potentially nasty interaction between R 1.8.0 and tetex

rossini@blindglobe.net (A.J. Rossini) writes:

I've been having problems building vignettes in bioconductor packages
with R-devel.  Turns out that Rdevel/share/texmf/hyperref.cfg wants
Blue and Red predefined, when only blue and red are defined (as of
rsync Rdevel, Sept 10th).  This is on a Debian unstable system (Sept
9th version).  Might not apply to all other tetex systems.  Seems to
have bitten the bioconductor build system, though.

Symptom:  


512$ R CMD build --force Biobase
* checking for file 'Biobase/DESCRIPTION' ... OK
* preparing 'Biobase':
* checking whether 'INDEX' is up-to-date ... OK
* creating vignettes ... ERROR
/usr/lib/R/bin/texi2dvi: pdflatex exited with bad status, quitting.
/usr/lib/R/bin/texi2dvi: see Bioconductor.log for errors.
Error in buildVignettes(dir = ".") : running texi2dvi on
Bioconductor.tex failed
Execution halted



Change produces: 

517$ R CMD build --force Biobase
* checking for file 'Biobase/DESCRIPTION' ... OK
* preparing 'Biobase':
* checking whether 'INDEX' is up-to-date ... OK
* creating vignettes ... OK
* removing junk files
* building 'Biobase_1.3.31.tar.gz'

Change being Blue --> blue, Red --> red in hyperref.cfg I presume? Odd
thing is that it doesn't happen with RedHat 8.0 tetex and ordinary
"make pdf". Shouldn't the hyperlinks in the manuals have the same
problem?

BTW, this uncovered another problem in that "make pdf" didn't work
when srcdir != builddir because of

$(texinputs_BASE): FORCE $(top_srcdir)/share/perl/build-help.pl

but configure makes $(top_builddir)/share/perl/build-help.pl from a
.in file

O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk)             FAX: (+45) 35327907

Peter Dalgaard BSA <p.dalgaard@biostat.ku.dk> writes:

Change being Blue --> blue, Red --> red in hyperref.cfg I presume? Odd
thing is that it doesn't happen with RedHat 8.0 tetex and ordinary
"make pdf". Shouldn't the hyperlinks in the manuals have the same
problem?

I'm not sure what is called in the case you mention.

Note that this doesn't happen if one uses texi2dvi from the command
line, either, which is how I traced it down when comparing output in
the tex log files.

(i.e. the inclusion of the RHOME/share/texmf/hyperref.cfg from "R CMD
build" and not present when using RHOME/bin/texi2dvi).

best,
-tony

A.J. Rossini     			
rossini@u.washington.edu            http://www.analytics.washington.edu/ 
Biomedical and Health Informatics   University of Washington
Biostatistics, SCHARP/HVTN          Fred Hutchinson Cancer Research Center
UW   :              FAX=206-543-3461 | moving soon to a permanent office
FHCRC: 206-667-7025 FAX=206-667-4812 | Voicemail is pretty sketchy/use Email

CONFIDENTIALITY NOTICE: This e-mail message and any attachme...{{dropped}}

On Wed, 10 Sep 2003 16:21:03 -0700,
A J Rossini (AJR) wrote:

> Peter Dalgaard BSA <p.dalgaard@biostat.ku.dk> writes:
  >> Change being Blue --> blue, Red --> red in hyperref.cfg I presume? Odd
  >> thing is that it doesn't happen with RedHat 8.0 tetex and ordinary
  >> "make pdf". Shouldn't the hyperlinks in the manuals have the same
  >> problem?

  > I'm not sure what is called in the case you mention.

  > Note that this doesn't happen if one uses texi2dvi from the command
  > line, either, which is how I traced it down when comparing output in
  > the tex log files.

  > (i.e. the inclusion of the RHOME/share/texmf/hyperref.cfg from "R CMD
  > build" and not present when using RHOME/bin/texi2dvi).

