y***@gmail.com
2009-05-05 22:34:36 UTC
Dear comp.soft-sys.octave:
First post! I started learning Matlab in my undergrad, and in my
graduate career, I have switched to the open-source octave very
smoothly! I love the piping.
Anyway, I am attempting to compile and make that tar.gz that is
downloaded on a local university cluster here. I do not have
administrative privileges. I do a "configure --disable-readline" and
it completes successfully. Most of my use of octave is programmatic,
so I'm not worried about the loss of readline functionality.
However! The make command fails complaining about not finding the
'makeinfo' package/command. As noted by
https://www-old.cae.wisc.edu/pipermail/bug-octave/2008-August/006468.html,
this can be rectified by installing something that provides makeinfo.
But, I'm not at liberty to do that, unfortunately. I have posted the
tail of the make log.
make[4]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5/src'
make -C ../../scripts DOCSTRINGS
make[4]: Entering directory `/home/johnson/prk17/pkg/octave-3.0.5/
scripts'
make[4]: `DOCSTRINGS' is up to date.
make[4]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5/
scripts'
making install.texi from install.txi
install.texi is unchanged
rm -f INSTALL
../../missing makeinfo -D INSTALLONLY \
--no-validate --no-headers --no-split --output INSTALL \
-I.. -I. -I./.. install.texi
WARNING: `makeinfo' is missing on your system. You should only need
it if
you modified a `.texi' or `.texinfo' file, or any other file
indirectly affecting the aspect of the manual. The spurious
call might also be the consequence of using a buggy
`make' (AIX,
DU, IRIX). You might want to install the `Texinfo' package
or
the `GNU make' package. Grab either from any GNU archive
site.
You may also need `makeinfo' if you obtained this package
from the CVS
archive because files that can be automatically generated are
not
considered source files and are not checked in to the CVS
archive.
You may also need `makeinfo' if you ran make maintainer-clean
and
removed files that can be automatically generated but that
are
normally distributed as part of the package for convenience.
make[3]: [../../INSTALL.OCTAVE] Error 1 (ignored)
mv INSTALL ../../INSTALL.OCTAVE
mv: cannot stat `INSTALL': No such file or directory
make[3]: *** [../../INSTALL.OCTAVE] Error 1
make[3]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5/doc/
interpreter'
make[2]: *** [../INSTALL.OCTAVE] Error 2
make[2]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5/doc'
make[1]: *** [INSTALL.OCTAVE] Error 2
make[1]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5'
make: *** [all] Error 2
What is confusing me is that earlier in the log, it also complains
about makeinfo being missing. Those complaints are also warnings, but
for some reason, only this last one is fatal. I'm not exactly sure
I'm 100% understanding what the makefile is doing towards the end, and
thus, I'm not sure I particularly understand why it's failing. I am
using gcc 4.0.2, the default GNU compiler package of the cluster. I
also have gcc 4.2.4 available to me, but it doesn't produce any
different results.
I can post the full make.log (make 2>&1 | tee make.log) if you wish.
If I should have posted somewhere else, tell me.
First post! I started learning Matlab in my undergrad, and in my
graduate career, I have switched to the open-source octave very
smoothly! I love the piping.
Anyway, I am attempting to compile and make that tar.gz that is
downloaded on a local university cluster here. I do not have
administrative privileges. I do a "configure --disable-readline" and
it completes successfully. Most of my use of octave is programmatic,
so I'm not worried about the loss of readline functionality.
However! The make command fails complaining about not finding the
'makeinfo' package/command. As noted by
https://www-old.cae.wisc.edu/pipermail/bug-octave/2008-August/006468.html,
this can be rectified by installing something that provides makeinfo.
But, I'm not at liberty to do that, unfortunately. I have posted the
tail of the make log.
make[4]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5/src'
make -C ../../scripts DOCSTRINGS
make[4]: Entering directory `/home/johnson/prk17/pkg/octave-3.0.5/
scripts'
make[4]: `DOCSTRINGS' is up to date.
make[4]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5/
scripts'
making install.texi from install.txi
install.texi is unchanged
rm -f INSTALL
../../missing makeinfo -D INSTALLONLY \
--no-validate --no-headers --no-split --output INSTALL \
-I.. -I. -I./.. install.texi
WARNING: `makeinfo' is missing on your system. You should only need
it if
you modified a `.texi' or `.texinfo' file, or any other file
indirectly affecting the aspect of the manual. The spurious
call might also be the consequence of using a buggy
`make' (AIX,
DU, IRIX). You might want to install the `Texinfo' package
or
the `GNU make' package. Grab either from any GNU archive
site.
You may also need `makeinfo' if you obtained this package
from the CVS
archive because files that can be automatically generated are
not
considered source files and are not checked in to the CVS
archive.
You may also need `makeinfo' if you ran make maintainer-clean
and
removed files that can be automatically generated but that
are
normally distributed as part of the package for convenience.
make[3]: [../../INSTALL.OCTAVE] Error 1 (ignored)
mv INSTALL ../../INSTALL.OCTAVE
mv: cannot stat `INSTALL': No such file or directory
make[3]: *** [../../INSTALL.OCTAVE] Error 1
make[3]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5/doc/
interpreter'
make[2]: *** [../INSTALL.OCTAVE] Error 2
make[2]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5/doc'
make[1]: *** [INSTALL.OCTAVE] Error 2
make[1]: Leaving directory `/home/johnson/prk17/pkg/octave-3.0.5'
make: *** [all] Error 2
What is confusing me is that earlier in the log, it also complains
about makeinfo being missing. Those complaints are also warnings, but
for some reason, only this last one is fatal. I'm not exactly sure
I'm 100% understanding what the makefile is doing towards the end, and
thus, I'm not sure I particularly understand why it's failing. I am
using gcc 4.0.2, the default GNU compiler package of the cluster. I
also have gcc 4.2.4 available to me, but it doesn't produce any
different results.
I can post the full make.log (make 2>&1 | tee make.log) if you wish.
If I should have posted somewhere else, tell me.