Michael A. Covington, Ph.D.
Senior Research Scientist
Adjunct Professor of Computer Science
Associate Director
Artificial Intelligence Center
The University of Georgia
This is a private web page not hosted or sponsored by the University.
Copyright 2006 Michael A. Covington.
Caching by search engines is explicitly permitted.
To go to the latest entry every day, bookmark

Previous months
About this notebook
Site search

Daily Notebook

Popular topics on this page:
Beckman Circuitmate DM25L schematic
Heath Malmstadt/Enke Instrumentation Laboratory
Free circuit simulation software
Carlingswitch VJD1 illuminated rocker switch
No, the comet is NOT going to hit us
Remove from server when deleted from 'Deleted Items'
Oxford English Dictionary hangs when you click "Search"
When Windows Installer goes wild
DirectShow and .NET
Astrophotos: 3C273 Comet 73P Comet 73P Comet 73P Cosmic rays Jupiter M3 M51 M104 Stargate

Scroll down or press Ctrl-F to find other topics, or check previous months.
BOOK SALE - I'm selling off some scholarly books via Amazon Marketplace. Have a look!


You are not a Homo sapien, and you don't have a bicep

I am mildly annoyed at the makers of the robot "Robosapien" for perpetuating a common Latin spelling error.

The species name for human beings is Homo sapiens ("intelligent human being") and is not plural.

If you wanted to apply the name to more than one of them, they'd be Homines sapientes, "intelligent human beings."

Some Latin words end in -s; deal with it.

A discus is not more than one discu.

A virus is not more than one viru.

So why should a biceps be more than one bicep...

or a Homo sapiens be more than one Homo sapien?

(And does anybody think a moose is more than one moo?)

In Latin, sapien is not a word. And the Latin plural of biceps is bicipites, although it's OK with me if you say "bicepses" when you want to denote more than one of them in English.

That's all until June. In the meantime, look up the correct plurals of octopus and virus. You'll be surprised!

Permanent link to this entry


When Windows Installer goes wild

Windows Installer pops up when you open almost any window
Windows Installer can neither install nor uninstall a program
Windows Installer complains that the patch file is corrupted or unreadable

Last night (May 28) I made two computers into one. To be precise, I moved the boot disk of Melody's old computer (with all her files and software) and made it the boot disk of our cheap but good eMachines desktop that is a few years younger. I followed the procedure here, and for the most part it worked.

But a couple of Adobe products were corrupted (no longer runnable), and Windows Installer would pop up with "Preparing to install..." (and then, after about a minute, log a failure event) almost every time I opened any window. My Computer and Internet Explorer were particularly afflicted.

It turns out that these were two faces of the same problem. Adobe Acrobat hooks itself to almost every window that can print, and when it's corrupted, it tries to fix itself at every inopportune moment.

But I couldn't uninstall or reinstall Acrobat. The installer (Add/Remove Programs) complained that the patch file was corrupt.

Fortunately, Microsoft has had the same problem (with installations of Office) and has worked out a solution.

The problem arises when the Registry contains incorrect or incomplete information about how the software is installed. This is in some deeply hidden Registry keys pertaining to the Installer, not anything obviously pertaining to the software.

Microsoft has published a Windows Installer cleanup utility for dealing with such situations. It doesn't uninstall software, nor does it remove anything from your disk or registry except the records of how to install it.

To fix the problem, you must therefore run the cleanup utility, then reinstall the exact same software in the same place, and finally uninstall it (if you want it uninstalled).

At least when they make a mess, they have to find a way to clean it up!

Permanent link to this entry


DirectShow meets .NET and C#

Microsoft Windows has two rival video systems. The old one, Video for Windows, is basically an extension of the Windows procedure-call API and is simple to use, but brittle – it breaks on lots of newer video file formats. In particular, it doesn't support .WMV, nor DV AVI. But it is still widely used for scientific work, and I've enjoyed Corinna John's C# wrappers for it.

The new (well, 10-year-old), powerful video system is called DirectShow. It's written in COM, and programming with it is a very strange experience. Like old IBM mainframe JCL, COM gives me the topsy-turvy feeling that the hard things are all easy and the easy things are all hard. Brilliant software insights are mixed with illegible and clumsy notations.

I understand the power of COM but don't enjoy working with it at all. Apparently, neither do a lot of other people, and that's why we now have .NET Framework.

