Bug#815734: regression: ESS cannot be compiled

2016-03-06 Thread Dirk Eddelbuettel 

On 6 March 2016 at 13:23, Norbert Preining wrote:
| Dear all
| 
| > BINGO -- once I add that things pass with the new texinfo.tex
| 
| Good. 
| 
| I have uploaded texinfo -5 just now, should be in the repos in
| no time.
| 
| After this, please remove the old texinfo.tex copies - we had to
| fight with outdated texinfo.tex for too often around the 4.8
| times, I try to make sure that this doesn't happen again.

Easy enough for us as doc/texinfo.tex sticks out in git.
 
| For -6 I have added a recommends on texlive-fonts-recommended, but
| this is for users and not targetting package builds. So please
| add the build-dep if necessary.

Done in the ess git repo which contains debian/.

I'll upload a new ess_15.09.2-3 with these changes too.

All good then -- so thanks all.

Dirk

-- 
http://dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org

Bug#815734: regression: ESS cannot be compiled

2016-03-05 Thread Norbert Preining 
Dear all

> BINGO -- once I add that things pass with the new texinfo.tex

Good. 

I have uploaded texinfo -5 just now, should be in the repos in
no time.

After this, please remove the old texinfo.tex copies - we had to
fight with outdated texinfo.tex for too often around the 4.8
times, I try to make sure that this doesn't happen again.

For -6 I have added a recommends on texlive-fonts-recommended, but
this is for users and not targetting package builds. So please
add the build-dep if necessary.

Thanks everyone

Norbert


PREINING, Norbert   http://www.preining.info
JAIST, Japan TeX Live & Debian Developer
GPG: 0x860CDC13   fp: F7D8 A928 26E3 16A1 9FA0  ACF0 6CAC A448 860C DC13

Bug#815734: regression: ESS cannot be compiled

2016-03-05 Thread Dirk Eddelbuettel 

On 6 March 2016 at 11:15, Norbert Preining wrote:
| Hi all,
| 
| On Sat, 05 Mar 2016, Dirk Eddelbuettel wrote:
| > kpathsea: Running mktexmf ecrm1440
| 
| apt-get install texlive-fonts-recommended

BINGO -- once I add that things pass with the new texinfo.tex
 
| Since 6.0 or so in some circumstances this is necessary, due to the
| usage of T1 encoded fonts. It depends on the texi input file,
| so it is not a strong dependency.
| 
| I might add a recommends to texinfo for that, though.

FWIW for r-base itself I appear to have added the depends on
texlive-fonts-recommended in April of 2008 or almost eight years ago.  When
texlive was probably a different beast.

In any event, MANY thanks to Norbert and texinfo upstream. I couldn't have
sorted this out myself. Or at least not quickly ...

Dirk

| 
| Norbert
| 
| 
| PREINING, Norbert   http://www.preining.info
| JAIST, Japan TeX Live & Debian Developer
| GPG: 0x860CDC13   fp: F7D8 A928 26E3 16A1 9FA0  ACF0 6CAC A448 860C DC13
| 

-- 
http://dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org

Bug#815734: regression: ESS cannot be compiled

2016-03-05 Thread Norbert Preining 
Hi all,

On Sat, 05 Mar 2016, Dirk Eddelbuettel wrote:
> kpathsea: Running mktexmf ecrm1440

apt-get install texlive-fonts-recommended

Since 6.0 or so in some circumstances this is necessary, due to the
usage of T1 encoded fonts. It depends on the texi input file,
so it is not a strong dependency.

I might add a recommends to texinfo for that, though.

Norbert


PREINING, Norbert   http://www.preining.info
JAIST, Japan TeX Live & Debian Developer
GPG: 0x860CDC13   fp: F7D8 A928 26E3 16A1 9FA0  ACF0 6CAC A448 860C DC13

Bug#815734: regression: ESS cannot be compiled

2016-03-05 Thread Dirk Eddelbuettel 

On 5 March 2016 at 19:11, Martin Maechler wrote:
| Well, I'm not so much of a "real texinfo user", but
| @Dirk:  You must have messed up badly:  In your run Metafont is called
| because tex can't find fonts, and then "tears indeed".