Hmm, we have this definition of hyperref.cfg since 1.5.0, i.e., quite
some time now. I'm not sure if we really should change anything if
problems are only on Debian unstable machines. I have Debian testing
and that works fine.

.f

Friedrich.Leisch@ci.tuwien.ac.at writes:

On Wed, 10 Sep 2003 16:21:03 -0700,
A J Rossini (AJR) wrote:

  > Peter Dalgaard BSA <p.dalgaard@biostat.ku.dk> writes:

  >> Change being Blue --> blue, Red --> red in hyperref.cfg I presume? Odd
  >> thing is that it doesn't happen with RedHat 8.0 tetex and ordinary
  >> "make pdf". Shouldn't the hyperlinks in the manuals have the same
  >> problem?

  > I'm not sure what is called in the case you mention.

  > Note that this doesn't happen if one uses texi2dvi from the command
  > line, either, which is how I traced it down when comparing output in
  > the tex log files.

  > (i.e. the inclusion of the RHOME/share/texmf/hyperref.cfg from "R CMD
  > build" and not present when using RHOME/bin/texi2dvi).

Hmm, we have this definition of hyperref.cfg since 1.5.0, i.e., quite
some time now. I'm not sure if we really should change anything if
problems are only on Debian unstable machines. I have Debian testing
and that works fine.

We'd probably want to take a closer look at what the issue really is,
including teTeX version numbers. It could be a permanent change in
teTeX v2+ in which case we're going to see the effect all over the
place as it gets into major Linux distros, or it could be an
installation problem with Debian-unstable which should go away. 

Then again, does "blue" and "Blue" really do different things? In a
"no-op or fix" situation, we might choose to do the change...

O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk)             FAX: (+45) 35327907

On 11 Sep 2003 10:30:39 +0200,
Peter Dalgaard BSA (PDB) wrote:

> Friedrich.Leisch@ci.tuwien.ac.at writes:
  >> >>>>> On Wed, 10 Sep 2003 16:21:03 -0700,

>> 
  >> > Peter Dalgaard BSA <p.dalgaard@biostat.ku.dk> writes:
  >> >> Change being Blue --> blue, Red --> red in hyperref.cfg I presume? Odd
  >> >> thing is that it doesn't happen with RedHat 8.0 tetex and ordinary
  >> >> "make pdf". Shouldn't the hyperlinks in the manuals have the same
  >> >> problem?
  >> 
  >> > I'm not sure what is called in the case you mention.
  >> 
  >> > Note that this doesn't happen if one uses texi2dvi from the command
  >> > line, either, which is how I traced it down when comparing output in
  >> > the tex log files.
  >> 
  >> > (i.e. the inclusion of the RHOME/share/texmf/hyperref.cfg from "R CMD
  >> > build" and not present when using RHOME/bin/texi2dvi).
  >> 
  >> Hmm, we have this definition of hyperref.cfg since 1.5.0, i.e., quite
  >> some time now. I'm not sure if we really should change anything if
  >> problems are only on Debian unstable machines. I have Debian testing
  >> and that works fine.

  > We'd probably want to take a closer look at what the issue really
  > is,

yes, definitely

  > including teTeX version numbers. It could be a permanent change in
  > teTeX v2+ in which case we're going to see the effect all over the
  > place as it gets into major Linux distros, or it could be an
  > installation problem with Debian-unstable which should go away. 

  > Then again, does "blue" and "Blue" really do different things?

probably not

  > In a
  > "no-op or fix" situation, we might choose to do the change...

I don't know ... the problem is that it is the first time we see this,
and I'd like to avoid that "blue" works for us, we put it in and then
it doesn't work on a lot of other systems. 

My system has:

galadriel:Leisch$ dpkg -l tetex*
ii  tetex-base     2.0.2-4.1      basic teTeX library files
ii  tetex-bin      2.0.2-4.2      teTeX binary files
ii  tetex-doc      2.0.2-4.1      teTeX documentation
ii  tetex-extra    2.0.2-4.1      extra teTeX library files



.f

  > We'd probably want to take a closer look at what the issue really
  > is,

yes, definitely

Yep.

