255:
warning: `Boolean' previously declared here
nsterm.m: In function `ns_term_init':
nsterm.m:3864: warning: assignment from incompatible pointer type
nsterm.m:3875: warning: assignment from incompatible pointer type
nsterm.m:3886: warning: assignment from incompatible pointer type
nsterm.m: In function `-[EmacsView keyDown:]':
nsterm.m:4340: warning: multiple declarations for method `initWithCapacity:'
/System/Library/Frameworks/Foundation.framework/Headers/NSArray.h:108:
warning: using `-(id)initWithCapacity:(unsigned)numItems'
nsterm.h:315: warning: also found `-initWithCapacity:(unsigned long)c'
gcc -c  -Demacs -DHAVE_CONFIG_H  -I. -I/Users/alex/Source/emacs/src
-Dtemacs    -g -O2  nsfns.m
In file included from nsfns.m:36:
keyboard.h:324: warning: redefinition of `Boolean'
/System/Library/Frameworks/CoreServices.framework/Headers/../Frameworks/CarbonCore.framework/Headers/MacTypes.h:255:
warning: `Boolean' previously declared here
nsfns.m: In function `interpret_services_menu':
nsfns.m:293: warning: assignment from incompatible pointer type
nsfns.m: In function `ns_do_applescript':
nsfns.m:2042: error: `typeUTF16ExternalRepresentation' undeclared
(first use in this function)
nsfns.m:2042: error: (Each undeclared identifier is reported only once
nsfns.m:2042: error: for each function it appears in.)
make[2]: *** [nsfns.o] Error 1
make[1]: *** [src] Error 2
make: *** [bootstrap] Error 2

Message #12 received at 1151 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: YAMAMOTO Mitsuharu <mituharu <at> math.s.chiba-u.ac.jp>
To: Alex <kensanata <at> gmail.com>, 1151 <at> debbugs.gnu.org
Subject: Re: bug#1151: Building Emacs on Mac OS 10.3.9
Date: Mon, 13 Oct 2008 11:41:39 +0900

>>>>> On Sun, 12 Oct 2008 22:27:41 +0200, Alex <kensanata <at> gmail.com> said:

> I also tried to build Emacs 22.3 on this iBook a while ago and it
> ended with the following error: ld:
> /usr/lib/gcc/darwin/3.3/libgcc.a(_fixunssfdi.o) illegal reference to
> symbol: ___cmpdi2 defined in indirectly referenced dynamic library
> /usr/lib/libgcc_s.1.dylib

Do some workarounds mentioned in etc/PROBLEMS work?

*** Mac OS X 10.3.9 (Carbon): QuickTime updater breaks build.

Some QuickTime updaters such as 7.0.4 and 7.2.0 are known to break
build at the link stage with the message like "Undefined symbols:
_HICopyAccessibilityActionDescription referenced from QuickTime
expected to be defined in Carbon".  A workaround is to use a QuickTime
reinstaller.  Alternatively, you can link with the frameworks in the
corresponding SDK by specifying LDFLAGS as
"-Wl,-F/Developer/SDKs/MacOSX10.3.0.sdk/System/Library/Frameworks".

				     YAMAMOTO Mitsuharu
				mituharu <at> math.s.chiba-u.ac.jp

Message #17 received at 1151 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Adrian Robert <adrian.b.robert <at> gmail.com>
To: 1151 <at> debbugs.gnu.org
Cc: Alex <kensanata <at> gmail.com>
Subject: #1151 - Building Emacs on Mac OS 10.3.9 - Emacs bug report logs
Date: Tue, 21 Oct 2008 09:42:03 -0400

Hi,

I am interested in having Emacs continue to work under Panther but  
don't have access to a test environment myself.  Could you try  
commenting out the part of the 'if' statement involving  
"typeUTF16ExternalRepresentation" and see where that gets you?

