Bell Tech 386 SysVr3

[Greg Woods]

[ WARNING: flames follow. ]

I am not familiar with Bell Tech's Sys V/386, but I am VERY familiar
with ISC's 386/ix, and somewhat familiar with Microport Sys V/386.

In article <11643 at steinmetz.ge.com> you write:
>In article <465 at sp7040.UUCP> jsp at sp7040.UUCP (John Peters) writes:
>| In article <25145 at ucbvax.BERKELEY.EDU>, jsilva at cogsci.berkeley.edu (John Silva) writes:
>| > How does Bell Technologies 386 based SysVr3 stack up to SCO?  It seems to have
>| > everything, but I am really unsure as to it's reliability.  Is it stable?
>	We have a few copies, and we're unsure of its reliability, too.
>	The serial ports still seem to take a lot of CPU, and lose data
>	at higher baud rates.

Anyone who thinks Xenix is reliable has NEVER seen a truely reliable system.
I worked with >10 Xenix 2.2.1 286 systems for a period 9 months.  At
least one of them crashed every (working) day, and often they were up and
down like yo-yo's.  Many of the crashes were due to kernel bugs.  Only a
small percentage could be attributed to hardware or local software (ie.
device drivers).  Many of the bugs are still un-found.

I've only managed to crash 386/ix once.  There's a (known) bug with shl.

The Xenix serial driver cannot share interrupt vectors with more than
one port.  It will lose data at 1200 baud.

Interrupt driven ports ALWAYS take a lot of cpu.  Try a really smart
board, such as Consensys PowerPorts.

>| > Do all the utilities work as they should?  Is it really a 'complete UNIX'
>| > as they advertise?

386/ix is a superset of what AT&T ship.  I wish they had ksh, and not
sendmail though.

>| 1.  SCO may have the "look and feel" of UNIX, but it is not UNIX and is not
>                                                    ^^^^^^^^^^^^^^
>	This is just blatently untrue. Xenix is based on AT&T code. It's

Yeah, a mess of V7, SysIII, BSD4.0, and other junk; all hooked together
by a buggy, non-standard, compiler.  As far as I know, they started the
2.x.x release with the AT&T 3B5 Sys V R2.1 tape.  However, I would guess
that only a tiny portion of the AT&T kernel was used (ie. IPC and termio).
Unfortunately, not all of the utilities were taken from that tape (ie. uucp).

The real problem is that SCO have built up a tremendous support
reputation, and they are very wary of making any changes that may have
dozens of users calling up.  Take for example the choice to make the
user utilities display disk blocks as 512 byte units, even though the
2.x kernel uses 1024 byte blocks.  One support person hinted to me that
they did not want to scare people who did a df, and found half their
disk was suddenly gone!

>	not unmodified UNIX, however. The only vendor I know of for that
>	is Bell Tech. MicroPort customizes theirs, as do Plexus, INter-
>	active Systems, etc.

Everyone customizes, but it's the porting base that counts.  Even Bell
Tech. try their hand at tuning and such.

>| 	at the level of System V Release 3.  
>	This is true, however virtual memory has been added, which is
>	the most visible to most users.

I'd of thought it was invisible.  Boy, what a can of worms it is too!
Just to "advertize", 386/ix 2.0 will have direct paging.

>| 
>| 2.  System V Release 3 has major enhancements of System V Release 2 and
>| 	SCO Zenix has not come up to its level of diversity of tools.
>| 
>  Any user who needs RFS and streams will have to buy another product or
>wait until November. Xenix has a lot of added stuff, and is generally
>quite reliable and has a number of BSD enhancements, as well as cross

       ????????
I mentioned that above.

BTW:  I'd rather have a superior product now, than wait for a lesser one.

>compile from 386 to Xenix/286, 8086 and DOS. Xenix has a lot of

This can be a very valuable selling point, but 386's (with VP/ix or
Merge) make it moot.

>utilities and tools not in SVR3 and vice versa. To imply that the lacks

Such as?  From a developer's point of view (and in my opinion), SysVR3
has more, newer, better, tools than any other version.

>are all in one direction is less than the whole truth.
>
>  You have to consider the requirements and decide if you need
>unmodified V.3 and its tools, or if you want to run SysV programs.

Now, that's a stupid statement!  What's the difference?  I want both!

>Xenix/386 passes SVID with the exception of one fairly obscure system

Is it binary (ie COFF) compatible yet?  Will the 286 version ever be?
Will there ever be a "SysVR4 compatible" Xenix 286?  [ I hope not! :-) ]

>call, and will allow you to develop programs which run portably on SVR2
>and SVR3. It won't run streams and RFS yet, and the compiler produces
>code which is (generally) better than the pcc in untouched SsyV. The

Except when it generates wrong code, or none at all.  I've ported over
20 Mb of source.  Microsoft have MANY problems with "register" and the
286 makes it almost impossible to get pointer expressions correct.

>386 SVR3 compiler would not compile about 30% of the programs I have

386/ix has trouble with function pointers.  Otherwise I've found it
better by far than the Xenix 286 compiler.  I wouldn't use the Xenix 386
compiler for any amount of money (given that I had to guarrantee the end
result).  [ I was told not to by SCO support! ]

I've heard that the Microport 286 (ie pcc/286) compiler is much better,
if slower.

>tested, generating code which crashed the assembler a suite which runs
>on BSD, Ultrix, Xenix, SunOS, SVR3 on 3B2/200, etc.

I don't want to be offensive, but are you sure you know what you are
doing?  Most code that is that portable will compile first time.  At
least it does for me.  Did you do the obvious things, like fix up the
makefile, and any configure/tuning headers, etc.?

>  Being biased is fine with me, but supporting the bias with statements
>which are false, or only part of the truth seems to be pretty useless to
>the person who asked the question.
>
>  I admit that I tested V/386, 386/ix and Xenix/386 and bought Xenix. I
>did it based on reliability of the products as of 11 months ago, and I
>have seen that the later versions of Xenix are still solid.

I bought (as have my clients who could), and STRONGLY recommend, 386/ix.
I STRONGLY recommend not touching Microsoft development tools with a 10
foot pole.

>-- 
>	bill davidsen		(wedu at ge-crd.arpa)
>  {uunet | philabs | seismo}!steinmetz!crdos1!davidsen
>"Stupidity, like virtue, is its own reward" -me


[ I apologize for the length.  I want to make it absolutely clear who
said what. ]
-- 
						Greg Woods.

UUCP: utgpu!woods, utgpu!{ontmoh, ontmoh!ixpierre}!woods
VOICE: (416) 242-7572 [h]		LOCATION: Toronto, Ontario, Canada