"Official" binary of Emacs-24.4 for MS-Windows

classic Classic list List threaded Threaded
37 messages Options
12
Reply | Threaded
Open this post in threaded view
|

"Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
Hello Eli and Glenn,

I've followed your advice for making an "official" binary distribution
of Emacs-24.4 for MS-Windows [1][2], and this is what I've got:

20 hours ago, I sent an email to "[hidden email]", asking for
upload rights to the GNU FTP site.  Still without answer (I don't know
how long this use to take).

I've just tried to make the binary, but after stripping temacs.exe,
this is what happens when I try to re-dump emacs.exe:

----------
Dani@LEG570 $ strip temacs.exe
Dani@LEG570 $ make
cd ../lisp; make  update-subdirs
make[1]: Entering directory `/usr/home/dani/emacs/emacs-24.4-build/lisp'
for file in `find ../../emacs-24.4/lisp -type d -print`; do  case
$file in  ../../emacs-24.4/lisp*/cedet* | ../../emacs-
24.4/lisp*/leim* ) ;;  *) wins="$wins${wins:+ }$file" ;;  esac;  done; \
for file in $wins; do \
   ../../emacs-24.4/lisp/../build-aux/update-subdirs $file; \
done;
make[1]: Leaving directory `/usr/home/dani/emacs/emacs-24.4-build/lisp'
if test "no" = "yes"; then \
  rm -f bootstrap-emacs.exe; \
  ln temacs.exe bootstrap-emacs.exe; \
else \
  ./temacs --batch --load loadup bootstrap || exit 1; \
  test "X" = X ||  -zex emacs.exe; \
  mv -f emacs.exe bootstrap-emacs.exe; \
fi
/bin/sh: line 4: ./temacs: Bad file number
Makefile:815: recipe for target `bootstrap-emacs.exe' failed
make: *** [bootstrap-emacs.exe] Error 1
Dani@LEG570 $ rm emacs.exe
Dani@LEG570 $ rm emacs-24.4.1.exe
Dani@LEG570 $ make
cd ../lisp; make  update-subdirs
make[1]: Entering directory `/usr/home/dani/emacs/emacs-24.4-build/lisp'
for file in `find ../../emacs-24.4/lisp -type d -print`; do  case
$file in  ../../emacs-24.4/lisp*/cedet* | ../../emacs-
24.4/lisp*/leim* ) ;;  *) wins="$wins${wins:+ }$file" ;;  esac;  done; \
for file in $wins; do \
   ../../emacs-24.4/lisp/../build-aux/update-subdirs $file; \
done;
make[1]: Leaving directory `/usr/home/dani/emacs/emacs-24.4-build/lisp'
if test "no" = "yes"; then \
  rm -f bootstrap-emacs.exe; \
  ln temacs.exe bootstrap-emacs.exe; \
else \
  ./temacs --batch --load loadup bootstrap || exit 1; \
  test "X" = X ||  -zex emacs.exe; \
  mv -f emacs.exe bootstrap-emacs.exe; \
fi
/bin/sh: line 4: ./temacs: Bad file number
Makefile:815: recipe for target `bootstrap-emacs.exe' failed
make: *** [bootstrap-emacs.exe] Error 1
----------

Do you know where the problem is?

TIA.

--
Dani Moncayo

---------------
[1] http://lists.gnu.org/archive/html/bug-gnu-emacs/2014-10/msg01000.html
[2] http://lists.gnu.org/archive/html/bug-gnu-emacs/2014-10/msg00995.html

Reply | Threaded
Open this post in threaded view
|

Windows build pipeline - Re: "Official" binary of Emacs-24.4 for MS-Windows

Nic Ferrier
I was going to do this as well. You're further down the line than me.

When you have a working build I'd be happy to collaborate on setting
something up to automate this.

If you need or want help let me know.


Nic

Reply | Threaded
Open this post in threaded view
|

Re: Windows build pipeline - Re: "Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
On Fri, Oct 24, 2014 at 5:17 PM, Nic Ferrier <[hidden email]> wrote:
> I was going to do this as well. You're further down the line than me.