If you get it working and send me a list of all places that you needed  
to change, I can put in #ifdefs etc. so it will work out of the box.

thanks,
Adrian

Message #22 received at 1151 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: "Alex Schröder" <kensanata <at> gmail.com>
To: "Adrian Robert" <adrian.b.robert <at> gmail.com>
Cc: 1151 <at> debbugs.gnu.org
Subject: Re: #1151 - Building Emacs on Mac OS 10.3.9 - Emacs bug report logs
Date: Tue, 23 Dec 2008 02:31:43 +0100

On Tue, Oct 21, 2008 at 2:42 PM, Adrian Robert
<adrian.b.robert <at> gmail.com> wrote:
> I am interested in having Emacs continue to work under Panther but don't
> have access to a test environment myself.  Could you try commenting out the
> part of the 'if' statement involving "typeUTF16ExternalRepresentation" and
> see where that gets you?
>
> If you get it working and send me a list of all places that you needed to
> change, I can put in #ifdefs etc. so it will work out of the box.

Hi Adrian

Indeed, that does work!

This is the only change I made:

cvs diff: Diffing src
Index: src/nsfns.m
===================================================================
RCS file: /sources/emacs/emacs/src/nsfns.m,v
retrieving revision 1.29
diff -c -r1.29 nsfns.m
*** src/nsfns.m	7 Dec 2008 05:00:56 -0000	1.29
--- src/nsfns.m	23 Dec 2008 01:27:36 -0000
***************
*** 2054,2061 ****
  	  *result = Qt;
  	  // script returned an AppleScript result
  	  if ((typeUnicodeText == [returnDescriptor descriptorType]) ||
! 	      (typeUTF16ExternalRepresentation
! 	       == [returnDescriptor descriptorType]) ||
  	      (typeUTF8Text == [returnDescriptor descriptorType]) ||
  	      (typeCString == [returnDescriptor descriptorType]))
  	    {
--- 2054,2061 ----
  	  *result = Qt;
  	  // script returned an AppleScript result
  	  if ((typeUnicodeText == [returnDescriptor descriptorType]) ||
! // 	      (typeUTF16ExternalRepresentation
! // 	       == [returnDescriptor descriptorType]) ||
  	      (typeUTF8Text == [returnDescriptor descriptorType]) ||
  	      (typeCString == [returnDescriptor descriptorType]))
  	    {

Message #27 received at 1151 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Adrian Robert <adrian.b.robert <at> gmail.com>
To: Alex Schröder <kensanata <at> gmail.com>
Cc: 1151 <at> debbugs.gnu.org
Subject: Re: #1151 - Building Emacs on Mac OS 10.3.9 - Emacs bug report logs
Date: Tue, 23 Dec 2008 10:54:20 +0200

On Dec 23, 2008, at 3:31 AM, Alex Schröder wrote:

> On Tue, Oct 21, 2008 at 2:42 PM, Adrian Robert
> <adrian.b.robert <at> gmail.com> wrote:
>> I am interested in having Emacs continue to work under Panther but  
>> don't
>> have access to a test environment myself.  Could you try  
>> commenting out the
>> part of the 'if' statement involving  
>> "typeUTF16ExternalRepresentation" and
>> see where that gets you?
>>
>> If you get it working and send me a list of all places that you  
>> needed to
>> change, I can put in #ifdefs etc. so it will work out of the box.
>
> Hi Adrian
>
> Indeed, that does work!

Great to hear, I will apply a version of this patch using appropriate  
#ifdefs soon after the holidays, or feel free to commit one yourself  
if you have access.

