Re: Request: New beta release

Hi Vladimir,

Vladimir Volovich schrieb am Wed 28. Feb, 20:00 (+0300):
> "JS" == Jörg Sommer writes:
> 
>  >> These variables are only used in actions that are triggered when
>  >> local (distributed!) files change; notably when the *.{m4,ac,am}
>  >> files are newer than the generated files. I found no other usage
>  >> of these files.  And in the distributed beta2 archive, the
>  >> {configure,Makefile.in} files are newer than
>  >> {configure.ac,Makefile.am}.
> 
>  JS> If you run make with -d, you can see what he is doing. But I
>  JS> didn't understand, why it does it.
> 
> OK, i've debugged this. there are dependencies in the Makefile which
> lead to regeneration of aclocal.m4, Makefile.in and configure because
> of wrong timestamps.
> 
> to prevent these regenerations, you need to:
> * first touch all aclocal.m4 files
> * then  touch all Makefile.in files
> * then  touch all configure files

I tried it and it works. The problem are be the time stamps in the
archive. I changed my debian/rules files and do a new upload next week.

> $ tar xfz ~/src/xindy/xindy-2.2-beta2.tar.gz
> $ cp -a xindy-2.2-beta2 xindy-2.2-beta2.orig
> $ cd xindy-2.2-beta2

Re: Request: New beta release

"JS" == Jörg Sommer writes:

 >> $ touch `find . -name aclocal.m4`
 >> $ sleep 2

 JS> touch has the option -d to give it a different time stamp than
 JS> now.

yes, but which specific date to put there? it was easier to put
current timestamp. it should be done once in the source tarball,
anyway (and upon further development, the timestamps should be
correct, if one always reautoconfs the sources after e.g. changing
Makefile.am).

 >> The file xindy-2.2-beta2/rte/clisp-2.33.2/src/Makefile~ is not
 >> removed

 JS> Where does it come from? I don't have this file.

This Makefile~ is generated by a rule in
xindy-2.2-beta2/rte/clisp-2.33.2/src/Makefile which is generated from
xindy-2.2-beta2/rte/clisp-2.33.2/src/makemake.in (look there for
"Makefile~").

Best,
v.

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your

Re: Request: New beta release

On Mon, 12 Feb 2007 01:17:15 +0100
Joachim Schrod <jschrod <at> acm.org> wrote:

Hi!

> The following showstopper issues are still open:
> 
>  -- Compilation with gcc 4 does not work [bug ticket 1515980]

Today I tried to build xindy and compilation fails indeed.

However, I've the same issue with compiling clisp-2.41 as well...

I'll watch Gentoo's bug to see how it will resolve, but I vote for the
attempt to try to make xindy-release with clisp-2.41, if possible.

> The first issue means that the rc3 tarball does not even compile on my
> own system; thus I really don't think that it's sensible to release
> it.

I can help with the distribution issues when the clisp ones are
resolved.

Sincerely,
Gour

p.s. Pls. excuse me for late reply, but I'm behind my email a bit...

-------------------------------------------------------------------------

Re: Request: New beta release

"G" == Gour  writes:

 G> Today I tried to build xindy and compilation fails indeed.

are the problems caused by a compiler bug, perhaps? what is the error message?

 G> However, I've the same issue with compiling clisp-2.41 as well...

i don't have problems with GCC 4.1.2 - xindy builds fine (even with
the included clisp sources).

Best,
v.

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Re: Request: New beta release

On Sun, 04 Mar 2007 16:32:14 +0300
Vladimir Volovich <vvv <at> vsu.ru> wrote:

> are the problems caused by a compiler bug, perhaps? what is the error
> message?

...
gmake[2]: Leaving directory
`/var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build/libcharset/lib' /bin/sh
../../libcharset/build-aux/mkinstalldirs /var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build
/usr/bin/install -c -m 644
include/libcharset.h
/var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build/libcharset.h /usr/bin/install
-c -m 644
include/localcharset.h /var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build/localcharset.h
gmake[1]: Leaving directory
`/var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build/libcharset'
gcc  -march=k8 -O2 -pipe -W -Wswitch -Wcomment -Wpointer-arith
-Wimplicit -Wreturn-type -Wmissing-declarations -Wno-sign-compare -O
-DUNICODE -DDYNAMIC_FFI -I. -c spvw.c In file included from
lispbibl.d:1979, from spvw.d:24: unix.d:178:56: error: asm/page.h: No
such file or directory make: *** [spvw.o] Error 1

> i don't have problems with GCC 4.1.2 - xindy builds fine (even with
> the included clisp sources).

Interesting...

In the upcoming week I'll try to take a closer look at the
issue...beta for lyx-1.5.0 is out bringing Unicode, so it would be nice

Re: Request: New beta release

"G" == Gour  writes:

 G> unix.d:178:56: error: asm/page.h: No such file or directory

are you building on linux? do you have the linux-kernel-headers
package installed (or it may be named differently in your
distribution)?

the error message you cited looks like caused by absent system headers.
it doesn't look like dependent on compiler version.

Best,
v.

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Re: Request: New beta release

On Tue, 06 Mar 2007 10:27:48 +0300
Vladimir Volovich <vvv <at> vsu.ru> wrote:

> are you building on linux? 