Actually I still haven't done too much.

> When you have a working build I'd be happy to collaborate on setting
> something up to automate this.
>
> If you need or want help let me know.

Have you tried to strip "temacs.exe" and re-dump "emacs.exe"?  Does it
work on your system?

If you are able to make the binary, the way Eli said [1], feel free to
upload it to the GNU FTP site.  (but I still would like to know why it
fails on my system)

Thanks

--
Dani Moncayo

[1] http://lists.gnu.org/archive/html/bug-gnu-emacs/2014-10/msg01000.html

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Eli Zaretskii
In reply to this post by Dani Moncayo
> Date: Fri, 24 Oct 2014 17:13:41 +0200
> From: Dani Moncayo <[hidden email]>
> Cc: Emacs development discussions <[hidden email]>
>
> Dani@LEG570 $ strip temacs.exe
> Dani@LEG570 $ make
> cd ../lisp; make  update-subdirs
> make[1]: Entering directory `/usr/home/dani/emacs/emacs-24.4-build/lisp'
> for file in `find ../../emacs-24.4/lisp -type d -print`; do  case
> $file in  ../../emacs-24.4/lisp*/cedet* | ../../emacs-
> 24.4/lisp*/leim* ) ;;  *) wins="$wins${wins:+ }$file" ;;  esac;  done; \
> for file in $wins; do \
>    ../../emacs-24.4/lisp/../build-aux/update-subdirs $file; \
> done;
> make[1]: Leaving directory `/usr/home/dani/emacs/emacs-24.4-build/lisp'
> if test "no" = "yes"; then \
>   rm -f bootstrap-emacs.exe; \
>   ln temacs.exe bootstrap-emacs.exe; \
> else \
>   ./temacs --batch --load loadup bootstrap || exit 1; \
>   test "X" = X ||  -zex emacs.exe; \
>   mv -f emacs.exe bootstrap-emacs.exe; \
> fi
> /bin/sh: line 4: ./temacs: Bad file number

What happens if you run the same command by hand, i.e.

  $ ./temacs --batch --load loadup bootstrap

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
>> if test "no" = "yes"; then \
>>   rm -f bootstrap-emacs.exe; \
>>   ln temacs.exe bootstrap-emacs.exe; \
>> else \
>>   ./temacs --batch --load loadup bootstrap || exit 1; \
>>   test "X" = X ||  -zex emacs.exe; \
>>   mv -f emacs.exe bootstrap-emacs.exe; \
>> fi
>> /bin/sh: line 4: ./temacs: Bad file number
>
> What happens if you run the same command by hand, i.e.
>
>   $ ./temacs --batch --load loadup bootstrap

Dani@LEG570 $ ./temacs --batch --load loadup bootstrap
sh: ./temacs: Bad file number

--
Dani Moncayo

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Eli Zaretskii
> Date: Fri, 24 Oct 2014 17:56:58 +0200
> From: Dani Moncayo <[hidden email]>
> Cc: Emacs development discussions <[hidden email]>
>
> >> if test "no" = "yes"; then \
> >>   rm -f bootstrap-emacs.exe; \
> >>   ln temacs.exe bootstrap-emacs.exe; \
> >> else \
> >>   ./temacs --batch --load loadup bootstrap || exit 1; \
> >>   test "X" = X ||  -zex emacs.exe; \
> >>   mv -f emacs.exe bootstrap-emacs.exe; \
> >> fi
> >> /bin/sh: line 4: ./temacs: Bad file number
> >
> > What happens if you run the same command by hand, i.e.
> >
> >   $ ./temacs --batch --load loadup bootstrap
>
> Dani@LEG570 $ ./temacs --batch --load loadup bootstrap
> sh: ./temacs: Bad file number

Ouch, sorry, you are right.  I forgot: the correct commands to run
are:

  $ cd src
  $ strip temacs.tmp
  $ ../nt/addsection temacs.tmp temacs.exe EMHEAP 27
  $ cd ..
  $ make


Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
> Ouch, sorry, you are right.  I forgot: the correct commands to run
> are:
>
>   $ cd src
>   $ strip temacs.tmp
>   $ ../nt/addsection temacs.tmp temacs.exe EMHEAP 27
>   $ cd ..
>   $ make

Ok thank you.  I now have an stripped binary that seems to work fine.