thanks,
Adrian



> This is the only change I made:
>
> cvs diff: Diffing src
> Index: src/nsfns.m
> ===================================================================
> RCS file: /sources/emacs/emacs/src/nsfns.m,v
> retrieving revision 1.29
> diff -c -r1.29 nsfns.m
> *** src/nsfns.m	7 Dec 2008 05:00:56 -0000	1.29
> --- src/nsfns.m	23 Dec 2008 01:27:36 -0000
> ***************
> *** 2054,2061 ****
>   	  *result = Qt;
>   	  // script returned an AppleScript result
>   	  if ((typeUnicodeText == [returnDescriptor descriptorType]) ||
> ! 	      (typeUTF16ExternalRepresentation
> ! 	       == [returnDescriptor descriptorType]) ||
>   	      (typeUTF8Text == [returnDescriptor descriptorType]) ||
>   	      (typeCString == [returnDescriptor descriptorType]))
>   	    {
> --- 2054,2061 ----
>   	  *result = Qt;
>   	  // script returned an AppleScript result
>   	  if ((typeUnicodeText == [returnDescriptor descriptorType]) ||
> ! // 	      (typeUTF16ExternalRepresentation
> ! // 	       == [returnDescriptor descriptorType]) ||
>   	      (typeUTF8Text == [returnDescriptor descriptorType]) ||
>   	      (typeCString == [returnDescriptor descriptorType]))
>   	    {

Message #32 received at 1151 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: "Alex Schröder" <kensanata <at> gmail.com>
To: "Adrian Robert" <adrian.b.robert <at> gmail.com>
Cc: 1151 <at> debbugs.gnu.org
Subject: Re: #1151 - Building Emacs on Mac OS 10.3.9 - Emacs bug report logs
Date: Mon, 29 Dec 2008 12:05:11 +0100

I've come back to my Mac after a few days and noticed a problem with
the patched version. My Mac has a Swiss German keyboard. The keys
producing ASCII characters all work. Therefore the layout is correct.
But the non-ASCII are wrong: The sequence çäöüàéè results in Á‰ˆ¸‡ÈË,
for example.

I'm not sure this is related to the patch, because the patch concerns
ns_do_applescript… Hm...

> On Tue, Oct 21, 2008 at 2:42 PM, Adrian Robert wrote:
>> commenting out the
>> part of the 'if' statement involving "typeUTF16ExternalRepresentation" and
>> see where that gets you?

Message #37 received at 1151 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: "Alex Schröder" <kensanata <at> gmail.com>
To: "Adrian Robert" <adrian.b.robert <at> gmail.com>
Cc: 1151 <at> debbugs.gnu.org
Subject: Re: #1151 - Building Emacs on Mac OS 10.3.9 - Emacs bug report logs
Date: Mon, 29 Dec 2008 13:56:16 +0100

More info on the strange mapping for äöü and related non-ASCII
characters. When I use C-h k ä I get the following information:

‰ (translated from ä) runs the command self-insert-command, which is
an interactive built-in function in `C source code'.

So Emacs does at some point in time recognize the ä.

On Mon, Dec 29, 2008 at 12:05 PM, Alex Schröder <kensanata <at> gmail.com> wrote:
> I've come back to my Mac after a few days and noticed a problem with
> the patched version. My Mac has a Swiss German keyboard. The keys
> producing ASCII characters all work. Therefore the layout is correct.
> But the non-ASCII are wrong: The sequence çäöüàéè results in Á‰ˆ¸‡ÈË,
> for example.
>
> I'm not sure this is related to the patch, because the patch concerns
> ns_do_applescript… Hm...

Message #42 received at 1151 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: "Alex Schröder" <kensanata <at> gmail.com>
To: "Adrian Robert" <adrian.b.robert <at> gmail.com>
Cc: 1151 <at> debbugs.gnu.org
Subject: Re: #1151 - Building Emacs on Mac OS 10.3.9 - Emacs bug report logs
Date: Tue, 30 Dec 2008 01:07:53 +0100

I was able to reproduce this bug on my unpatched Emacs running on Mac
OS 10.4 so I will open a new bug.

On Mon, Dec 29, 2008 at 1:56 PM, Alex Schröder <kensanata <at> gmail.com> wrote:
> More info on the strange mapping for äöü and related non-ASCII
> characters. When I use C-h k ä I get the following information:
>
> ‰ (translated from ä) runs the command self-insert-command, which is
> an interactive built-in function in `C source code'.
>
> So Emacs does at some point in time recognize the ä.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.