Yes.

> do you have the linux-kernel-headers package installed (or it may be
> named differently in your distribution)?

Yes, I'm on Gentoo, so everything is built from the source ;)

> the error message you cited looks like caused by absent system
> headers. it doesn't look like dependent on compiler version.

I'll investigate a bit...

Sincerely,
Gour

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your

Re: Request: New beta release

Hallo Gour,

Gour schrieb am Sun 04. Mar, 20:19 (+0100):
> On Sun, 04 Mar 2007 16:32:14 +0300
> Vladimir Volovich <vvv <at> vsu.ru> wrote:
> 
> > are the problems caused by a compiler bug, perhaps? what is the error
> > message?
> 
> ...
> gmake[2]: Leaving directory
> `/var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build/libcharset/lib' /bin/sh
../../libcharset/build-aux/mkinstalldirs /var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build
> /usr/bin/install -c -m 644
> include/libcharset.h
/var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build/libcharset.h /usr/bin/install
> -c -m 644
> include/localcharset.h /var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build/localcharset.h
> gmake[1]: Leaving directory
> `/var/tmp/portage/dev-lisp/clisp-2.41/work/clisp-2.41/build/libcharset'
> gcc  -march=k8 -O2 -pipe -W -Wswitch -Wcomment -Wpointer-arith
> -Wimplicit -Wreturn-type -Wmissing-declarations -Wno-sign-compare -O
> -DUNICODE -DDYNAMIC_FFI -I. -c spvw.c In file included from
> lispbibl.d:1979, from spvw.d:24: unix.d:178:56: error: asm/page.h: No
> such file or directory make: *** [spvw.o] Error 1

Maybe the build log from the debian package helps you:
http://buildd.debian.org/fetch.cgi?pkg=clisp;ver=1%3A2.41-1;arch=amd64;stamp=1161063779

Schöne Grüße, Jörg.

Fix index with Xindy

Hi,

for whatever reasons, I get the idx file

\indexentry{Kruskel|hyperpage}{46}
\indexentry{$\ <at> mathrm {NC}$|hyperpage}{53}
\indexentry{hochgradig parallelisierbar|hyperpage}{53}
\indexentry{$\ <at> mathrm {NC}$-berechenbar|hyperpage}{53}
\indexentry{$\ <at> mathrm {NC}$-reduzierbar|hyperpage}{53}
\indexentry{$\ <at> mathrm {P}$-vollst\"andig|hyperpage}{53}
\indexentry{polylogarithmisch|hyperpage}{53}
\indexentry{polynomiell|hyperpage}{53}

As you see, the \ <at> mathrm entries are broken. It's a bug in a LaTeX package.

My question is now: Can I repair this with Xindy? tex2xindy produces:

(indexentry :tkey (("Kruskel")) :attr "hyperpage" :locref "46")
(indexentry :tkey (("$\\" "mathrm {NC}$")) :attr "hyperpage" :locref "53")
(indexentry :tkey (("hochgradig parallelisierbar")) :attr "hyperpage" :locref "53")
(indexentry :tkey (("$\\" "mathrm {NC}$-berechenbar")) :attr "hyperpage" :locref "53")
(indexentry :tkey (("$\\" "mathrm {NC}$-reduzierbar")) :attr "hyperpage" :locref "53")
(indexentry :tkey (("$\\" "mathrm {P}$-vollst\\\"andig")) :attr "hyperpage" :locref "53")
(indexentry :tkey (("polylogarithmisch")) :attr "hyperpage" :locref "53")
(indexentry :tkey (("polynomiell")) :attr "hyperpage" :locref "53")

Is it possible to repair this with a module file? Can I preprocess the
input and concatenate the two strings "$\\" and "mathrm{P}$"? How?

Regards, Jörg.

Re: Fix index with Xindy

>>>>> "JS" == Jörg Sommer <joerg <at> alea.gnuu.de> writes:

Hello,

JS> for whatever reasons, I get the idx file

JS> \indexentry{Kruskel|hyperpage}{46}
JS> \indexentry{$\ <at> mathrm {NC}$|hyperpage}{53}
JS> \indexentry{hochgradig parallelisierbar|hyperpage}{53}
JS> \indexentry{$\ <at> mathrm {NC}$-berechenbar|hyperpage}{53}
JS> \indexentry{$\ <at> mathrm {NC}$-reduzierbar|hyperpage}{53}
JS> \indexentry{$\ <at> mathrm {P}$-vollst\"andig|hyperpage}{53}
JS> \indexentry{polylogarithmisch|hyperpage}{53}
JS> \indexentry{polynomiell|hyperpage}{53}

JS> As you see, the \ <at> mathrm entries are broken. It's a bug in a LaTeX package.

JS> My question is now: Can I repair this with Xindy?

Sadly, no. (And I consider that a bug in tex2xindy; but alas, it's not
easy to repair.)

You can get a workaround locally by recompiling tex2xindy: In
tex2xindy.l, change the line with `ACTUAL' to an other character that
you don't use (e.g., % or so) and recompile. Then xindy won't break up
the index key in the preprocessing step.

Sorry that I can't help more,

	Joachim