> actually... it has been my thought that Y2K will be just what you described...
> a 'trigger' of a kind. I just didnt want to state that for fear of offending
> the resident Keely deities with conspiracy theories. <g>
Too true. ;->
> Frank, I'm aware that UNIX boxes are set for 2038 and that they comprise
> the majority of net notes, servers, etc... my concern was regarding the PLC's
> et al. If these go...well, UNIX Y2K compliancy won't help.
That's the thing. I think you're going to find that it's going to be
isolated,
largely irrelavent systems that are going to be impacted (Financial
stuff is,
sadly, largely irrelevent compared to communications, power, etc.).
> > The chips issue is nowhere as bad as the pundits for Y2K claim it is.
>
> well, the question is...which pundits do you believe? The ones that are in
> our industry trying to calm everyone down and make a ton of money off stirring
> the pot of fear just enough to make people pay or the frothing at the mouth
> militant survivalist who sees conspiracies everywhere? Personally I have little
> faith inthe claims of either.
Neither. I think we'll see problems, to be sure, but just not what the
doom
and gloom sayers claim either. Yes, it's a domino effect, waiting to
happen-
but it's not going to play out the way they're claiming it will.
> Reasonable self reliancy IMHO is never a bad thing.
If you knew me personally, you'd know that I've always kept that to
heart and
always plan my life around that simple concept. Right now, I lack a few
things
to be self-reliant no matter what goes down- which are to be obtained
over the
next year from now. If I or someone else find over-unity and
super-luminal
communications, some of the items will no longer be needed and may
render all
of these conversations as moot.
> which reality would that be Frank? to my limited, inferior knowledge, I was under
> the general understanding that when it came to programmable logic controllers
> there was hardly any sort of glaobal standard in place for operation or design.
Doesn't need to be a standard for design. Most of them do not need a
year to operate.
A large number of them don't even need the time of day. I know what I'm
talking about
here- been in the industry for 8+ years and been playing with this stuff
as a hobby for
17+ years (including the 8 for the professional stuff).
> At least for many of the now aged plc's that were fab'd and implemented between
> the 70's and mid 80's? Who knows what conventions are used in each situation
> in places such as Eastern Europe.
PLCs are generally simple devices. They are no more complex than the
Sinclair
Radionics ZX-80 in design and programming- recent devices have things
like 486's
and Pentiums on them, but they're used for rather sophisticated things
like these
fancy-schmancy ATMs with the pretty pictures on the screens. Largely,
you're
going to find that most of the PLCs are going to be of the complexity of
a ZX80
because it's dirt-cheap to make these days. Cost is the _main_ issue in
PLCs.
> Any one who claims to have 100% solid knowledge about all the different protocols
> and date related functions that are set in place IMHO is either a con man or a fool.
But in the case of what I speak of, date is largely irrelavent. Anybody
that USED
dates in the context of most PLCs is a poor programmer or a fool for
making the code
largely more complex than it _ever_ needed to be. The problem with
dates is with
systems that deal in dates- mostly database and financial systems.
> Why else would organizations like the Aussie govt be sweating so much publicly
> JUST to get essential services in place?
Because they don't know the stuff in question and they're freaking out
like the
rest of the world will be come January 1, 2000. Y2K is the new
apocalypse theory.
> well heres the thing... you say don't 'normally' go off. even if we happened to
> have lucked out and the whole world had decided 15 to 20 years ago that they were all
> going to more or less use the same design specs for plc's (which I am convinced
> pretty much, this was not the case), and if there was only a very small fraction
> of these devices that WERE sensitive to a year date and some of them happened to be
> used in critical places susch as plants, factories, utilities.... you'd only
> need a few small ones going screwy to cause one big monstrous effect.
That's just it. Dates are needed for things like VCR clocks to start
and stop
recording. Dates _aren't_ needed to monitor or control processes. Like
I said
before, only a fool would complicate their designs this way and the
people doing
that sort of thing quickly get darwinized out of the profession. You
WON'T find
things like date sensitive stuff in power plants, etc. The old stuff
_might_ have
needed it, but the resources available to the PLCs of the time
_precluded_ doing
that sort of thing as there wouldn't be enough space for both things in
the PLC.
> you obviously have MUCH more faith in the intelligence of companies and computer
> professionals.
Yeah, could be because I am a computer professional- software engineer
in fact.
Could be because your experience with all of this is derived from
_applications_
development situations and not _engineering_ or _systems_ level
development.
Surprisingly, there's a LARGE gap in all of that- I know, been on both
sides of the
coin and would rather be doing REAL software/systems engineering (Which
is what I
do for a living nowadays...) than writing applications software.
> I have never know any dev group to be able to get to a prod
> release date under budget and according to spec without having some major bugs
> shake loose. The standard thing I've seen is a systems/programmer guy or
> team promise the moon to save his butt or get a client and then end up being
> hideously late in delivering and having to scale way back on what he had promised
> in the first place.
