USB microscopes for very small SMT

[Spehro Pefhany]

Joel Koltner wrote:

[...]

Software bloat doesn't bother me nearly as much as the fall in software
quality (from the bug-free perspective)...

Even if the programmers exhibited the same level of diligence today as
they did in the 80's, a code that is bloated 100-1000 times in size is
bound to have a huge increase in the number of bugs.

They'd be bankrupt.

Well, when people like me start looking and shopping for older versions
of Acrobat or Works or whatever that would disturb me if I was a manager
there.

Take a look at Vista: IMHO a blatant slap in the face for the designers.
Clients of mine only buy PCs when they come with XP or the right to
downgrade. Until a few minutes ago (when I placed the order) I was in
the market for a netbook -> All XP. Else I wouldn't have bought one.

www.oldapps.com or
http://www.oldversion.com for "light" programs that won't bog down
your netbook.

See, for example, ACDSee


ACDSee 1.0 for Windows 95 (0.5 MB)
ACDSee 2.1 (0.6 MB)
ACDSee 2.22 (0.9 MB)
ACDSee 2.4 (1.9 MB)
ACDSee 2.42 (1.9 MB)
ACDSee 2.43 (1.6 MB)
ACDSee 2.45 (German) (2.0 MB)
ACDSee 3.0 (3.9 MB)
ACDSee 3.10 (5.7 MB)
ACDSee 4.01 (11.4 MB)
ACDSee 4.02 (10.8 MB)
ACDSee 5.0.1 (11.1 MB)
ACDSee 7.0 (14.8 MB)
ACDSee 8.0 (12.1 MB)
ACDSee 8.1 (24.2 MB)

V2.22 (0.9 MB) will work with most non-raw photos.


Best regards,
Spehro Pefhany

 

[krw]

I annoy kids when I come up with numbers in a couple of seconds, while
they're still fiddling with their Blackberry.


We still do a lot of arithmetic in our heads, at meetings and such.
But it's not classic adding/subtraction of digits, it's more like a
mental slide rule, rough "analog computing" estimates or rounding to
cardinal points (like 3/8 = 0.375) and then tossing in an estimated
interpolation to get a little closer. There are times when a 20%
accurate estimate is still very valuable.

The order of magnitude is usually good enough in such
circumstances. One significant digit is almost always enough.
That sort of "accuracy" was what I lost by using calculators as a
crutch. Annoying but I wasn't a bad trade, IMO.

But yeah, I'm not sure if I can still do long division, and even
subtraction seems silly by hand.

I never worried too much about multiplication and division beyond
one digit. That's what the slip stick was for. One digit was
enough to verify the process, something now missing with calculator
(or spice) results.

 

[krw]

Sure, the problem always never lies with the kids but their mentors.
IMHO parents should take a pivotal role here and not dump all that
responsibility on a school.

Schools don't get a pass. Schools are often better equip ed than
parents and often have more face time. Schools are charged with
teaching facts rather than the squishy subjects they're so fond of.
Face it, they fail.

Yep. This was meant to explain how technology makes us complacent. I've
recently re-learned the way our great-great-grandparents made break,
over a wood fire. Not by turning a knob on the oven and later shoving
the dough into it.

If that's what you enjoy doing, so be it. There are so many things
previous generations had to do that are no longer required that
there isn't time to master them all. I don't have time to do the
"modern" things I'd like to do.

It's good to grab a slide rule once in a while. I recently upgraded to a
round Scientific Instruments No.250, from a friend who passed away.
Unfortunately the hot summers out here have shrunk its naugahide pouch
and it no longer fits into it.

I take out my Post VersaLog once in a while. I can still use it,
but that's not the issue. I don't use it for serious work so no
longer have to "know" the answer before. The slip stick just gives
the significant digits, not the order of magnitude. A calculator
gives it all so the part of the brain that supplies the other half
atrophies.

The maps from this area show the street quite clearly. All of them. But
yeah, if I were a trucker I'd also have GPS. With the emphasis on "also".

I can assure you that maps don't show every street in the world.
If I were a trucker, I'd do exactly what I do now. I'd have both,
but the paper map is a backup. I don't have a map in the car for
every city in the country though. Neither does the trucker.

Oh no, I am involved in quite a few large projects and thoroughly enjoy
them because I get to work in teams. But those are all well managed. I
fail to see that level of management competence on some software projects.

So you're saying that only hardware projects can be managed
effectively?

Many engineers think managers are non-essential. Thing is, they are very
essential. The head of R&D at my first employer was a radio/TV tech by
trade. He often didn't have the foggiest what we were doing but he knew
he could trust us and we knew we could trust him. Excellent manager.