A couple more questions:

To produce the binary distribution, should I do a plain "make
install"? Or perhaps:
  make install GZIP_INFO= GZIP_PROG=
?

Also, the README file in [1] was written for 24.3.  So I think it
should be renamed (README-24.3), and perhaps a new README-24.4 file
should be written (by someone knowledgeable enough, not me :-) ).

--
Dani Moncayo

[1] http://ftp.gnu.org/gnu/emacs/windows/

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Eli Zaretskii
> Date: Fri, 24 Oct 2014 20:03:51 +0200
> From: Dani Moncayo <[hidden email]>
> Cc: Emacs development discussions <[hidden email]>
>
> To produce the binary distribution, should I do a plain "make
> install"? Or perhaps:
>   make install GZIP_INFO= GZIP_PROG=
> ?

I don't see a reason not to compress files.  Emacs can display and use
them just fine, even if gzip.exe is not installed/available, right?

> Also, the README file in [1] was written for 24.3.  So I think it
> should be renamed (README-24.3), and perhaps a new README-24.4 file
> should be written (by someone knowledgeable enough, not me :-) ).

Just replace that README with README.W32 that we have in the
distribution.

Thanks.

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
>> To produce the binary distribution, should I do a plain "make
>> install"? Or perhaps:
>>   make install GZIP_INFO= GZIP_PROG=
>> ?
>
> I don't see a reason not to compress files.  Emacs can display and use
> them just fine, even if gzip.exe is not installed/available, right?

I've always used "GZIP_INFO= GZIP_PROG=".  Now I've tried a plain
"make install", and something goes wrong: if I open the installed
Emacs "runemacs -Q" and do "C-h r", the manual does not open.  The
*info* buffer is empty and the *Messages* buffer has:
  user-error: No such node or anchor: Top

I've just made a second install (from the same build), this time with
"GZIP_INFO= GZIP_PROG=".  In this installation the above problem does
not happen.

--
Dani Moncayo

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Eli Zaretskii
> Date: Fri, 24 Oct 2014 22:21:49 +0200
> From: Dani Moncayo <[hidden email]>
> Cc: Emacs development discussions <[hidden email]>
>
> >> To produce the binary distribution, should I do a plain "make
> >> install"? Or perhaps:
> >>   make install GZIP_INFO= GZIP_PROG=
> >> ?
> >
> > I don't see a reason not to compress files.  Emacs can display and use
> > them just fine, even if gzip.exe is not installed/available, right?
>
> I've always used "GZIP_INFO= GZIP_PROG=".  Now I've tried a plain
> "make install", and something goes wrong: if I open the installed
> Emacs "runemacs -Q" and do "C-h r", the manual does not open.  The
> *info* buffer is empty and the *Messages* buffer has:
>   user-error: No such node or anchor: Top
>
> I've just made a second install (from the same build), this time with
> "GZIP_INFO= GZIP_PROG=".  In this installation the above problem does
> not happen.

I don't object to not compressing the files, if compressing them
causes trouble on some systems.

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
> I don't object to not compressing the files, if compressing them
> causes trouble on some systems.

This system is Windows 8.1 64-bit, and, according to what you said:

>> I don't see a reason not to compress files.  Emacs can display and use
>> them just fine, even if gzip.exe is not installed/available, right?

not being able to show the manual (nor visiting .gz files -- I've just
tried) seemed like a bug.  It this expected (when gzip.exe is not
available) or should I report this as a bug?

Regarding the format of the binary archive, there are several options:

  Format  Archive Size
  zip     50 MB
  tar.gz  50 MB
  tar.xz  32 MB
  7z      31 MB

As you see, the last two formats produce an archive considerably
smaller.  Which format should I choose?

