[tex-live] Xindy broken in TL2011 pretest?

[Joachim Schrod]

Richard Koch wrote:

Dick, Simon,

I think I don't understand Simon's situation completely.

The issue, as far as I see it, is that (a) clisp (a.k.a. xindy.run)
is called and (b) the call aborts without any error message. To run
properly, the executable and the .mem file must be found, and the
.mem file must fit to the executable. (Just imagine TeX and
non-portable .fmt files and you have the right picture.)

Simon, do you use the Intel 64-bit distribution, or the 32-bit
universal distribution?

> We make two distributions. In the 64 bit distribution, which only runs
> on Intel,
> ls -l /usr/texbin/xindy.run gives
> -rwxr-xr-x 1 root wheel 2284984 May 28 13:31 xindy.run
> 
> and ls -l /usr/texbin/xindy.mem gives
> -rw-r--r-- 1 root wheel 1019780 May 28 13:31 xindy.mem

The Intel 64-bit case: Here we would need to check if (a) xindy.run
can be called and (b) that this .mem file fits and works.

> The second distribution is 32 bits universal, for both PowerPC and Intel.
> There ls -l /usr/texbin/xindy.run gives
> -rwxr-xr-x 1 root wheel 4017928 May 28 13:31 xindy.run
> 
> but ls -l /usr/texbin/xindy.mem gives
> ls: xindy.mem: No such file or directory
> 
> But that is no surprise since (as per last year) this directory has two
> mem images, one for PowerPC and one for Intel:
> 
> ls -l /usr/texbin/xindy-i386.mem
> -rw-r--r-- 1 root wheel 943707 May 28 13:31 xindy-i386.mem
> 
> and ls -l /usr/texbin/xindy-ppc.mem
> -rw-r--r-- 1 root wheel 989901 May 28 13:31 xindy-ppc.mem

The universal 32-bit case: Here we need to check (a) if it could be
called by hand, (b) if the respective .mem file fits, and (c) why
that .mem file got selected last year but isn't this year.

Simon, can you provide information what your situation is? Then I
can tell you the next commands for debugging the situation.

Best,
	Joachim

-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Joachim Schrod				Email: jschrod at acm.org
Roedermark, Germany