porting problems with AIXv3.1 for the Risc System/6000

[Paul Pomes - UofIllinois CSO]

*Sigh*

How hard can it be to put the FQDN in the From: line?  Really folks, it
doesn't inspire confidence that IBM products can interoperate with the rest
of the world.  Now if only the DNS records for austin.ibm.com reflected
reality (or were there at all).

>From comp.unix.aix Sat Nov 24 09:46:16 1990
Path: ux1.cso.uiuc.edu!julius.cs.uiuc.edu!usc!elroy.jpl.nasa.gov!jarthur!uunet!panews.awdpa.ibm.com!kbohrer
From: kbohrer at panews (Kathy Bohrer)
Newsgroups: comp.unix.aix
Subject: porting problems with AIXv3.1 for the Risc System/6000
Message-ID: <1990Nov22.004211.26024 at panews>
Date: 22 Nov 90 00:42:11 GMT
Organization: IBM AWD  Palo Alto
Lines: 12

To answer the posting, yes some software is difficult to port to AIX.  The
reasons are usually missing structures from #include files.  Some were
relocated to other #include files or gratuitously re-named.  Moving large
portions of libc.a into the kernel hasn't helped.  Please explain how I
can find a bug in malloc when the debugger leaps into assembler?  Or how to
use my own when other system routines continue to use the kernel version?

IBM could sell a lot more RS6000 boxes by offering a straight 4.3 or 4.4 port.
AIX is needlessly complex for little gain in functionality.  System management
demands more time because it does everything differently than other UNIX
systems.  Adding a local daemon is a real education in AIX weirdness.  Rumor
has it that mach is coming for the RS6000.  If so then perhaps we can kiss
AIX goodbye soon.

/pbp
--
         Paul Pomes

UUCP: {att,iuvax,uunet}!uiucuxc!paul   Internet, BITNET: paul at uxc.cso.uiuc.edu
US Mail:  UofIllinois, CSO, 1304 W Springfield Ave, Urbana, IL  61801-2910