I don't think many engineers thing they're non-essential, just that
theirs is useless. ;-) Yes, some of my best managers haven't had a
clue what I was doing but defended me anyway. Others could do my
job. I've been fortunate (though perhaps I'm easy) and can count
the bad managers on one hand. The key is that good managers,
manage. As the corporate world is set up, that usually means
defense. There is a reason I never wanted to be a manager.

 

[Joerg]

In real life, they are. Most reasonably complex hardware projects are
successful, and most reasonably complex programming projects aren't.

It succeeds or fails with the skills of the software manager and also
the organizational diligence, version control, documentation and such.
Some programmers seriously think a comment line here and there
constitutes proper documentation.

That's not to say that hardware designers are always better in that
respect. I have met a few who do documentation after the project and see
it as a chore.

 

[krw]

That is why so many kids are home-schooled in this area. I know a father
who has to explain physics stuff to his offspring because obviously the
school failed to do so. And this is not a kid with a lack of motivation
or IQ, not at all.

What about parents that don't have the skills to teach physics?
What about parents who aren't equipped (mentally, financially,
whatever reason) to home school? Should these kids be left behind
because schools fail? Withdrawing from the system may be the answer
for individuals but it doesn't help the system. Schools must be
held accountable no matter what the parents do.

Thing is, it doesn't require any extra time. While it bakes I continue
to work off the Saturday honey-do list. When the bread is done done I
give the coals 2-3 minutes to get back to full heat and do my regular
barbecue session.

It takes time to learn how to do things another way. The first
time, for instance, the bread requires more "tending". It would
certainly take me significant time to make a wood fire. That time
can be spent in only so many ways. If it's recreation, fine.

That atrophy is what I want to avoid

I'd rather it hadn't, but there were many years I didn't even need
a calculator. I finally bought another one (HP-35s) a year ago
because I finally had need for it.

In my line of work there always comes that time where you are in the
middle of an EMC session with a hard time limit and there ain't no time
to break out the calculator. Got to rattle it off in your head, change a
few parts, run outside and close that big steel door.

The calculator isn't in prison. ;-)

I did not say that. I said "some software projects". For example,
certain operating systems ...

If you want to see a real mind-boggler try TLAVu (to view logic analyzer
recordings). That ought to be the most sluggish SW I've ever used,
drives me nuts.

It seemed you were railing against all software as being bloated
with the bloat being management (mis)directed. I counter that that
bloat may be management directed, but management is directed by the
customer. Paying customers like feature-itis, whether they
complain about it or not.

I thoroughly enjoyed being a manager. Sure, as a consultant I have a lot
more freedom now but I do miss all those people and the team work.

I've worked in large teams, at small independent tasks in large
teams, and in small teams. They're all fun, as long as management
doesn't F**K it up. The company I work for now seems to be pretty
well run, though doesn't have infinite resources. Management is a
little bit optimistic WRT schedules, though (the owner is getting a
little antsy ;-).

 

[Joerg]

On Fri, 05 Dec 2008 09:24:18 -0800, Joerg

[snip]

That is why so many kids are home-schooled in this area. I know a father
who has to explain physics stuff to his offspring because obviously the
school failed to do so. And this is not a kid with a lack of motivation
or IQ, not at all.

[snip]

I have the whole gamut right now: tutoring a second grader in
subtraction, and a junior in high school in Calculus ;-)

I find that quite sad. In our school back then we learned things right
then and there, no tutoring. I only had to ask my father about the
things that weren't in the curriculum, like how an oil furnace works so
I could fix it when he was on a biz trip and the thing went on the
fritz. Among the best teachers were lay people. For example one guy from
our ham radio club prepared an elaborate short seminar teaching us how
PAL system color TVs work. I was around 15 years old and he made it so
clear that I understood it. Still amazes me how he did that.

 

[Joerg]

I think that good coders can write large-scale programs in C++ significantly
faster than in straight C, but the problem is coming up with "good
programmers" -- there's also much more potential to shoot yourself in the foot
with complex C++ programs than with complex C programs. Heck, even with C,
some people consider something like pointers to be "kinda dangerous" -- these
are not the sorts of people I'd want to go on a hunting trip with.

I wouldn't want to go hunting with a guy who is haphazard with pointers.
At least not if he handles his rifle the same way.

There's nothing wrong with straight C, and for smaller projects it might make
more sense than C++. I've used C++ on small embedded systems where you have
to waste time making your classes and their methods static so that you don't
immediately blow out the mere 8kB of RAM that you have... it was an
interesting learning experience, but in retrospect C probably would have been
just as good for 90+% of the code.


I don't think it's as guided as you think it is ... although certainly the
paid guys doing distributions -- Canoncial, Red Hat, etc. -- are trying hard
to make it all a little saner.