The observation is correct, the inference is not.

This is the exact same way we build packages, so this is almost certainly
indicative of a larger, or different issue.  To put it more plainly: the
exact same setup (from my side) built all other ESS uploads over the years.

[ But underlying Debian packages change too so I might need a new font
package, on the other hand I just built 15.09.2-2 this same very a few days
ago.  And, just to make this point very explicit, I rebuilt it now. ]
 
| If I just do what Norbert suggested, placing his new  texinfo.tex  in
| my local directory  then the minimal example runs fine and (after

That is not a pristine, empty, virgin, ... build environment, which is our
requirement and meaurement standard.  [ You can (easily) simulate something
like with by installing the docker binary package and starting from an
(almost empty) Debian or Ubuntu container. It is a powerful too. ]

| 'make clean'), 'make' does also run fine with the ESS manual and all
| other such files there. and I did make sure that indeed Norbert's
| new file is used everywhere in the process:
| (./texinfo.tex Loading texinfo [version 2016-02-05.07.deb2])
| 
| So I'd say Norbert (and his team)'s changes were successful and have
| solved the regression.
| Thank you guys so much!

I disagree (strongly) as I have identical build setups differing only between

   edd@max:/tmp/ess$ grep texinfoversion{2 ess-15.09.2-?/doc/texinfo.tex
   ess-15.09.2-2/doc/texinfo.tex:\def\texinfoversion{2015-06-24.23}
   ess-15.09.2-3/doc/texinfo.tex:\def\texinfoversion{2016-02-05.07.deb2}
   edd@max:/tmp/ess$ 

and the former builds whereas the latter fails.

So, still 'nope' from here.

Dirk
 
| Martin
| 
| On Sat, Mar 5, 2016 at 4:25 PM, Dirk Eddelbuettel  wrote:
| >
| > On 5 March 2016 at 23:37, Norbert Preining wrote:
| > | Hi all (Debian only)
| > |
| > | On Sat, 05 Mar 2016, Gavin Smith wrote:
| > | > Done in SVN revision 7041. Please email if there are further problems.
| > |
| > | I have prepare a new version of texinfo package, can you please check
| > | whether the attached texinfo.tex fixes the problem for you?
| > |
| > | That is, make a built with the attached file included in the source
| > | dir instead of the old version.
| > |
| > | Please check that the log file mentions the following texinfo.tex
| > | version:
| > |   2016-02-05.07.deb2
| > |
| > | Thanks for letting me know
| >
| > Negative.
| >
| > Using the same pbuilder chroot, starting from the same ess 15.09.2-2 
package I
| > successfully prepared with the texinfo 6.0 file texinfo.tex, I now get a
| > fail once I now include the new texinfo.tex you just emailed.
| >
| > I tickle a grep for its version in debian/rules as you requested:
| >
| > dh_testdir
| > echo "*** For Norbert  ***"
| > *** For Norbert  ***
| > grep texinfoversion doc/texinfo.tex
| > \def\texinfoversion{2016-02-05.07.deb2}
| > \message{Loading texinfo [version \texinfoversion]:}
| > \everyjob{\message{[Texinfo version \texinfoversion]}%
| > @c time-stamp-start: "deftexinfoversion{"
| > echo "*** For Norbert  ***"
| > *** For Norbert  ***
| >
| > and it ends in tears a little later -- though now with a different issue.
| > Real texinfo users like Kurt or Martin may have an idea:
| >
| >
| > Chapter 11 [55] [56] [57] [58] [59] [60] Chapter 12 [61]
| > (/build/ess-15.09.2/doc/help-s.texi [62] [63] [64] [65] [66]) Chapter 13
| > [67] (/build/ess-15.09.2/doc/help-sas.texi [68] [69] [70] [71] [72] [73]
| > [74] [75]) Chapter 14 [76] (/build/ess-15.09.2/doc/help-bugs.texi) Chapter 
15
| > [77] (/build/ess-15.09.2/doc/help-jags.texi [78]) Chapter 16 [79]
| > (/build/ess-15.09.2/doc/bugs.texi) (/build/ess-15.09.2/doc/bugrept.texi)
| > (/build/ess-15.09.2/doc/mailing.texi [80]) Appendix A [81] (Indices) [82]
| > kpathsea: Running mktextfm ecrm1440
| > mkdir: cannot create directory '././nonexistent': Permission denied
| > mktextfm: Running mf-nowin -progname=mf \mode:=ljfour; mag:=1; nonstopmode; 
input ecrm1440
| > This is METAFONT, Version 2.7182818 (TeX Live 2015/Debian) (preloaded 
base=mf)
| >
| > kpathsea: Running mktexmf ecrm1440
| >
| > ! I can't find file `ecrm1440'.
| > <*> ...ljfour; mag:=1; nonstopmode; input ecrm1440
| >
| > Please type another input file name
| > ! Emergency stop.
| > <*> ...ljfour; mag:=1; nonstopmode; input ecrm1440
| >
| > Transcript written on mfput.log.
| > grep: ecrm1440.log: No such file or directory
| > mktextfm: `mf-nowin -progname=mf \mode:=ljfour; mag:=1; nonstopmode; input 
ecrm1440' failed to make ecrm1440.tfm.
| > kpathsea: Appending font creation commands to missfont.log.
| >
| > ./ess.texi:3643: Font \thisecfont=ecrm1440 at 14.0pt not loadable: Metric 
(TFM)
| >  file not found.
| > 
| >