I've NEVER promised the moon. It's damned unprofessional AND
unethical. Oftentimes
the reasons for the problems with software development reside in either
unrealistic
expectations/offerings (i.e. Some marketing droid offers something that
is pure
vaporware and the hapless developers end up scrambling to fill the
void.), subtly or
dramatically shifting specs, unrealistic deadlines that are motivated by
marketing
reasons, or pure and simple mismanagement of the project. I've seen it
all and these
are the most common reasons for why there's bugs in the software and
it's over budget
and out of spec.
> As for plc's, from what I understand many of them ARE year dependent
> since their service cycles check against the date in some cases to flag for
> system replacement or component replacement.
I know of none that have been sold in the US, UK, or Germany that had
this
sort of feature in them. Most stuff when it needs to have a service
cycle
happen to have a clock of elapsed time that runs until it goes off,
signaling
the maintenance cycle for the unit or it's parts. A clock that keeps
track of
all of the current year, day, hour information is another chip that adds
to the
cost of the unit. These chips run about $4 or so in quantities of
1000. A PC
needs one of them to set the OS up with this information. A PLC's
margin cost
would be increased by this sort of thing. Code to duplicate this
functionality
is extra bulk that eats up precious ROM space for the main
functionality.
> IF your point is that many of the recently made
> plc's are strictly minute / hourly logged...sure I'd agree.
> But old ones, from what I understand were prone to all kinds of failure based
> on year slot averages. A good case in point is power supply replacement.
> I could be wrong about all this.
A year slot type failure does not require a year to signal things- this
is merely
a function of how many hours the unit is in operation. It's very likely
that even
the oldest ones operate under this premise.
> > Citicorp supposedly has their "critical" software corrected.
>
> right, and when they say 'supposedly' we are supposed to stick our thumbs back
> in our mouths and go wandering in our pajamas back to bed cos daddy says everythings
> ok. Sorry, I don't buy it. You get a letter from Citicorp stating that they
> are 100% compliant in their critical code as you mentioned it above and then lets talk again.
Which is why I'm going to have a couple thousand US dollars socked away
in hard
currency for what might come. I think they'll have it largely working
but it
*never* hurts in the slightest to have cash in hand. (besides, I don't
like buying
things on credit- I'm a cash and carry man, myself! ;->
> agreed about the govt. and mass hysteria. My point is that mass hysteria
> occurs because people are NOT self reliant and they know it. I'm also concerned
> that entities like Keelynet may no longer be available to everyone if grid
> sections go down.
If they do go down, it won't be for long and the phone system will not
be
likely to go down. I'm sure that if there's power down here in the
Dallas
area, Keelynet will come back as at least a BBS- come hell or high
water.
(Suffice it to say, there's enough people here in Dallas to make it a
point
to keep the communication lines for our researches going- myself
included.)
> stupid companies abound everywhere. most firms are run at the top from patronage
> not competence. I know. I recently worked for one of the largest firms in America,
> a fortune 500 company thats run by a very wealthy family. I have seen first hand...
> the crap that goes down. I was allowed into some fairly high level meetings with TOP
> people involved with global IT operations. IF anything, I am convinced they are so far behind
> Y2K compliancy and have no idea how bad off they are...well...it wont be pretty. But then
> next to their other screw ups that were literally 'criminal' well.... it made Y2K look like a cake
> walk. The hilarious thing is that the main guy in this 'family' is heralded across the country
> as a visionary, american cowboy hero, rah rah sis boom bah entrepreneurial wizard in control of
> his destiny and a boon to society and american apple pie yadayada... from what I saw behind the scenes
> this guy seemed like simply a crook with no sense of ethics. But then look at how well crooks lie,
> like Clinton for instance?
Yeah, I know where you're coming from. And, the stupidity factor is
what has me
purchasing a 2kW light plant for myself. It's one of the few remaining
things I
lack.
> > It'll be from people rioting and looting in the
> > streets
>
> yup... it'll be from the ones who never took the time to think for themselves
> and just relied on what 'daddy' said.
You know, that's basically the same thinking I've got in this matter.
> Right, and Frank... I'm more than willing to acknowledge that your particular
> brand of the truth is one I should buy into.... all I ask is that you offer
> some sort of hard evidence that says there's nothing substantial to worry about.
I'll work on it. What I'm basing this all on is my personal and
professional
experience. I know a lot about computers and software (Or, at least
that's
what everyone that knows me, including my employer, accuses me of! ;->
and I know
that the Y2K problem is larger than the "there is no problem" crowd is
claiming
and it's way smaller than the "this is going to be the end of
civilization as
we know it" crowd is claiming it to be. This whole Y2K thing is just as
Jerry's
labeling it as- it's a distraction for the majority of the populace so
that they
don't focus on the really important problems that we need to face.
> Fairy tales go both ways.... they are either stories of monsters or of princes
> and princesses and fluffy wuffy bunnies and happily ever after ending tales.
> But they are both still 'fairy tales'.
My bet is that reality is somewhere in the middle. It _always_ is.
-- Frank Earl