The instant I learned about forks and stuff I became a bit skeptical.
It's like a wagon trail and the last three take a turn-off but all the
others don't.

They do, but they also have more annoying little glitches than similar
mechanisms from years ago -- old casette players, TVs, VCRs, compasses, cars,
etc. had far fewer potential failure modes, and it was usually pretty obvious
what the root cause when something did fail.

Yep. A Swiss army knife was often all you needed to fix it. Mostly.

Not that I want to rewind the technological clock by 25+ years here -- not at
all. But I think the way that software and hardware is designed in this world
needs some revisiting, since I'm convinced that considerably better quality
results can be obtained without any more costs for development.

I purposely bought a car with the least amount of electronics I could
find. That has paid off for me, zero defects in >10 years, looks and
drives like new. Also, I remember a CEO years ago who asked me whether I
new anyone who'd be willing to part with a 2-4 year old 7-series BMW.
Money was not an issue, he just didn't want a new one with dozens of
micro controllers in there. And guess what, he also had a masters degree
in EE.

 

[Vladimir Vassilevsky]

Joerg wrote:

I purposely bought a car with the least amount of electronics I could
find.

Having some practical involvement with the ECUs for the modern vehicles,
I must say that you made a wise decision

That has paid off for me, zero defects in >10 years, looks and
drives like new.

I guess you can still buy russian made "Lada" where the only piece of
electronics is in the ignition. But that doesn't mean it is a good car

Also, I remember a CEO years ago who asked me whether I
new anyone who'd be willing to part with a 2-4 year old 7-series BMW.
Money was not an issue, he just didn't want a new one with dozens of
micro controllers in there. And guess what, he also had a masters degree
in EE.

Aha, the guy knows what this stuff is really worth


Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com

 

[Joerg]

Vladimir Vassilevsky wrote:

[...]

I guess you can still buy russian made "Lada" where the only piece of
electronics is in the ignition. But that doesn't mean it is a good car

I remember those, basically they were Fiat copies maybe with the
exception of the Lada Niva. In the 70's they came with a tire iron _and_
a crowbar. Impressive. They must have figured that everyone will easily
pull the tire off the rim and mount a new one. Not that it's impossible,
I've done it. But afterwards my knuckles hurt.

 

[Joerg]

We just ordered a new gas stove. The rule seems to be, the less
electronics, the higher the price. The Kenmore and Bosch type stuff is
cheap, full of processors and touch screens and displays, and will
fail in a few years. The starting price for a gas stove with minimal
electronics - just the ignitor - is about $2K. If the ignitor fails, I
can still cook.

When we bought a dishwasher I asked the clerk whether he had one without
electronics behind the buttons. "I knew you'd say that, you must be an
engineer".

I do *not* need a coffee maker with a CPU inside. I need a good kettle
and a filter cone.

We came by a garage sale today and I was tempted by an old wok set for
gas stoves that would fit the charcoal barbie nicely. Said to myself
that we don't need more stuff. Now I am having second thoughts ...

 

[Joerg]

I presume you mean a cast-iron wok? Run back to the garage sale and
see if it's still there. I've had one for years, and it's dandy on
the bar-b ;-)

No, unfortunately not. Then I would have bought it immediately. It's a
steel version with a ring to set it onto. Come to think of it, they did
have a large cast iron skillet but it was quite crusted up inside.

We've got two small cast iron skillets, roughly twice the size of a
multimeter but only 1/2" rims. We did already bake them in with oil so
they are ready. I am still pondering what to cook with those on the
Weber. Maybe mini pizzas or so.

 

[Rich Grise]

Not that I want to re" here -- not
at all. But I think the way that software and hardware is designed in
this world needs some revisiting, since I'm convinced that considerably
better quality results can be obtained without any more costs for
development.

You wouldn't have to "rewind the technological clock by 25+ years" (as
much as I'd like to rewind my biological clock! ;-D); just find somebody
who can use a modicum of common sense in their designs.

Cheers!
Rich

 

[Joerg]

You wouldn't have to "rewind the technological clock by 25+ years" (as
much as I'd like to rewind my biological clock! ;-D); just find somebody
who can use a modicum of common sense in their designs.

That is becoming harder and harder. Many folks think that every mundane
kitchen gizmo needs a uC with an RTOS on there.

Buying simple things is becoming tougher as well. For example the number
of toasters without chips in them is dwindling. People in Germany told
me it's next to imposible over there to spot a version with a simple
bi-metal thermostat.

We have a gourmet coffee maker in repair right now. I don't think the
mfg makes any money on this repair considering it's $30 inclucing both
ways shipping. What had croaked? The electronics. Of course.

