Is Film Going Away? by Ken Rockwell

"William Graham" wrote in message
...
snip
I spent my entire life as an electronics technician fixing
electrical/physical devices, and the first thing I always did was to fully
understand any machine that I was responsible for, and that meant I had to
know exactly how it worked before I could hope to be able to fix it.
Today, I notice that the people who fix these things don't actually know
how they work. They just replace PC cards and/or chips that each house
thousands of individual circuits, and hope that the bad circuit is the one
they replace. - I used to call this process, "Shotgunning" and it was
always a very poor way to fix anything, because, even when it worked, you
didn't have the faintest idea why it worked, or what, exactly, was wrong
with the machine to begin with. IOW, you learned nothing, and learning
something was the path to becoming a better and better tech, and to making
more and more money at doing what you did.
Today we are in a situation where everyone just shotguns everything, and
no one has the faintest idea why it works or not. IOW, it is impossible to
troubleshoot to the component level anymore, and that's a scary thing to
me. It seems that it is the beginning of the end. The end being where the
machines will know more than we do, and perhaps, someday, will take over
control of the world from us.

Very good point. Suppose for example, that an edge connector had an
intermittant bad solder connection. If you replace the circuit card that
plugs into it, you might flex the connector in such a way that the
connection will be good... until it gets bumped the wrong way. Knowing how
to trace a signal and what each part of the system does to the signal can
make quality repairs (or improvements) easier. Amazing how much
troubleshooting I've done with a heat gun, a can of Insta-Chill, and a
rubber mallet!

Bringing it on topic: If you know what each chemical does, you can diagnose
processing problems better. If you pull a roll of film out of the tank and
it's _completely_ blank, odds are that you never developed it, only fixed
it. If it's blank except for the edge numbers, odds are that it wasn't
exposed and there's a camera problem.

As for the "control of the world" part, isn't that the job of those apes
we've been domesticating? Or was Charleton Heston lying to us?!

On Tue, 22 Jan 2008 10:40:24 -0500, "Michael Benveniste"
wrote:

"." wrote:
For those of you too young to remember Super-8, they were film
cartridges that held 50 feet of film. Super-8 cameras and film were
not sensitive to light:

Super-8 film was not sensitive to light?

:-) Perhaps someone was running leader through the camera.

It wasn't very sensitive to light. The standard cartridge was 40 ASA
tungsten-balanced, so at 18fps with an f/1.8 lens it could manage down
to a light value of around 8. However, there were 500 ASA cartridges
available so Super-8 could be used under domestic lighting.

I used to use standard-8 until a few years ago, but that has the
advantage that it's trivial to create by reperforating 16mm stock.
It was expensive, though: black and white worked out at about £10
per minute once development costs were added in, while colour was
over £20 per minute.

--
Matthew Winn
[If replying by mail remove the "r" from "urk"]

William Graham wrote:

Today we are in a situation where everyone just shotguns everything, and no
one has the faintest idea why it works or not. IOW, it is impossible to
troubleshoot to the component level anymore, and that's a scary thing to me.
It seems that it is the beginning of the end. The end being where the
machines will know more than we do, and perhaps, someday, will take over
control of the world from us.

I shotgun even knowing what the equipment is doing just because it's
more intuitive to see the results on the LCD (rather than pull out a
calculator so I only have to take one shot). I take a first-stab shot &
inspect it, then I know what needs to be adjusted, try another, etc.
With time, I learn more and it takes fewer stabs. The instant feedback
is a great way to learn.

On Jan 23, 12:14*am, "William Graham" wrote:
(snip)

Today we are in a situation where everyone just shotguns everything, and no
one has the faintest idea why it works or not. IOW, it is impossible to
troubleshoot to the component level anymore, and that's a scary thing to me.
It seems that it is the beginning of the end. The end being where the
machines will know more than we do, and perhaps, someday, will take over
control of the world from us.-

I have no problem with the shotgun approach to resolving a problem
with a electronic device. The twelve gage does wonders on a Mac. Just
the replacement costs are getting a bit high.

...and what do you mean "someday"? That time is here now. Mankind
would be in a sorry state if the machines All stopped working.