Regarding the name of the binary archive, the convention so far has
been "emacs-VERSION-bin-i386".  But if I eval `system-configuration'
from that Emacs I get `i686-pc-mingw32'.  So, perhaps we should name
the archive as "emacs-24.4-bin-i686", or (even better IMO),
"emacs-24.4-bin(i686-pc-mingw32)".  I like the latter format, because
it is a general and compact way of telling not only the hardware
architecture, but also OS flavor it was built for.  It is quite
self-explanatory.

But of course the choice is yours. Which filename should I choose?

And finally, note that I haven't got an answer yet to the request I
sent 2 days ago (to [hidden email]) for ftp upload rights.  So I
still can't upload anything.  (I could send the archive to anyone with
proper upload right...)

-- Dani Moncayo

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Eli Zaretskii
> Date: Sat, 25 Oct 2014 10:00:23 +0200
> From: Dani Moncayo <[hidden email]>
> Cc: Emacs development discussions <[hidden email]>
>
> > I don't object to not compressing the files, if compressing them
> > causes trouble on some systems.
>
> This system is Windows 8.1 64-bit, and, according to what you said:
>
> >> I don't see a reason not to compress files.  Emacs can display and use
> >> them just fine, even if gzip.exe is not installed/available, right?
>
> not being able to show the manual (nor visiting .gz files -- I've just
> tried) seemed like a bug.  It this expected (when gzip.exe is not
> available) or should I report this as a bug?

It's probably a missing feature, since we have zlib-decompress-region
now.  I misremembered: we still need gzip to display compressed files.

> Regarding the format of the binary archive, there are several options:
>
>   Format  Archive Size
>   zip     50 MB
>   tar.gz  50 MB
>   tar.xz  32 MB
>   7z      31 MB
>
> As you see, the last two formats produce an archive considerably
> smaller.  Which format should I choose?

IMO, you should use zip, because unpacking it does not require any
external tools to be installed on Windows.  The large size of the
archive is unfortunate, but asking users to install additional
programs is IMO worse.

> Regarding the name of the binary archive, the convention so far has
> been "emacs-VERSION-bin-i386".  But if I eval `system-configuration'
> from that Emacs I get `i686-pc-mingw32'.  So, perhaps we should name
> the archive as "emacs-24.4-bin-i686", or (even better IMO),
> "emacs-24.4-bin(i686-pc-mingw32)".  I like the latter format, because
> it is a general and compact way of telling not only the hardware
> architecture, but also OS flavor it was built for.  It is quite
> self-explanatory.
>
> But of course the choice is yours. Which filename should I choose?

I prefer emacs-24.4-i686-pc-mingw32-bin.zip.  I don't like parentheses
in the file names, because they are special to some shells.

> And finally, note that I haven't got an answer yet to the request I
> sent 2 days ago (to [hidden email]) for ftp upload rights.  So I
> still can't upload anything.  (I could send the archive to anyone with
> proper upload right...)

I hope Glenn will be able to help here, I have no experience with
uploading to GNU FTP sites.

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Glenn Morris-3
In reply to this post by Dani Moncayo
Dani Moncayo wrote:

> And finally, note that I haven't got an answer yet to the request I
> sent 2 days ago (to [hidden email]) for ftp upload rights.

It took 5 days for me (but that was over Christmas) and they said they
found my mail in their spam folder...
So in a few more days, maybe send a non-encrypted enquiry asking if they
got your first mail ok.

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
In reply to this post by Eli Zaretskii
>> Regarding the name of the binary archive, the convention so far has
>> been "emacs-VERSION-bin-i386".  But if I eval `system-configuration'
>> from that Emacs I get `i686-pc-mingw32'.  So, perhaps we should name
>> the archive as "emacs-24.4-bin-i686", or (even better IMO),
>> "emacs-24.4-bin(i686-pc-mingw32)".  I like the latter format, because
>> it is a general and compact way of telling not only the hardware
>> architecture, but also OS flavor it was built for.  It is quite
>> self-explanatory.
>>
>> But of course the choice is yours. Which filename should I choose?
>
> I prefer emacs-24.4-i686-pc-mingw32-bin.zip.  I don't like parentheses
> in the file names, because they are special to some shells.

