Very interesting discussion at Christmas dinner this year.A family friend works in upper management at a local co-gen plant(utilizes fossil fuel to generate dual energy resources-in this case burns culm{coal by-product}to produce steam and electricity)This plant,which is apx. 15 years old,produces steam to 10 acres of greenhouses as well as electricity which it sells back to the utility co.s and/or stores (to an extent)I questioned if his plant had upgraded their computer systems lately.He responded that they had.I gave him a nod and a smile.Then asked what they upgraded to.He stated the control system has been upgraded to a 386 system,they had been running 8086(286)prior.(apparently not a recent upgrade)I said nothing more through dinner.Afterwards I was compelled to confront the 386 issue and again questioned what the 386 was used for.He stated the control system and ,sensing my concern, that any change in command would take from 3-30 minutes to make it through the process.(ie.increase temp.would require more culm through the crusher onto the conveyor and finally through the turbine)I don't know what solace that provides but it's apparently what Tech has told him.I only stated "Well,it's a good thing you're off for the next 2 weeks"(he's been from Canada to Portugal this year working alot of 12/7 on maintenance and repair,not tech,and had just finished a local shift{NE PA} at 6:00 A.M.Christmas day.)His response was,"yeah,I hope so.We had to fill out a form stating our location and contact information,so we're basically on call"I left it at that.As far as I know NO 386 system is Y2K compliant nor can it be made compliant.Their network server and workstation PC's are pentium based.3-30 minutes is not a contigency plan.He also stated he has assumed the task to "drive" ,if necessary,to anywhere in upper 48 to procure parts if need be(this is assuming primary delivery services are inoperable).That was as far as I wanted to take the issue on Christmas Day 1999.Again,386 systems,to my knowledge cannot be made compliant.If one of your contingency plans is to "pick up"needed parts because delivery systems are down,how can you assume parts will be available and I'm not talking bolts and washers.This facility is primed for disruption.The upside is,it is not a primary provider and is a for profit enterprise which provides surplus electricity to the local grid.My concern is if older,primary electrical generating facilities are anywhere near the state of preparedness as this 15 year old co-gen,we are in trouble.

-- Dragnet (just@the.facts), December 25, 1999

Answers

Without any spacing, your post is extremely hard to read. I just gave up.

-- .... (....@....com), December 25, 1999.

Dragnet, thanks for the information. I don't know about the 386 systems but your conversation with your friend was interesting to me.

-- Carol (glear@usa.net), December 25, 1999.

...@..com

Sorry,forgot your still spoon feeding.

-- Dragnet (just@the.facst), December 25, 1999.

There are compliant 386's. The ones I am familiar with are pC-104 type which are designed (and whose BIOS's were written) to be embedded.

-- biker (y2kbiker@hotmail.com), December 25, 1999.

Dragnet;

And a Merry Christmas to you.

I do not know of a 486 that is compliant, but if you set them back 10 years and do not care about the date well, they will still work.

Things will get worse before they get better...

-- Helium (Heliumavid@yahoo.com), December 25, 1999.

There's no reason a 386, or a 486, or even a 286, 188, 8088, 8086, or so on cannot be in a compliant *system*. Compliance proper is not an issue to those CPUs, as there's no time or date functionality in them at all. In fact, there's no *firmware* in them at all.

Compliance or noncompliance for those CPUs is determined entirely by the *external* programming (firmware and/or software).

-- Ron Schwarz (rs@clubvb.com.delete.this), December 25, 1999.

y2kbiker:thanks for the input

helium:he also stated clocks are being set to 1995.Questioned why,not a logical date.Didn't know.

-- Dragnet (just@the.facts), December 25, 1999.

Ron:

Point well taken.But,is typical BIOS on 386 compliant?

-- Dragnet (just@the.facts), December 25, 1999.

C'mon Ron,

I think that most IT people KNOW that the 'CPU' in and of itself does not have firmware and does not have any compliance issues. But what Dragnet is stating, and that is so blatantly obvious, is that the CPU is sitting on a motherboard with a non-compliant BIOS. To most folks, the 386, 486, etc. describes their entire system, not just the processor and he is assuming that most of US would get that...correct me if I am wrong Dragnet.

So exactly what version of BIOS is he running on a 386 motherboard? Non-compliant for sure without a flash upgrade....and tell me, have YOU ever flashed a 386 BIOS? is it even possible? or does the little IC chip have to be replaced to make this board compliant? HMMMM????

You are splitting hairs here, Ron, and missed the big picture.

-- LZach (lisa@texasnetworks.com), December 25, 1999.

BIOS is 'burned' into a ROM chip, not in the CPU.

-- hiding in plain (sight@edge. of no-where), December 25, 1999.

hiding

Thank you but I knew that.My question to Ron was is typical BIOS for a 386 chip compliant or can it be made(flashed)compliant?I say not.