Bug#815734: regression: ESS cannot be compiled

2016-03-05 Thread Martin Maechler 
Well, I'm not so much of a "real texinfo user", but
@Dirk:  You must have messed up badly:  In your run Metafont is called
because tex can't find fonts, and then "tears indeed".

If I just do what Norbert suggested, placing his new  texinfo.tex  in
my local directory  then the minimal example runs fine and (after
'make clean'), 'make' does also run fine with the ESS manual and all
other such files there. and I did make sure that indeed Norbert's
new file is used everywhere in the process:
(./texinfo.tex Loading texinfo [version 2016-02-05.07.deb2])

So I'd say Norbert (and his team)'s changes were successful and have
solved the regression.
Thank you guys so much!

Martin

On Sat, Mar 5, 2016 at 4:25 PM, Dirk Eddelbuettel  wrote:
>
> On 5 March 2016 at 23:37, Norbert Preining wrote:
> | Hi all (Debian only)
> |
> | On Sat, 05 Mar 2016, Gavin Smith wrote:
> | > Done in SVN revision 7041. Please email if there are further problems.
> |
> | I have prepare a new version of texinfo package, can you please check
> | whether the attached texinfo.tex fixes the problem for you?
> |
> | That is, make a built with the attached file included in the source
> | dir instead of the old version.
> |
> | Please check that the log file mentions the following texinfo.tex
> | version:
> |   2016-02-05.07.deb2
> |
> | Thanks for letting me know
>
> Negative.
>
> Using the same pbuilder chroot, starting from the same ess 15.09.2-2 package I
> successfully prepared with the texinfo 6.0 file texinfo.tex, I now get a
> fail once I now include the new texinfo.tex you just emailed.
>
> I tickle a grep for its version in debian/rules as you requested:
>
> dh_testdir
> echo "*** For Norbert  ***"
> *** For Norbert  ***
> grep texinfoversion doc/texinfo.tex
> \def\texinfoversion{2016-02-05.07.deb2}
> \message{Loading texinfo [version \texinfoversion]:}
> \everyjob{\message{[Texinfo version \texinfoversion]}%
> @c time-stamp-start: "deftexinfoversion{"
> echo "*** For Norbert  ***"
> *** For Norbert  ***
>
> and it ends in tears a little later -- though now with a different issue.
> Real texinfo users like Kurt or Martin may have an idea:
>
>
> Chapter 11 [55] [56] [57] [58] [59] [60] Chapter 12 [61]
> (/build/ess-15.09.2/doc/help-s.texi [62] [63] [64] [65] [66]) Chapter 13
> [67] (/build/ess-15.09.2/doc/help-sas.texi [68] [69] [70] [71] [72] [73]
> [74] [75]) Chapter 14 [76] (/build/ess-15.09.2/doc/help-bugs.texi) Chapter 15
> [77] (/build/ess-15.09.2/doc/help-jags.texi [78]) Chapter 16 [79]
> (/build/ess-15.09.2/doc/bugs.texi) (/build/ess-15.09.2/doc/bugrept.texi)
> (/build/ess-15.09.2/doc/mailing.texi [80]) Appendix A [81] (Indices) [82]
> kpathsea: Running mktextfm ecrm1440
