From: "Brett H. Williams" <brett_williams@...>
Date: 2003-04-12T06:15:33+09:00
Subject: Re: Building Ruby on HP-UX B.11.22

On Apr 11, Daniel Berger wrote:
> "Brett H. Williams" <brett_williams@agilent.com> wrote in message news:<20030403232221.GB31723@vor-lord.ftc.agilent.com>... 
> > They don't have 3.2 for IA64 on the testdrive box?  It sure seems a lot of more
> > recent fixes in GCC have been for IA64. I'd try to get as recent as you can
> > with GCC for IA64.
> 
> The gcc compiler has since been updated to 3.2:
> 
> However, now it's failing on the curses lib.  Same result with both
> make and gmake:
> 
> >make
> ....gcc -g -O2 -Wl,-E main.o dmyext.o libruby.a -ldld -ldl -lm   -o
> miniruby
> compiling Win32API
> compiling curses
>         gcc -fPIC -g -O2  -I/house/djberg96/ruby-1.6.8
> -I/house/djberg96/ruby-1.6.8 -DHAVE_CURSES_H -DHAVE_ISENDWIN
> -DHAVE_UNGETCH -DHAVE_BEEP -DHAVE_DOUPDATE -DHAVE_FLASH
> -DHAVE_DELETELN -DHAVE_WDELETELN  -c curses.c
> curses.c: In function `window_maxy':
> curses.c:581: structure has no member named `_maxy'
> curses.c: In function `window_maxx':
> curses.c:602: structure has no member named `_maxx'
> curses.c: In function `window_begy':
> curses.c:620: structure has no member named `_begy'
> curses.c: In function `window_begx':
> curses.c:637: structure has no member named `_begx'
> *** Error exit code 1

This looks much more tractable, as it is a compile error and not a link error.
The curses lib appears to not match what is expected--a little detective work
should pinpoint the problem on that machine.  Any luck yet?