-- Dragnet (just@the.facts), December 25, 1999.

LZach

Thank you.You obviously understand.

-- Dragnet (just@the.facts), December 25, 1999.

Dragnet,

In 10 years in the computer service/IT industry, I have NEVER seen a motherboard designed for a 386 CPU with a compliant BIOS. The "Flash" process was not available until the 486 generation of system boards. I will check my facts and get back to you on what versions of the BIOS's, i.e., Phoenix and Award, were first capable of being flashed. Short of being flashed, the ROM chip has to be replaced.

In every instance of Y2K remediation we have done that involved early generations of motherboards and CPU's, we have always replaced the equipment with the latest generation, and in some cases where TPTB were tight wads, one previous generation of components.

It sounds as if they moved a "newer" existing computer to take over the control system.

-- LZach (lisa@texasnetworks.com), December 25, 1999.

Dragnet, I'm not an IT specialist but your posts was understandable. The moron that posted directly under your thread cannot be taught to comprehend. Some excellent points were made.

-- ~~~~~ (~~~~@~~~.xcom), December 25, 1999.

There are several fixes that provide viable workarounds to non- compliant BIOSs. IBM has one which comes in several flavors that cover the vast majority of systems, for nearly all OSs, and they've made it available for free download.

Yes, I know what a BIOS is, I know what firmware is, and I know the difference between flash RAM and PROMs (as a former Phoenix licensee, I burned my share of EPROMs for systems I built and sold to various government agencies). The issue that I addressed pertained to the increasingly common question of whether a particular *CPU* was or was not compliant.

-- Ron Schwarz (rs@clubvb.com.delete.this), December 26, 1999.

LZach

Thanks again for your insight.I'm hoping that's the case.My impression was it is not.Ironically,I worked on the construction of this facility 15 years ago and there is a certain amount of pride attached to it.

-- Dragnet (just@the.facts), December 26, 1999.

Dragnet,

The 286, as well as the 386 can be made Y2K compliant. The BIOS and Reat Time Clock are the least of the issues, though. Several approaches are available to do this. Of more pressing concern is the Process Control firmware (ROM). Merely upgrading to a 386 won't fix broken ROM.

The 286 is a great microprocessor. Properly deployed with a suitable operating system (NOT DOS!) it is the chip that kicked the Embedded Revolution into orbit. The combination of its (non DOS supported)inherent power coupled with the cratering of its price (due to its use in PC's and production under license by numerous 3rd party vendors) caused Control Engineers to go CRAZY and put them everywhere.

However, each system must be checked, preferably with an Embedded Science DELTA T Probe, for compliance issues. There are no shortcuts.

-- K. Stevens (kstevens@ It's ALL going away in just five days.com), December 26, 1999.

K- You mean there is actually someone else in the world who knows about the Delta-T Probe? I've heard it is the only tool for embeddeds that will identify potential problems. Also heard that less than 100 of them exist in the U.S. Can you enlighten us????

-- Michigander (onbubble@michigan.com), December 26, 1999.

There are exactly sixty DELTA T Probes in the U.S.A. Go to www.embedded-science.com for a really informative discussion.

Back when I got out of school, we were using Xerox Data Systems process control computers to run cement plants, and steel mills. A typical system cost half a million, and had 128 Kilobytes of RAM and a fast disk of 500 Kilobytes. Notice Kilobytes. And yet we did it. Typicaly used Punched Paper Tape readers at 300 Bytes a second if we were lucky, if not, the Teletype machines included a 15 character per second paper tape reader.

The reason I bring this up is that the Process Control tasks did NOT get more difficult as time went on, but the price of the computers cratered! Think about it...less total memory than just one 3 1/2 inch floppy. When the 286 arrived in 1982, Control Engineers thought they had died and gone to Heaven! Set up non DOS, the 80286 will directly address 16 Meg of Ram (and or ROM), and because of the Personal Computer craze the hardware got dirt cheap. See, we never demanded cheap high performance Peripherals...we just didn't need them in the bad old days of Watergate. Punched Paper Tape was perfectly adequate. However, think about the implications...the Operating System almost never crashed. We DEMANDED reliable software...an outfit like Microsoft would have been regarded as a Criminal enterprise for shipping software with known bugs.

One regretable development that flowed from the dirt cheap hardware was a drive to replace people with systems. Back before 1980, Process Control computers were used to enhance operator performance, not to replace operators. The lack of trained operators who can step in and run Processes manually with essentially what is muscle memory has contributed mightily to this whole Embedded Systems mess. And try to train digital gameheads to concentrate on analog process like runnung a water or a wastewater plant...no long term attention span.

In just 5 days, it WILL get interesting.

-- K. Stevens (kstevens@ It's ALL going away in 121 hours.com), December 26, 1999.