Well... Some very generous and right-thinking programmers on SourceForge have created and shared a powerful set of open-source C# wrappers for DirectShow, with a fine set of sample programs. You can get them here, free of charge. They're under the LGPL, which, if I understand it right, means that the DLLs are freeware but can be used in commercial products. And they are genuinely easy to use.

Permanent link to this entry


Who remembers Malmstadt/Enke?

If you studied electronics in college in the late 1960s (or played around in a college lab, as I did), you may well have encountered the amazing, bluish-green Heath Malmstadt/Enke Instrumentation Laboratory system. This was a set of coordinated test equipment and breadboards that provided a practical way of breadboarding circuits without soldering in the vacuum-tube era, an impressive achievement.

Occasionally, Malmstadt/Enke instruments turn up on eBay; the breadboards are probably the rarest item in the set. Search for Heath products with model numbers that begin with EU and EUW.

The test equipment was notable in its own time for being much less expensive than competitors. Like most Heath test equipment, it was cheaply made, but not too cheaply; the price-performance ratio was exceptionally good.

The breadboard had no competition as far as I can determine. Multiple breadboards could be bolted together to build complex circuits.

I do not presently own any of this equipment, although I admired it in catalogues and occasionally got to use it as a high school student at Valdosta State College in the late 1960s. I've collected some information in this PDF file. Rather full documentation is in the book Electronics for Scientists, by Malmstadt, Enke, and Toren, which can often be found secondhand on Amazon and is worth reading for its own sake.

Permanent link to this entry


Quote for the day

"The only alternative to tradition is bad tradition."Jaroslav Pelikan.

Permanent link to this entry


Jupiter, Callisto, Io, and Europa

Here's last night's catch. We had unusually steady air. The other thing that was unusual is that the earth and Jupiter are positioned in their orbits in such a way that we can see satellites pass north or south of the planet; usually they pass in front of it or behind it as they orbit from west to east.

8-inch Meade LX200 telescope, prime focus, Philips ToUCam, 2000 frames of video recorded and about the best 1000 stacked and enhanced.

Permanent link to this entry

Some interesting reading

Interesting things I've found on line recently...

The London Stone is an ancient pre-Roman artifact, mentioned by Shakespeare, and now housed in a sporting-goods store.

Deep thinking: St. Bonaventure's Journey of the Mind (Itinerarium). I haven't read it yet.

The Purdue University Horticultural Humor Page.

An easy-to-build tiny FM transmitter for electronics beginners. The construction technique, with unetched printed circuit board used as a ground plane, is a time-honored one that I've never tried. I may build one, just to see how well it works.

Permanent link to this entry


Red, white, and blue clouds

This isn't a distant nebula. It's a cloud formation I saw outside just a few minutes ago. I grabbed my Canon Digital Rebel, took the picture, brought it onto the laptop, and enhanced the color and contrast with Photoshop. Voilà!

Permanent link to this entry



What are those things on Inuyasha's face in the picture? Apparently, they are a Japanese representation of veins throbbing on an anguished person's forehead. They remind us that many of the things we see in drawings – arrows, starbursts, etc. – are not pictures at all; they're arbitrary symbols that we've gotten used to.

(They are not the logo of the Atlanta airport, much as they look like it!)

Speaking of drawings, I'm beginning to suspect that Google Sketchup may have finally figured out a good user interface for 3-dimensional drafting. Maybe Sketchup is doing for 3D what Visio did for diagrams. Try it and see; it's free for noncommercial use.

Today I made minor revisions to my existing blog entries about fixing a Lexmark Optra R printer and about The Gospel of Judas (readers will be relieved to know the two are unrelated).
Sunday, December 12, 1965, was a memorable day. At least, one event was memorable. My father and I were watching the launch of Gemini 6.

The rocket ignited right on time, smoke billowed out from under it, and then it didn't move.

After a while the astronauts came out and the launch was rescheduled for December 15.

At the time, we had no idea what a fright this was for the astronauts. If the rocket had lifted off even an inch, it would have exploded coming back down, and the astronauts would have had to use their ejection seats, which were unproven and thought to be very dangerous.

Fortunately, Commander Schirra made the correct decision not to eject; he hadn't felt even an inch of movement.

