octave on DUnix 3.0

From: Kevin Maguire <kcm_at_tca1.dl.ac.uk>
Date: Tue, 9 Apr 1996 19:51:04 +0100 (BST)

Hi

Has anyone successfully compiled octave-1.1.1 on DUnix 3.0. To this
end I compiled and installed gcc/g++ and libg++, this process seems to
have gone okay. However when trying to compile octave I get warnings like:

---
DAE.cc: In function `int ddassl_f(double *, double *, double *, double
*, int *, double  *, int *)':
DAE.cc:212: warning: name lookup of `i' changed for new ANSI `for' scoping
DAE.cc:200: warning:   using obsolete binding at `i'
DAE.cc: In method `class Matrix DAE::integrate(const class
ColumnVector &, class Matrix  &)':
DAE.cc:376: warning: name lookup of `i' changed for new ANSI `for' scoping
DAE.cc:363: warning:   using obsolete binding at `i'
DAE.cc: In method `class Matrix DAE::integrate(const classColumnVector
&, class Matrix  &, const class ColumnVector &)':
DAE.cc:465: warning: name lookup of `i' changed for new ANSI `for' scoping
DAE.cc:398: warning:   using obsolete binding at `i'
---
etc. until:
---
echo making all in liboctave
making all in liboctave
cd liboctave ; gmake all
gmake[2]: Entering directory
`/usr/users/kcm/Inst/octave-1.1.1/liboctave'
c++ -c  -I. -I../src -I.. -I../liboctave -I../src -DHAVE_CONFIG_H
-fno-implicit-templates  -g -O CColVector.cc
In file included from dMatrix.h:32,
                 from mx-base.h:31,
                 from CColVector.cc:32:
mx-defs.h:59: conflicting types for `struct Complex'
/usr/local/lib/g++-include/Complex.h:8: previous declaration as `class
complex<double>'
mx-defs.h:59: conflicting types for `struct Complex'
/usr/local/lib/g++-include/Complex.h:82: previous declaration as
`class complex<double>'
Array.h: In method `class Complex DiagArray<Complex>::get(int)':
In file included from mx-base.h:32,
                 from CColVector.cc:32:
Array.h:233: return-type `Complex' is an incomplete type
Array.h:233: cannot lookup method in incomplete type `Complex'
Array.h:233: in base initialization for class `Complex'
Array.h: In method `void DiagArray<Complex>::set(const class Complex
&, int)':
...
---
libg++ passed all the 'gmake check' tests okay, gcc seems okay too
(compiles other things okay, including itself!).
On an unrelated topic is there a timer function available for
alpha/osf1 with granularity better (finer) than 1/60 sec?  clock's man
page says it reports time in usec, but only in 1/60s which is no good.
Cheers,
Kevin
         ___________________________
      _ /       Kevin Maguire       \  _
     / )|  Higher Scientific Officer |( \
    / / |   Daresbury Laboratory     | \ \
  _( (_ |    K.Maguire_at_dl.ac.uk      | _) )_
 (((\ \> \/->____<-=-=-=-=->_____<-\/ </ /)))
 (\\\\ \_/ /                      \ \_/ ////)
  \       /                        \       /
   \    _/                          \_    /
   /   /                             \   \
  /___/                               \___\

Received on Tue Apr 09 1996 - 21:32:45 NZST