Of course this is my two cents worth of retoric.


Draco

On Jan 22, 7:14*pm, "William Graham" wrote:

*I spent my entire life as an electronics technician fixing
electrical/physical devices, and the first thing I always did was to fully
understand any machine that I was responsible for, and that meant I had to
know exactly how it worked before I could hope to be able to fix it. Today,
I notice that the people who fix these things don't actually know how they
work. They just replace PC cards and/or chips that each house thousands of
individual circuits, and hope that the bad circuit is the one they
replace. - I used to call this process, "Shotgunning" and it was always a
very poor way to fix anything, because, even when it worked, you didn't have
the faintest idea why it worked, or what, exactly, was wrong with the
machine to begin with. IOW, you learned nothing, and learning something was
the path to becoming a better and better tech, and to making more and more
money at doing what you did.
Today we are in a situation where everyone just shotguns everything, and no
one has the faintest idea why it works or not. IOW, it is impossible to
troubleshoot to the component level anymore, and that's a scary thing to me.
It seems that it is the beginning of the end. The end being where the
machines will know more than we do, and perhaps, someday, will take over
control of the world from us.

For a product that is in production this might be true but when
bringing up a new product for the first time you pretty much need to
understand at the component level why something is not working. It
is not so much the complexity of things that make this harder to do
today then say 20 years ago, but the size of the leads on the parts.
And if you are working with a ball grid array package life gets very
hard very fast.

Scott

"Paul Furman" wrote in message
...
William Graham wrote:

Today we are in a situation where everyone just shotguns everything, and
no one has the faintest idea why it works or not. IOW, it is impossible
to troubleshoot to the component level anymore, and that's a scary thing
to me. It seems that it is the beginning of the end. The end being where
the machines will know more than we do, and perhaps, someday, will take
over control of the world from us.

I shotgun even knowing what the equipment is doing just because it's more
intuitive to see the results on the LCD (rather than pull out a calculator
so I only have to take one shot). I take a first-stab shot & inspect it,
then I know what needs to be adjusted, try another, etc. With time, I
learn more and it takes fewer stabs. The instant feedback is a great way
to learn.

When I first went to work for IBM many years ago, they put me with a tech
specialist for the first few weeks I was on the job. When we went out on a
call, while he was inspecting the circuit diagram of the machine, I pulled a
relay and proceeded to clean it. "What are you doing?" He asked. Well, I
said, "The trouble might very well be in this relay". "And if it is, (he
said) I will never know it, because you have messed with it. - This guy
taught me some valuable lessons. He was afraid to touch a machine unless he
knew exactly what he was doing and why. When he fixed something, he knew
that he had fixed it, and exactly what was wrong with it. Or, if he didn't,
he was unsatisfied, and knew that he would have to eventually come back.
Before he actually fixed anything, he would put a scope on one side of the
failure and then the other side, and prove beyond the shadow of a doubt that
that exact spot was the cause of the failure. Or, at least, this was his
ideal, and what he strove to do. Unfortunately, with highly intermittent
problems, one can't always do this. But shotgunning was something he never
did, and the longer I worked there, the less shotgunning I did too. There is
nothing quite like the knowledge that you have really found the problem, and
would never have to come back and work on it again.

"Scott W" wrote in message
...
On Jan 22, 7:14 pm, "William Graham" wrote:

I spent my entire life as an electronics technician fixing
electrical/physical devices, and the first thing I always did was to fully
understand any machine that I was responsible for, and that meant I had to
know exactly how it worked before I could hope to be able to fix it.
Today,
I notice that the people who fix these things don't actually know how they
work. They just replace PC cards and/or chips that each house thousands of
individual circuits, and hope that the bad circuit is the one they
replace. - I used to call this process, "Shotgunning" and it was always a
very poor way to fix anything, because, even when it worked, you didn't
have
the faintest idea why it worked, or what, exactly, was wrong with the
machine to begin with. IOW, you learned nothing, and learning something
was
the path to becoming a better and better tech, and to making more and more
money at doing what you did.
Today we are in a situation where everyone just shotguns everything, and
no
one has the faintest idea why it works or not. IOW, it is impossible to
troubleshoot to the component level anymore, and that's a scary thing to
me.
It seems that it is the beginning of the end. The end being where the
machines will know more than we do, and perhaps, someday, will take over
control of the world from us.

