[ On Fri, March 10, 1995 at 08:28:00 (-0500), John F. Woods wrote: ]
> Subject: Re: Netscape. 
>
> The non-x86 NetBSDs lean on commercial UNIXes for their "ABI"s, which change
> relatively infrequently, hence make easier targets to track.  I think, though,
> that the 386 BSD world is headed for the same market fragmentation problem
> that was well understood by the commercial UNIX world by 1988.

A "better" term might be the de facto standard one:  Binary
Compatability Standard.

Like it or not, there is already a well known and accepted set of
standards for binary compatability in most of the AT&T System V world,
and it's probably a good idea to follow that one.  There exist versions
of this standard for most major variants of SysVr4 ports, such as WE32k,
mc68k, mc88k, Sparc, iAPX86, MIPS-R4k, etc.  My understanding is that
the mc88k edition is the "most" standard between vendors, according to
level of hype, though the mc68k edition is probably the "best"
standardised in practical terms (I've never had problems running
Motorola UNIX targeted binaries on most any other SysV mc68k machine
with the same level of CPU).

I think the iAPX86 version is still called the Intel Binary
Compatability Standard, or IBCS, and I think version II is the most
recent, though I'm probably out to lunch on that too!  ;-)  I do know
there's been a lot of hype and fighting about this, esp. since the SCO
folks try to keep their own camp well armoured and well separated from
the other SysVr4 (true R4! ;-) folks.

As you all probably know, SysVr4 uses an object file format called ELF
(and SysVr2-3 used COFF, and there were some variants, like XCOFF).
Note that this is only a tiny part of the BCS problem.  The big kicker,
esp. in the workstation and personal computer world, seems to be getting
all the custom driver ioctl()'s to work.  My favourite example is the
SCO UNIX console driver, of which lots of software developers are sucked
into taking advantage of, to the detriment of all other iAPX86 platforms.

Does anyone know what the state of the OSF/1 world is w.r.t. BCS'?

> It's not just that someone might make up the expense doing a NetBSD-specific
> port (and don't count on it, they're going to have to buy hardware and blow
> big bucks on salaries -- QA people have to eat, too).  From their standpoint,

QA people don't just eat, but usually know they're worth a lot of money
and thus demand it!  ;-)

-- 
						Greg A. Woods

+1 416 443-1734			VE3TCP		robohack!woods
Planix, Inc. <woods@planix.com>; UniForum Canada <woods@uniforum.ca>