Ok, no parentheses.  But I think that the system configuration,
since is not part of the program version (is just a build-time
setting), should should go after "-bin", not before:

   emacs-24.4-bin-i686-pc-mingw32.zip

Do you agree?

--
Dani Moncayo

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Eli Zaretskii
> Date: Sat, 25 Oct 2014 10:39:55 +0200
> From: Dani Moncayo <[hidden email]>
> Cc: Emacs development discussions <[hidden email]>
>
> > I prefer emacs-24.4-i686-pc-mingw32-bin.zip.  I don't like parentheses
> > in the file names, because they are special to some shells.
>
> Ok, no parentheses.  But I think that the system configuration,
> since is not part of the program version (is just a build-time
> setting), should should go after "-bin", not before:
>
>    emacs-24.4-bin-i686-pc-mingw32.zip
>
> Do you agree?

We are down to nitpicking, but...  Why after?

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
>> Ok, no parentheses.  But I think that the system configuration,
>> since is not part of the program version (is just a build-time
>> setting), should should go after "-bin", not before:
>>
>>    emacs-24.4-bin-i686-pc-mingw32.zip
>>
>> Do you agree?
>
> We are down to nitpicking, but...  Why after?

Because, reading the filename from left to right, it seemed more
natural to me to say first that "this is a binary distribution", and
then its system configuration than the other way around.

But yes, there isn't a compelling reason for choosing one format above
the other.

--
Dani Moncayo

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Eli Zaretskii
> Date: Sat, 25 Oct 2014 11:31:18 +0200
> From: Dani Moncayo <[hidden email]>
> Cc: Emacs development discussions <[hidden email]>
>
> >> Ok, no parentheses.  But I think that the system configuration,
> >> since is not part of the program version (is just a build-time
> >> setting), should should go after "-bin", not before:
> >>
> >>    emacs-24.4-bin-i686-pc-mingw32.zip
> >>
> >> Do you agree?
> >
> > We are down to nitpicking, but...  Why after?
>
> Because, reading the filename from left to right, it seemed more
> natural to me to say first that "this is a binary distribution", and
> then its system configuration than the other way around.

Fine, go for it.  And thanks.

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
Hi Eli,

I've tried (for the first time) the "make install-strip" target from a
trunk build, and I've noticed something strange: the size of the
resulting stripped "emacs.exe" is ~ 16 MB, whereas the same file for
the 24.4 release (also stripped) is ~ 9 MB.

Such big difference is strange, isn't it?

Also, I've noticed this suspicious errors in the output of "make
install-strip", which don't appear in a "make install" from the same
build:

------------------------------------------------------
if [ "`cd /C/msys/home/Dani/emacs/trunk/info && /bin/pwd`" =
"$exp_infodir" ]; then \
  true; \
else \
   [ -f "/c/usr1/share/info/dir" ] || \
      [ ! -f /C/msys/home/Dani/emacs/trunk/info/dir ] || \
      /bin/install -c -m 644 /C/msys/home/Dani/emacs/trunk/info/dir
"/c/usr1/share/info/dir"; \
   info_misc=`MAKELEVEL=0 make -s -C doc/misc echo-info`; \
   cd /C/msys/home/Dani/emacs/trunk/info ; \
   for elt in emacs.info eintr.info elisp.info ${info_misc}; do \
      test "yes" = "no" && test ! -f $elt && continue; \
      for f in `ls $elt $elt-[1-9] $elt-[1-9][0-9] 2>/dev/null`; do \
       (cd "${thisdir}"; \
        /bin/install -c -m 644 /C/msys/home/Dani/emacs/trunk/info/$f
"/c/usr1/share/info/$f"); \
        [ -n "" ] || continue ; \
        rm -f "/c/usr1/share/info/$f.gz"; \
         -9n "/c/usr1/share/info/$f"; \
      done; \
     (cd "${thisdir}"; \
      /bin/install-info --info-dir="/c/usr1/share/info"
"/c/usr1/share/info/$elt"); \
   done; \
fi
install-info: No such file or directory for /c/usr1/share/info/make:
install-info: No such file or directory for /c/usr1/share/info/Entering
install-info: No such file or directory for /c/usr1/share/info/directory
install-info: No such file or directory for
/c/usr1/share/info/`/home/Dani/emacs/trunk-build/doc/misc'
install-info: No such file or directory for /c/usr1/share/info/make:
install-info: No such file or directory for /c/usr1/share/info/Leaving
install-info: No such file or directory for /c/usr1/share/info/directory
install-info: No such file or directory for
/c/usr1/share/info/`/home/Dani/emacs/trunk-build/doc/misc'
Makefile:633: recipe for target `install-info' failed
make[1]: [install-info] Error 1 (ignored)
------------------------------------------------------



