home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!spool.mu.edu!sol.ctr.columbia.edu!emory!swrinde!cs.utexas.edu!asuvax!ncar!noao!arizona!arizona.edu!arizona.edu!news
- From: d_mc@mccube.tucson.az.us (Don McCollam)
- Newsgroups: comp.sys.next.sysadmin
- Subject: Re: bash+3.0 (SUMMARY)
- Message-ID: <1992Nov7.224851.3960@arizona.edu>
- Date: 8 Nov 92 05:48:50 GMT
- Distribution: na,local
- Lines: 81
- Nntp-Posting-Host: mccube.tucson.az.us
-
- Hi, folks.
-
- In an earlier post, I said:
- >My copy of bash1.12 does not "make" straight out of the box. Before I
- try
- >to "reason why", I'd like to know if anyone else has successfully
- >installed bash1.12 since installing 3.0. Please e-mail your response.
- >Thanks.
-
- Thanks to all who responded. I now have bash up and running (correctly, I
- hope.) The most useful response (from Frederick Korz) is the bug report
- partially quoted below:
- ---BEGIN QUOTE---
- It took a few minutes of looking at a comparision between building
- bash 1.12 on a NeXTSTEP 2.1 system and a 3.0 system. Once again
- bitten by the great NeXT header file reorganization!
- Here is a copy of the report I submitted. Bash still has some bugs
- (e.g. the pwd builtin and occasional hangs) but is pretty good
- othewise.
-
- BTW: I compiled using 'gcc-2.3.1 -g -O2'. The NeXT supplied should
- work too.
-
- Enjoy
-
-
- Return-Path: <korz>
- Date: Thu, 5 Nov 92 13:52:12 -0500
- From: korz (Frederick Korz)
- To: bash-maintainers@ai.mit.edu
- Cc: korz@bach.cs.columbia.edu
- Subject: configuration bug and fix
-
- * bash version 1.12.1
-
- * NeXTcube, 68040, 48MB mem + much disk, running NeXTSTEP 3.0
-
- * In the 3.0 release of NeXTSTEP, Next reorganized the header files
- somewhat. The script support/mksysdefs looks for wait.h and
- resource.h, does not find them in the traditional places and concludes
- they do not exist. The compilation the procedes misconfigured,
- builtins/ulimits.def results in an object that needs ulimit() and the
- compilation fails when linking the final bash executable.
-
- * repeat by unpacking the 1.12 bash distribution and making
-
- * fix - the unelegant one is included below. I have not seriously
- looked at the GNU autoconf-1.2 package. Would that be better than a
- shell script that admitted pokes around a bit?
-
- --- support/mksysdefs~ Thu Jan 9 02:45:08 1992
- +++ support/mksysdefs Thu Nov 5 12:29:02 1992
- @@ -151,7 +151,7 @@
- echo "#endif /* HAVE_UNISTD_H */" >>$sysdefs
- fi
-
- -if [ -f /usr/include/sys/wait.h ]; then
- +if [ -f /usr/include/sys/wait.h -o -f /usr/include/bsd/sys/wait.h ];
- then
- echo "" >>$sysdefs
- echo "#if !defined (HAVE_WAIT_H)" >>$sysdefs
- echo "# define HAVE_WAIT_H" >>$sysdefs
- @@ -165,7 +165,7 @@
- echo "#endif /* HAVE_ALLOCA_H */" >>$sysdefs
- fi
-
- -if [ -f /usr/include/sys/resource.h ]; then
- +if [ -f /usr/include/sys/resource.h -o -f
- /usr/include/bsd/sys/resource.h ]; then
- echo "" >>$sysdefs
- echo "#if !defined (HAVE_RESOURCE)" >>$sysdefs
- echo "# define HAVE_RESOURCE" >>$sysdefs
-
- ---END QUOTE---
- This worked for me. "Your mileage may vary."
-
- My thanks to Frederick, and all others who responded to my plea for help.
-
- Don
-
- d_mc@mccube.tucson.az.us
-