> mkdir: cannot create directory '././nonexistent': Permission denied
> mktextfm: Running mf-nowin -progname=mf \mode:=ljfour; mag:=1; nonstopmode; 
> input ecrm1440
> This is METAFONT, Version 2.7182818 (TeX Live 2015/Debian) (preloaded base=mf)
>
> kpathsea: Running mktexmf ecrm1440
>
> ! I can't find file `ecrm1440'.
> <*> ...ljfour; mag:=1; nonstopmode; input ecrm1440
>
> Please type another input file name
> ! Emergency stop.
> <*> ...ljfour; mag:=1; nonstopmode; input ecrm1440
>
> Transcript written on mfput.log.
> grep: ecrm1440.log: No such file or directory
> mktextfm: `mf-nowin -progname=mf \mode:=ljfour; mag:=1; nonstopmode; input 
> ecrm1440' failed to make ecrm1440.tfm.
> kpathsea: Appending font creation commands to missfont.log.
>
> ./ess.texi:3643: Font \thisecfont=ecrm1440 at 14.0pt not loadable: Metric 
> (TFM)
>  file not found.
> 
>\thisecfont
> \indexlbrace ->{\ifmonospace \else \ecfont
>\fi \char 123}
>  ... \kern -0.05em \secbf {\indexlbrace
>   }
> \leftline #1->\line {#1
>\hss }
> \initialx ...e {\secfonts \kern -0.05em \secbf #1}
>   \nobreak \vskip 
> .33\baseli...
>
> \thisline ->\initial {{\indexlbrace }}
>
> ...
> l.3643 @printindex ky
>
> ?
> ./ess.texi:3643: Emergency stop.
> 
>\thisecfont
> \indexlbrace ->{\ifmonospace \else \ecfont
>\fi \char 123}
>  ... \kern -0.05em \secbf {\indexlbrace
>   }
> \leftline #1->\line {#1
>\hss }
> \initialx ...e {\secfonts \kern -0.05em \secbf #1}
>   \nobreak \vskip 
> .33\baseli...
>
> \thisline ->\initial {{\indexlbrace }}
>
> ...
> l.3643 @printindex ky
>
> ./ess.texi:3643:  ==> Fatal error occurred, no output PDF file produced!
> Transcript written on ess.log.
> /usr/bin/texi2dvi: pdfetex exited with bad status, quitting.
> Makefile:43: recipe for target 'ess.pdf' failed
> make[1]: *** [ess.pdf] Error 1
> make[1]: Leaving directory '/build/ess-15.09.2/doc'
> debian/rules:35: recipe for target

Bug#815734: regression: ESS cannot be compiled

2016-03-05 Thread Dirk Eddelbuettel 

On 5 March 2016 at 23:37, Norbert Preining wrote:
| Hi all (Debian only)
| 
| On Sat, 05 Mar 2016, Gavin Smith wrote:
| > Done in SVN revision 7041. Please email if there are further problems.
| 
| I have prepare a new version of texinfo package, can you please check
| whether the attached texinfo.tex fixes the problem for you?
| 
| That is, make a built with the attached file included in the source
| dir instead of the old version.
| 
| Please check that the log file mentions the following texinfo.tex
| version:
|   2016-02-05.07.deb2
| 
| Thanks for letting me know

Negative.

Using the same pbuilder chroot, starting from the same ess 15.09.2-2 package I
successfully prepared with the texinfo 6.0 file texinfo.tex, I now get a
fail once I now include the new texinfo.tex you just emailed.

