Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!ucsd!sdcc6!sdcc13!pa1412
From: pa1412@sdcc13.ucsd.edu (pa1412)
Newsgroups: gnu.g++.lib.bug
Subject: Re: pretesters needed
Keywords: G++ LIBG++ Configurations
Message-ID: <5078@sdcc6.ucsd.edu>
Date: 14 Nov 89 17:59:33 GMT
References: <8911121121.AA02969@g.oswego.edu>
Sender: news@sdcc6.ucsd.edu
Reply-To: pa1412@sdcc13.ucsd.edu (pa1412)
Distribution: gnu
Organization: University of California, San Diego
Lines: 31

In article <8911121121.AA02969@g.oswego.edu> dl@oswego.oswego.edu writes:
+
+So far, people have kindly volunteered to serve as libg++ pretesters
+on the following machines/systems
+
+Sequent Symmetry/Dynix
+Sun3/SunOs3.x
+Sun3/SunOs4.x
+Sun4/SunOs4.x
+Encore/Mach
+Vax/Mach
+386/MicroPort SysV.3
+Masscomp 5600/RTU
+HP9000xx/HPUX
+
+It does appear that there will be a need to create config.libg++
+scripts that create or edit .h files.

In regard to porting to the MASSCOMP RTU 4.0. The main area of
problems is in the standard I/O. In particular _IOB structure. The
fields are in different orders, have different sizes, and some
fields, _bufsiz, are absent.

I don't know of a good way to get around this. Possibly a
pre-processor tool which will extract the local stdio.h and setup
the configuration flags to enable _BUFSIZ or VPRINTF code 
--
John Clark
jclark@ucsd.edu
pa1412@iugrad2.ucsd.edu
ucsd!scubed!irt!clark