My system has:

galadriel:Leisch$ dpkg -l tetex*
ii  tetex-base     2.0.2-4.1      basic teTeX library files
ii  tetex-bin      2.0.2-4.2      teTeX binary files
ii  tetex-doc      2.0.2-4.1      teTeX documentation
ii  tetex-extra    2.0.2-4.1      extra teTeX library files

Sure, that was public knowledge as we know you run testing. What we need is
some 'meta diff' over tetex, just like Peter suggested.

Tony:  Is there a changelog entry somewhere in tetex which reveals anything?

Dirk

Those are my principles, and if you don't like them... well, I have others.
                                                -- Groucho Marx

Dirk Eddelbuettel <edd@debian.org> writes:

On Thu, Sep 11, 2003 at 11:00:13AM +0200, Friedrich.Leisch@ci.tuwien.ac.at wrote:

  > We'd probably want to take a closer look at what the issue really
  > is,

yes, definitely

Yep.

My system has:

galadriel:Leisch$ dpkg -l tetex*
ii  tetex-base     2.0.2-4.1      basic teTeX library files
ii  tetex-bin      2.0.2-4.2      teTeX binary files
ii  tetex-doc      2.0.2-4.1      teTeX documentation
ii  tetex-extra    2.0.2-4.1      extra teTeX library files

Sure, that was public knowledge as we know you run testing. What we need is
some 'meta diff' over tetex, just like Peter suggested.

501$ dpkg -l tetex*
Desired=Unknown/Install/Remove/Purge/Hold
|
Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed
|/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err:
uppercase=bad)
||/ Name           Version        Description
+++-==============-==============-============================================
un  tetex          <none>         (no description available)
ii  tetex-base     2.0.2-4.2      Basic library files of teTeX
ii  tetex-bin      2.0.2-4.3      The teTeX binary files
pn  tetex-brev     <none>         (no description available)
un  tetex-dev      <none>         (no description available)
pn  tetex-doc      <none>         (no description available)
pn  tetex-eurosym  <none>         (no description available)
ii  tetex-extra    2.0.2-4.2      Additional library files of teTeX
un  tetex-french   <none>         (no description available)
pn  tetex-frogg    <none>         (no description available)
pn  tetex-frogg-do <none>         (no description available)
pn  tetex-lib      <none>         (no description available)
un  tetex-nonfree  <none>         (no description available)
pn  tetex-src      <none>         (no description available)

Tony:  Is there a changelog entry somewhere in tetex which reveals anything?

Looking over the Debian changelogs reveals nothing -- just maintainer
stuff.  

Fritz, where is "Blue" and "Red" defined in your distribution?

best,
-tony

A.J. Rossini     			
rossini@u.washington.edu            http://www.analytics.washington.edu/ 
Biomedical and Health Informatics   University of Washington
Biostatistics, SCHARP/HVTN          Fred Hutchinson Cancer Research Center
UW   :              FAX=206-543-3461 | moving soon to a permanent office
FHCRC: 206-667-7025 FAX=206-667-4812 | Voicemail is pretty sketchy/use Email

CONFIDENTIALITY NOTICE: This e-mail message and any attachme...{{dropped}}

Looking over the Debian changelogs reveals nothing -- just maintainer
stuff.  

Just to provide more (potentially useful) info, I'm having the same
problem on a Solaris 8 machine (with the same workaround on the
Blue/blue/Red/red thing) and the teTeX installed has definitely not been
changed for a long while (perhaps ever).

-J

On Thu, 11 Sep 2003 09:21:09 -0700,
A J Rossini (AJR) wrote:

> Fritz, where is "Blue" and "Red" defined in your distribution?

Hard to say ... seems like graphics/dvipsnam.def contains a couple of
name definitions, all starting with uppercase
letters. graphics/color.sty contains lowercase definitions for the
primary colors, though. So maybe we should simply try

	Red -> red
	Blue -> blue

and see whether that survives the upcoming testing cycle?

.f