I tickle a grep for its version in debian/rules as you requested:

dh_testdir
echo "*** For Norbert  ***"
*** For Norbert  ***
grep texinfoversion doc/texinfo.tex
\def\texinfoversion{2016-02-05.07.deb2}
\message{Loading texinfo [version \texinfoversion]:}
\everyjob{\message{[Texinfo version \texinfoversion]}%
@c time-stamp-start: "deftexinfoversion{"
echo "*** For Norbert  ***"
*** For Norbert  ***

and it ends in tears a little later -- though now with a different issue.
Real texinfo users like Kurt or Martin may have an idea:


Chapter 11 [55] [56] [57] [58] [59] [60] Chapter 12 [61]
(/build/ess-15.09.2/doc/help-s.texi [62] [63] [64] [65] [66]) Chapter 13
[67] (/build/ess-15.09.2/doc/help-sas.texi [68] [69] [70] [71] [72] [73]
[74] [75]) Chapter 14 [76] (/build/ess-15.09.2/doc/help-bugs.texi) Chapter 15
[77] (/build/ess-15.09.2/doc/help-jags.texi [78]) Chapter 16 [79]
(/build/ess-15.09.2/doc/bugs.texi) (/build/ess-15.09.2/doc/bugrept.texi)
(/build/ess-15.09.2/doc/mailing.texi [80]) Appendix A [81] (Indices) [82]
kpathsea: Running mktextfm ecrm1440
mkdir: cannot create directory '././nonexistent': Permission denied
mktextfm: Running mf-nowin -progname=mf \mode:=ljfour; mag:=1; nonstopmode; 
input ecrm1440
This is METAFONT, Version 2.7182818 (TeX Live 2015/Debian) (preloaded base=mf)

kpathsea: Running mktexmf ecrm1440

! I can't find file `ecrm1440'.
<*> ...ljfour; mag:=1; nonstopmode; input ecrm1440
  
Please type another input file name
! Emergency stop.
<*> ...ljfour; mag:=1; nonstopmode; input ecrm1440
  
Transcript written on mfput.log.
grep: ecrm1440.log: No such file or directory
mktextfm: `mf-nowin -progname=mf \mode:=ljfour; mag:=1; nonstopmode; input 
ecrm1440' failed to make ecrm1440.tfm.
kpathsea: Appending font creation commands to missfont.log.

./ess.texi:3643: Font \thisecfont=ecrm1440 at 14.0pt not loadable: Metric (TFM)
 file not found.
 
   \thisecfont 
\indexlbrace ->{\ifmonospace \else \ecfont 
   \fi \char 123}
 ... \kern -0.05em \secbf {\indexlbrace 
  }
\leftline #1->\line {#1
   \hss }
\initialx ...e {\secfonts \kern -0.05em \secbf #1}
  \nobreak \vskip .33\baseli...

\thisline ->\initial {{\indexlbrace }}
   
...
l.3643 @printindex ky
 
? 
./ess.texi:3643: Emergency stop.
 
   \thisecfont 
\indexlbrace ->{\ifmonospace \else \ecfont 
   \fi \char 123}
 ... \kern -0.05em \secbf {\indexlbrace 
  }
\leftline #1->\line {#1
   \hss }
\initialx ...e {\secfonts \kern -0.05em \secbf #1}
  \nobreak \vskip .33\baseli...

\thisline ->\initial {{\indexlbrace }}
   
...
l.3643 @printindex ky
 
./ess.texi:3643:  ==> Fatal error occurred, no output PDF file produced!
Transcript written on ess.log.
/usr/bin/texi2dvi: pdfetex exited with bad status, quitting.
Makefile:43: recipe for target 'ess.pdf' failed
make[1]: *** [ess.pdf] Error 1
make[1]: Leaving directory '/build/ess-15.09.2/doc'
debian/rules:35: recipe for target 'build-stamp' failed
make: *** [build-stamp] Error 2



I put the complete build-log here if anybody wants to take a closer look:

  http://dirk.eddelbuettel.com/tmp/texinfo-regression/ess_15.09.2-3.log

Dirk

-- 
http://dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org

Bug#815734: regression: ESS cannot be compiled

2016-03-05 Thread Norbert Preining 
Thanks Gavin,

> Done in SVN revision 7041. Please email if there are further problems.