--
Dani Moncayo

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Eli Zaretskii
> Date: Sat, 25 Oct 2014 13:08:38 +0200
> From: Dani Moncayo <[hidden email]>
> Cc: Emacs development discussions <[hidden email]>
>
> I've tried (for the first time) the "make install-strip" target from a
> trunk build, and I've noticed something strange: the size of the
> resulting stripped "emacs.exe" is ~ 16 MB, whereas the same file for
> the 24.4 release (also stripped) is ~ 9 MB.
>
> Such big difference is strange, isn't it?

I think it's expected.  You will see the same 7MB difference between
the unstripped src/emacs.exe in the emacs-24 branch vs the trunk.

The reason is that on the trunk we now store the memory allocated
during dumping in a static array, instead of a special section of
emacs.exe.  That static array is larger than is strictly needed for an
Emacs dumped after all Lisp files were compiled to *.elc, but we keep
the larger size because it is needed during bootstrap, when the
preloaded Lisp files are loaded in their source form.  We didn't yet
find a way to reduce the size of the array when we build a
non-bootstrap Emacs.

It would be nice to find a solution to this, but it's not a bug.

> Also, I've noticed this suspicious errors in the output of "make
> install-strip", which don't appear in a "make install" from the same
> build:

Looks like some redirection snafu: for some reason, install-info is
called with arguments taken from Make messages that announce entering
and leaving directories.  Maybe it comes from this line:

>    info_misc=`MAKELEVEL=0 make -s -C doc/misc echo-info`; \

Looks like MAKELEVEL=0 doesn't work with your Make.  Try adding the
"--no-print-directory" switch to the "make" command, and see if that
helps.

This sounds like bug #13962, but since we now require GNU Make on the
trunk, I think it's OK to use --no-print-directory instead of the
MAKELEVEL=0 trick.

Reply | Threaded
Open this post in threaded view
|

Re: "Official" binary of Emacs-24.4 for MS-Windows

Dani Moncayo
>> I've tried (for the first time) the "make install-strip" target from a
>> trunk build, and I've noticed something strange: the size of the
>> resulting stripped "emacs.exe" is ~ 16 MB, whereas the same file for
>> the 24.4 release (also stripped) is ~ 9 MB.
>>
>> Such big difference is strange, isn't it?
>
> I think it's expected.  You will see the same 7MB difference between
> the unstripped src/emacs.exe in the emacs-24 branch vs the trunk.

Between unstripped executables, the difference is close to 10 MB, but
ok, this seems expected.

>> Also, I've noticed this suspicious errors in the output of "make
>> install-strip", which don't appear in a "make install" from the same
>> build:
>
> Looks like some redirection snafu: for some reason, install-info is
> called with arguments taken from Make messages that announce entering
> and leaving directories.  Maybe it comes from this line:
>
>>    info_misc=`MAKELEVEL=0 make -s -C doc/misc echo-info`; \
>
> Looks like MAKELEVEL=0 doesn't work with your Make.  Try adding the
> "--no-print-directory" switch to the "make" command, and see if that
> helps.

Yes, it helps.  The install-info errors are gone with that switch.

> This sounds like bug #13962, but since we now require GNU Make on the
> trunk, I think it's OK to use --no-print-directory instead of the
> MAKELEVEL=0 trick.

Fine with me.  Thank you.

--
Dani Moncayo

12