building octave 2.0.14 comments

[Himanshu Gohel]

So I'm rebuilding octave 2.0.14 on my SGI IRIX 6.5.2f system,
and I notice the sequence going something like this:

c++ -c  -I. -I.. -I../liboctave -I../src -I../libcruft/misc  -I../glob 
-I../glob -DHAVE_CONFIG_H -fno-rtti -fno-exceptions -fno-implicit-templates 
-g -O2 -Wall ov-re-mat.cc
ar rc  liboctinterp.a ov-re-mat.o
c++ -c  -I. -I.. -I../liboctave -I../src -I../libcruft/misc  -I../glob 
-I../glob -DHAVE_CONFIG_H -fno-rtti -fno-exceptions -fno-implicit-templates 
-g -O2 -Wall ov-cx-mat.cc
ar rc  liboctinterp.a ov-cx-mat.o
c++ -c  -I. -I.. -I../liboctave -I../src -I../libcruft/misc  -I../glob 
-I../glob -DHAVE_CONFIG_H -fno-rtti -fno-exceptions -fno-implicit-templates 
-g -O2 -Wall ov-range.cc
ar rc...

so it's compile-archive-compile-archive-...  Just wondering,
would it not be better to call 'ar' just once at the end of
compilation of all the objects instead of invoking ar about
a hundred (or more) times for each library archive?

The build has been running on this O2/R1000, 150MHz/64MB for
just about 3 hours now, and it's not done yet...still waiting...

--
Himanshu Gohel, address@hidden
WEB URL: http://www.csee.usf.edu/~gohel/
University of South Florida, Tampa, FL.  USA.