Wikipedia's articles on the U.S. space program are very informative and worth reading.

Two creative ways to get yourself in trouble:
  • (1) Fall off a cruise ship. It's surprisingly common for people to go overboard and drown, and this CNN story sheds some light on why: Missing-person reports on ships are very common, and nearly all of them are mistaken.

    You'd think it would be a simple matter to monitor the sides of the ship with security cameras and detect it the moment someone goes over. Why not?

  • (2) Dial random numbers on the telephone in search of adventure. This was done by a 12-year-old girl near Boston who got herself kidnapped.

    And apparently, it's a fairly common practice. Maybe this explains some of the wrong numbers I get. I make a point of answering the phone "This is Michael Covington, may I help you?" rather than "Hello," and a lot of people don't catch a word of it!

Permanent link to this entry


Free circuit simulation software

The other day I wrote about Multisim circuit simulation software, which is available as a 45-day free trial, but otherwise quite expensive (on the order of $2000).

Is there anything like this that is free of charge, at least in limited form, and doesn't expire after 45 days?

Yes; so far, I've found three.

One is the demo version of OrCAD, which is limited only as to the size of the circuit. It's a huge download (over 200 MB) and comes with copious documentation.

I haven't gotten started with OrCAD yet. It's apparently very powerful but is an order of magnitude harder to get started with than Multisim. (See these handy notes about an earlier version, and note that what you normally want is a transient analysis, i.e., a virtual oscilloscope.) Since OrCAD is widely used by professionals, the effort of learning it is likely to be worthwhile.

I had some trouble getting OrCAD Demo. OrCAD's web site is supposed to send you e-mail once you register for the download. After three tries, I concluded those e-mails just weren't being sent. I e-mailed OrCAD's webmaster, as directed on the web page, and got an autoreply from a software support company that didn't identify itself as OrCAD. Finally I contacted OrCAD's North American sales representative and got immediate help.

The parts library for OrCAD's free demo is rather restricted. Also, curiously, OrCAD doesn't seem to know about My Documents. You have to navigate to your My Documents folder by finding it under C:\Documents and Settings.

The second is Micro-Cap, whose demo version is designed for students and is relatively easy to get started with (though not as easy as Multisim). In the free demo version, the parts library is limited but consists mostly of commonly used items, with no strong preference as to manufacturer.

Of the three, this one is the most like Multisim.

The third is LTSPICE, also known as SwitcherCAD. This is what you see below. It's small, quick, and easy to get started with once you realize that the power supply is hidden under Components, Misc, Voltage.

LTSPICE was created by Linear Technology Corporation to promote their ICs, many of which are used in switching power supplies (hence the name SwitcherCAD). Switching power supplies involve oscillators, inductance, and calculations you can't do in your head, hence the need for the software. LTSPICE also includes many parts made by other companies – but only when Linear doesn't have a competing product. All the op-amps, for instance, are from Linear and have LT numbers.

Compared to Multisim, LTSPICE has a less copious selection of parts available and has more involvement with traditional SPICE (e.g., you see, and can edit, the traditional .tran command that starts each simulation). Some of the parts in the library are just symbols to use with a SPICE model of your own devising. My favorite chip, the 555 timer, is present only as an "idealized" model (whereas Multisim gives you realistic bipolar and CMOS versions).

There were two "gotchas" with LTSPICE. When installed, it puts its shortcuts only into the start menu of the person who installed it. To put it onto a computer for general use, install as Administrator and then move the shortcuts from Administrator\Start Menu to All Users\Start Menu, then add "Users" to the list of those authorized to read and execute it.

Second, by default it wants to save your files in C:\Program Files, where you probably don't have permission to write.

Permanent link to this entry


Seven easy pieces

What's remarkable about the following seven astrophotos is how little work they were. I put my Canon Digital Rebel on the back of my 8-inch f/10 telescope using a Celestron compressor lens to get effective f/5.6. The telescope, a Meade LX200, has a "smart drive" with permanent periodic error correction, so it tracks very smoothly, and its permanent pier is polar-aligned to within about 0.1 degree.

I focused the camera with the Canon Angle Finder and did not use an autoguider; no guiding corrections were made during the exposures. Each exposure was 3 minutes; most of the pictures are stacks of three of these.

Dark frame subtraction, alignment, and stacking were done with Images Plus; final processing, with Photoshop (usually involving some unsharp masking).