I have prepared a new version for Debian and asked the ess team
to see if that works out. We will surely come back if there
are any problems.

Thanks a big big many times for your great support!!!

Norbert


PREINING, Norbert   http://www.preining.info
JAIST, Japan TeX Live & Debian Developer
GPG: 0x860CDC13   fp: F7D8 A928 26E3 16A1 9FA0  ACF0 6CAC A448 860C DC13

Bug#815734: regression: ESS cannot be compiled

2016-03-05 Thread Gavin Smith 
On 4 March 2016 at 21:28, Gavin Smith  wrote:
> On 4 March 2016 at 04:01, Norbert Preining  wrote:
>> The output is:
>> ./bla.texi:9: Argument of @gobblespaces has an extra }.
>
> I got the same. I think it's easy enough to fix. The only possible
> problem is that this error comes about due to an attempt to remove
> whitespace from the start of a macro argument when the macro only
> takes one argument. To fix, this would be removed, so any whitespace
> after the opening brace of the macro invocation wouldn't be ignored. I
> think this is okay, because people will only expect spaces following
> commas separating macro arguments to be discarded.

Done in SVN revision 7041. Please email if there are further problems.

Bug#815734: regression: ESS cannot be compiled

2016-03-04 Thread Norbert Preining 
Dear all,

down here at Debian we got another bug report about a FTBFS, which
went away when switching to an older version of texinfo.tex:

On Thu, 03 Mar 2016, Dirk Eddelbuettel wrote:
> Ok, the obvious fix of fattening the source directory with a working copy of
> texinfo.tex did the trick.

I have reduced the problematic case to the following minimal example:

\input texinfo   @c -*-texinfo-*-
@setfilename ess.info
@settitle ESS -- Emacs Speaks Statistics

@macro ESS{ARG}
@acronym{ESS\ARG\}
@end macro

@ESS{}: bad

@ESS{[R]}: good

@bye

The output is:
./bla.texi:9: Argument of @gobblespaces has an extra }.
 
@par 
 
   }
@doacronym ...#3@finish ->{@selectfonts @lsize #1}
  @def @temp {#2}@ifx @temp ...
l.1 @acronym{ESS@gobblespaces }
   @texinfoc
@scanmacro ...atspaces }@scantokens {#1@texinfoc }
  @aftermacro 
l.9 @ESS{}
  : bad
? 

Can anyone comment on this?

Thanks a lot

Norbert


PREINING, Norbert   http://www.preining.info
JAIST, Japan TeX Live & Debian Developer
GPG: 0x860CDC13   fp: F7D8 A928 26E3 16A1 9FA0  ACF0 6CAC A448 860C DC13

Bug#815734: regression: ESS cannot be compiled

2016-03-04 Thread Dirk Eddelbuettel 

Just a quick 'thumbs up' from here -- it is actually quite remarkable that we
do maintain buildability of these pieces over so many years and decaded, and
I really appreciate how we're going calmly about the business of maintaining
that tradition.

Thanks to Norbert for distilling that minimal example out of the larger ESS
documentation, and to the texinfo folks for looking at it.

Cheers, Dirk

-- 
http://dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org

Bug#815734: regression: ESS cannot be compiled

2016-03-04 Thread Martin Maechler 
On Fri, Mar 4, 2016 at 10:38 PM, Gavin Smith  wrote:
>> > I have reduced the problematic case to the following minimal example:
>
> Thank you for taking the time to do this, Norbert.
>
>> From these two facts I'd declare that the new texinfo.tex that
>> you have in Debian is broken.
>> (Yes, that is a strong statement; but still breaking something
>>  that worked dozens of year w/o problems needs VERY good arguments
>>  before you can call it "progress" !!)
>
> No-one's said that breaking this usage is progress. It wasn't intended
> and we will fix it.
>
> The problem comes from trying to support the Texinfo macro facility,
> the different ways that it's been used in different documents and
> managing multiple incompatible implementations of it. It's caused
> problems ever since it was introduced, maybe about 20 years ago.
> Evidently these problems aren't at an end yet, despite hope. When the
> fix is finalized I'll add this breaking usage to our test file to
> prevent it breaking again.

