[OS X TeX] Re: Xindy?

[Justin C. Walker]

On Jul 17, 2006, at 08:13 , Joachim Schrod wrote:

> Thanks a lot for your interest in xindy. While I don't have a Mac  
> myself and don't have access to one either, I'm very interested in  
> improving the support for xindy on Mac OS X. As Maarten already  
> mentioned, if there are any changes that we can make to improve the  
> compilation, I'm all for it. E.g., we can explicitly unset *FLAG  
> variables if we detect a Darwin environment, if that would be  
> sufficient. But currently the thread reads as if still some issues  
> are out there.

I have both types of Macs, and am happy to help resolve this.  See  
below.

> Jonathan Kew wrote:
>
>> And looks very similar to what I built a couple of weeks back,  
>> though I didn't have this exact source archive as my starting  
>> point. (See http://www.jfkew.plus.com/xindy-macosx.html for the  
>> ready-to-use package, as well as notes on what I used.)
>
> The changes that are described on that page (to the C header file  
> ordrules.h, don't use readline) have been folded into the source  
> tarball that is used in these tests. Of course, if there are more  
> changes

Is there a problem with readline?  I know that Mac OS X (10.4) tried  
to get cute with readline by equating it with some oddball library  
(libedit); I have moved that out of the way.  GNU readline is in /usr/ 
local.

>> (1) I suspect it's a good idea to purge any hint of Fink's /sw  
>> tree from your paths, as the build is known to work on a Fink-free  
>> system but seems liable to have problems when other versions of  
>> stuff -- could be headers & libraries, autotools stuff, other  
>> utilities.... who knows! -- get mixed in. Different versions of  
>> autotools support files, for example, could easily affect things.
>
> The next release will try to avoid usage of autotools stuff outside  
> of xindy, maybe that helps here, too.
>
>> (2) There might be a real issue with Intel-based Macs, as this is  
>> NOT listed as a supported platform in rte/clisp-2.33.2/ffcall/ 
>> trampoline/PLATFORMS. (Maybe it's OK, and the file just hasn't  
>> been updated, but I'd be concerned as this is low-level CPU and OS- 
>> specific stuff.)
>
> I haven't followed this thread fully -- Maarten only recently  
> brought it to my attention. Does anybody here knows if the current  
> version of CLISP (that is 2.38) compiles on Intel-based Macs? If  
> that's the case, that would be enough reason to use it in xindy's  
> distribution.

I have SAGE built on both platforms; this uses clisp 2.38.p1, and I  
don't yet know whether there was a problem building the "stock" 2.38  
version on either platform.

Now for some real weirdness:

I note the following:

1) I don't see "/sw" show up in any environment variable, although I  
do have a symlink to "/sw/lib/libfac.a" in /usr/local/lib.

2) I renamed "/sw" to make sure it wasn't interfering.  I reinstalled  
xindy and did configure/make.  This time, it passed the tests, but  
blew up in the build of clisp (tail of the 'make' output below).

3) I moved "/opt" (DarwinPorts) and "/usr/local" out of the way.  Bad  
Ideal (xindy wants to see TeX support around, and it's in /usr/local).

4) I moved "/usr/local" back, and left "/opt" moved out of the way.   
I reinstalled xindy, and rebuilt.  Back to blowing up at the  
trampoline test.

5) I put "/opt" back, reinstalled, rebuilt.  Unlike (2), it blew up  
at the trampoline test.

So now I am officially Confused.

It's almost as if we're fighting Cosmic Rays.

Regards,

Justin

=====tail of 'make' output:

Compiling file /private/tmp/xindy-2.2-beta2/rte/clisp-2.33.2/src/ 
regexp/regexp.lisp ...

Wrote file /private/tmp/xindy-2.2-beta2/rte/clisp-2.33.2/src/regexp/ 
regexp.fas
The following functions were used but not defined:
REGEXP:REGEXP-COMPILE REGEXP:REGEXP-EXEC
0 errors, 0 warnings
/tmp/xindy-2.2-beta2/rte/clisp-2.33.2/src/lisp.run -M /tmp/xindy-2.2- 
beta2/rte/clisp-2.33.2/src/lispinit.mem -B /tmp/xindy-2.2-beta2/rte/ 
clisp-2.33.2/src -Efile UTF-8 -Eterminal UTF-8 -norc -q ../ 
modprep.fas regexi.c
;; MODPREP: "regexi.c" --> #P"regexi.m.c"
;; MODPREP: reading "regexi.c": 4,049 bytes,
*** - UNIX error 45 (EOPNOTSUPP): Operation not supported on socket
Real time: 0.009905 sec.
Run time: 0.009778 sec.
Space: 57876 Bytes

make[3]: *** [regexi.m.c] Error 1
make[2]: *** [regexp] Error 2
make[1]: *** [xindy.run] Error 2
make: *** [all-recursive] Error 1

--
Justin C. Walker, Curmudgeon-At-Large
Institute for the Enhancement of the Director's Income
--------
Experience is what you get
   when you don't get what you want.
--------



------------------------- Info --------------------------
Mac-TeX Website: http://www.esm.psu.edu/mac-tex/
          & FAQ: http://latex.yauh.de/faq/
TeX FAQ: http://www.tex.ac.uk/faq
List Archive: http://tug.org/pipermail/macostex-archives/