 

[Rich Grise]

I wouldn't want to go hunting with a guy who is haphazard with pointers.
At least not if he handles his rifle the same way.

Oddly, I took to pointers like a duck takes to water - they seemed just
sort of "natural". And yes, I'm adamant about keeping track of who does
what to whom. ;-)

Interestingly, it's a rag-tag group of random programmers worldwide, much
like the Free Market. ;-)

The instant I learned about forks and stuff I became a bit skeptical.
It's like a wagon trail and the last three take a turn-off but all the
others don't.

One difference - forks have the equivalent of cell phones. It's like half
the group goes to the good hunting places and the other half goes
foraging in the good forage territory. Then, when their tasks are
completed, they come back together and make stew. ;-D

Cheers!
Rich

 

[Rich Grise]

I do *not* need a coffee maker with a CPU inside. I need a good kettle and
a filter cone.

I once had a plastic filter cone; just the cone with a flange at the small
end; it could sit on top of a mug or on top of an ordinary coffeepot. It
was about $1.99 for just the cone. I've looked up "Mellita" and didn't
immediately see the one I had, but they might have one in the "coffee"
department at your local store.

Hope This Helps!
Rich

 

[Rich Grise]

Anyways I have a way to make rice in the microwave that really works.

Do Tell! I do it on the stovetop and usually burn it. )-;

Thanks!
Rich

 

[Rich Grise]

Yeah, but you might have lost enough in, say, fuel economy that you've
more than paid for the repair bills you'd otherwise have had.

Maybe I'm having selective memories, but I remember seeing cars get
mileages in a really close ballpark as these EPA numbers; they just
kept them tuned properly!!

I had a 1.6L Pinto once, and routinely got about 30 on the highway -
that's a kinda rough estimate; I had an 8-gallon tank and could go
at least 240 miles. Admittedly, it's a small car, but I know how
to drive, so it was never in any danger of getting run over. >:->

Cheeers!
Rich

 

[Rich Webb]

Two cups of water in a small pot. Bring to boil. Add a bit of butter
and a little salt. While stirring, add rice. Reduce to very slow
simmer, cover, and wait 10 minutes. Turn off the heat and wait another
10 minutes.

Jasmine rice is best.

That's it.

I've become a big fan of my little Zojirushi cooker. It just works. I'd
really recommend giving a real [*] rice cooker a try.

But jasmine rice? Oh yes, little elephant, oh yes. Wonderful stuff!

[*] "real" in this context meaning that the wording embossed on the dew
cup doesn't use a roman typeface. Black and Decker may make great socket
wrenches but I'll go with Japanese rice cookers. ;-)

 

[JosephKK]

Sure, but my point is that are some "hard" problems out there that are worth
pursuing, and these problems benefit greatly as a direct result of the cheap
CPU cycles and terabytes that come as a result of more "frivolous" needs such
as storing your MP3s, DVD collection, having an animated paper clip ask you if
you're writing a suicide note, etc. For instance, this Christmas season
there are several $99 (and below) GPS navigation units available. Now, at
least to me, GPS navigation units do address a worthy problem -- helping
people get from point A to point B more easily and with fewer errors.
Internally the map database for these units is often on the order of a
gigabyte and you've got some ~hundreds of MIPS CPU running the show, driving a
color LCD touchscreen and often having a Bluetooth phone interface as well.
Without the sort of "quick and sloppy" programming techniques that drove the
demand for cheap but fast CPUs and large storage devices, it'd probably be
another decade before the price for something like this would be under $10k.


Well, I'd agree it's likely to have an increase, although I can't really say
"just how huge" that increase would be. Some people adopt the rather
pessimistic view that there are so-many bugs per lines of code, so the more
lines of code... the more bugs. Personally I think that with good software
development techniques, for the *overall product*, the number of bugs per
lines of code should drop as the LOC increases, because the overall product
isn't going to exercise every possible execution path and thereby be hit with
every bug.

---Joel

Each to their own, the studies that i have read show the flaws per
KLOC is mostly dependant on the software process, and not so much on
the size of the project. The most disciplined processes have to use
flaws per MLOC. Ad hoc processes use flaws per LOC.

YMMV

 

[JosephKK]

Don't blame GPSs, or any technology, for the lack of map reading
skills when schools and parents are truly guilty here.


They never RTFM. That's not technology. Again, put blame where
blame belongs.


That one you can blame on technology. Since calculators, I've
completely lost any ability to estimate values by quick inspection.
Don't need Spice for most things, other than it's very nice for
documentation.

How very strange, for most of the last three decades i have had to do
actual measurements instead of estimating and my ability to estimate
has improved.

YMMV