Thanks a lot, Gavin!  The above sounds very good to me,
I'm sorry for my former misunderstanding..

Best regards,  and thank you for all the efforts to make info /
texinfo even more 'future-proof'!

Martin

Bug#815734: regression: ESS cannot be compiled

2016-03-04 Thread Gavin Smith 
> > I have reduced the problematic case to the following minimal example:

Thank you for taking the time to do this, Norbert.

> From these two facts I'd declare that the new texinfo.tex that
> you have in Debian is broken.
> (Yes, that is a strong statement; but still breaking something
>  that worked dozens of year w/o problems needs VERY good arguments
>  before you can call it "progress" !!)

No-one's said that breaking this usage is progress. It wasn't intended
and we will fix it.

The problem comes from trying to support the Texinfo macro facility,
the different ways that it's been used in different documents and
managing multiple incompatible implementations of it. It's caused
problems ever since it was introduced, maybe about 20 years ago.
Evidently these problems aren't at an end yet, despite hope. When the
fix is finalized I'll add this breaking usage to our test file to
prevent it breaking again.

Bug#815734: regression: ESS cannot be compiled

2016-03-04 Thread Gavin Smith 
On 4 March 2016 at 04:01, Norbert Preining  wrote:
> The output is:
> ./bla.texi:9: Argument of @gobblespaces has an extra }.

I got the same. I think it's easy enough to fix. The only possible
problem is that this error comes about due to an attempt to remove
whitespace from the start of a macro argument when the macro only
takes one argument. To fix, this would be removed, so any whitespace
after the opening brace of the macro invocation wouldn't be ignored. I
think this is okay, because people will only expect spaces following
commas separating macro arguments to be discarded. In the past, spaces
were trimmed at the beginning and end of macro arguments regardless of
the number of arguments they took, but this broke when the macro
argument had newlines which had to be significant (as far as I
remember).

Bug#815734: regression: ESS cannot be compiled

2016-03-04 Thread Martin Maechler 

> Norbert Preining 
> on Fri, 4 Mar 2016 13:01:23 +0900 writes:

> Dear all,
> down here at Debian we got another bug report about a FTBFS, which
> went away when switching to an older version of texinfo.tex:

> On Thu, 03 Mar 2016, Dirk Eddelbuettel wrote:
>> Ok, the obvious fix of fattening the source directory with a working 
copy of
>> texinfo.tex did the trick.

> I have reduced the problematic case to the following minimal example:

> \input texinfo   @c -*-texinfo-*-
> @setfilename ess.info
> @settitle ESS -- Emacs Speaks Statistics

> @macro ESS{ARG}
> @acronym{ESS\ARG\}
> @end macro

> @ESS{}: bad

> @ESS{[R]}: good

> @bye

> The output is:
> ./bla.texi:9: Argument of @gobblespaces has an extra }.
>  
> @par 
>  
> }
> @doacronym ...#3@finish ->{@selectfonts @lsize #1}
> @def @temp {#2}@ifx @temp ...
> l.1 @acronym{ESS@gobblespaces }
> @texinfoc
> @scanmacro ...atspaces }@scantokens {#1@texinfoc }
> @aftermacro 
> l.9 @ESS{}
> : bad
> ? 

> Can anyone comment on this?

Yes:

- This macro use has always worked in the ESS sources for
  around 20 years maybe.
- It still works for me (with a TeX Live 2015 version of texinfo.tex
  that declares  'Loading texinfo [version 2015-07-01.07]'

>From these two facts I'd declare that the new texinfo.tex that
you have in Debian is broken.
(Yes, that is a strong statement; but still breaking something
 that worked dozens of year w/o problems needs VERY good arguments
 before you can call it "progress" !!)
 
Martin



ess-tst.texi
Description: Binary data