For a product that is in production this might be true but when
bringing up a new product for the first time you pretty much need to
understand at the component level why something is not working. It
is not so much the complexity of things that make this harder to do
today then say 20 years ago, but the size of the leads on the parts.
And if you are working with a ball grid array package life gets very
hard very fast.

Scott
Yes, sometimes you have no choice but to replace a whole bunch of circuits
at one time. But even then, it is good if you can send the group back to the
factory, or to some bench tech that can acertain just what the problem is
within the array and feed that information back to the designers so they
will know what went wrong.....IOW. there should be some closure of the
accountability loop. I can remember my specialist friend stomping on a
circuit card and then throwing it in the trash. When I asked him why he
didn't send it back to the factory for rebuild, (which was the current
policy) he said. "I have been troubleshooting this highly intermittant
problem for over a month now. When I toured the factory, they put these
modules on a machine and checked them out for about 30 seconds. When they
didn't fail during that half-minute, they put them back in a new box, and
shipped them out to someone else as new parts. What I am doing now is making
sure that doesn't happen with this particular module."

William Graham wrote:
"Paul Furman" wrote in message
...
William Graham wrote:

Today we are in a situation where everyone just shotguns everything, and
no one has the faintest idea why it works or not. IOW, it is impossible
to troubleshoot to the component level anymore, and that's a scary thing
to me. It seems that it is the beginning of the end. The end being where
the machines will know more than we do, and perhaps, someday, will take
over control of the world from us.
I shotgun even knowing what the equipment is doing just because it's more
intuitive to see the results on the LCD (rather than pull out a calculator
so I only have to take one shot). I take a first-stab shot & inspect it,
then I know what needs to be adjusted, try another, etc. With time, I
learn more and it takes fewer stabs. The instant feedback is a great way
to learn.

When I first went to work for IBM many years ago, they put me with a tech
specialist for the first few weeks I was on the job. When we went out on a
call, while he was inspecting the circuit diagram of the machine, I pulled a
relay and proceeded to clean it. "What are you doing?" He asked. Well, I
said, "The trouble might very well be in this relay". "And if it is, (he
said) I will never know it, because you have messed with it. - This guy
taught me some valuable lessons. He was afraid to touch a machine unless he
knew exactly what he was doing and why. When he fixed something, he knew
that he had fixed it, and exactly what was wrong with it. Or, if he didn't,
he was unsatisfied, and knew that he would have to eventually come back.
Before he actually fixed anything, he would put a scope on one side of the
failure and then the other side, and prove beyond the shadow of a doubt that
that exact spot was the cause of the failure. Or, at least, this was his
ideal, and what he strove to do. Unfortunately, with highly intermittent
problems, one can't always do this. But shotgunning was something he never
did, and the longer I worked there, the less shotgunning I did too. There is
nothing quite like the knowledge that you have really found the problem, and
would never have to come back and work on it again.

Reboot fixes most things :-)

Paul Furman wrote:

Reboot fixes most things :-)


You didn't finish that sentence. I think you meant to say:

"Reboot fixes most things that should never have needed fixing in the
first place. But that's Microsoft for you." :-)

On Jan 22, 1:29 pm, Michael wrote:


Well yes, But why on January 21, 2008 post something about the May 31,
2005 edition of the NYT? He tells us to pick it up. OK, I'll go to my
library and dig it out, then pick it up. And Switzerland will be
processing Super 8 Kodachrome all the way into the future of.... last
month???

An interesting post but a liitle bit late, don't you think?

Michael

Akshally, Kodak just came out with a new film
for super 8. But that little detail will of course go
unnoticed in the general "digital is clean, film
is dirty" bull**** in this ng...

Oh, BTW: Ken published that article in 2005, so his
text is up to date. Of course, the OP only noticed it now,
so the problem is perhaps with the OP?