[Note added May 20: Add Images Plus to the list of software that won't run unless you're logged on as Administrator. Even altering the file permissions won't fix this.]

First, the "Stargate asterism," so named because it resembles a logo formerly used by the TV show Stargate SG-1. This is a small group of stars in Virgo (not actually a cluster; they're different distances from us) in almost the same direction as M104. Here you see a single 3-minute exposure.

Next, the galaxy M104 itself, a stack of three 3-minute exposures:

The quasar 3C273, same technique, showing stars down to 16th magnitude. This quasar is probably the most distant object I've photographed. It's one of the brighter stars near the middle; the arrow indicates it.

The spiral galaxy M51, same technique again:

Coming back into our own galaxy, the globular cluster M3, a composite of two 3-minute exposures:

And finally, very close to home (but still much farther than the Moon), some fragments of Comet 73P/Schwassmann-Wachmann. Here's Fragment C:

And here's Fragment B along with another fainter fragment (arrow).

Each of these comet pictures is a single 3-minute exposure. Even in so short a time, the image of each piece of the comet is elongated by the comet's orbital motion.

Permanent link to this entry


It was against the law...

It turns out laws were violated when our local jeweler autodialed everybody in town and played recordings into their telephones. The Do-Not-Call List deals with telemarketing by human beings. Use of autodialers to deliver ads to residences is illegal even if they're not on the Do-Not-Call List ( 47 USC 227 (b)(1)(B)).

Autodialing the University probably wasn't illegal, but it wasn't wise either.

Permanent link to this entry

Oxford English Dictionary hangs when you click "Search"

The Oxford English Dictionary on CD-ROM is a very handy piece of software; you can copy all the data files to your PC and run it without the discs.

I've been having problems with it starting up slowly, or hanging altogether, when I click "Search" to open up the dictionary.

The cure is to tell F-Secure Antivirus not to search the OED's DATA folder. You can exclude specific folders from automatic virus searching (even if you're not Administrator), and this greatly speeds up programs that have to open large data files. Since the F-Secure data files have the .htm extension, they get searched as if they were web pages.

Permanent link to this entry


Yes, it's that bad
The Da Vinci Code opens to very negative reviews

Besides annoying all the Christians in the world with its distortions of history and misrepresentations of real people and organizations, The Da Vinci Code is, to my surprise and delight, getting catcalls, not applause at the Cannes Film Festival.

The plain fact seems to be that a puzzle does not make a very good movie, and even if it did, this one isn't very well made.

Much of the furor over The Da Vinci Code comes from the fact that the author didn't draw the line between fact and fiction in any sensible place. He was taken in by a hoax (the Priory of Sion) while researching the "real history" that he claims is behind the book. He uses a real organization (Opus Dei) and makes it into a murderous cult. He also uses a living person (art historian Maurizio Seracini) as a character without his permission.

In short, it's neither fact nor fiction – it's confusion.

Fiction about an alternate world can be enjoyable. What if the South won the Civil War? Or the Americans lost the War for Independence?

But The Da Vinci Code isn't that. It's apparently designed to confuse and spread misconceptions. What would happen if someone did the same thing to the history of the American Revolution or World War II?

Permanent link to this entry

An interesting bylaw

Speaking of things Roman...

I've just learned that the Catholic Church has a bylaw for the electing of Popes that solves a potential problem in an interesting way. Managers and politicians, take note.

The bylaw says that if any of the Cardinals is under excommunication or censure, that excommunication or censure is set aside for the purpose of electing or being elected Pope. (Details here.)

The effect? No papal election can be set aside by proving, later, that one or more of the Cardinals was under excommunication or censure at the time.

It is possible to incur excommunication automatically, without a court passing sentence, by doing certain things, such as joining certain anti-Catholic organizations. Thus, a Cardinal could be under excommunication without anyone knowing it at the time. Allegations could be brought later.

If a papal election were invalidated ex post facto, all sorts of things would go wrong. The cardinals appointed by the invalidated Pope would be invalidated. Thus, so would any further papal elections they had carried out.

In fact, there are people now who contend that the election of John XXIII was invalid; thus the cardinals that he appointed are invalid; thus Paul VI, John Paul I, John Paul II, and Benedict XVI are not validly elected because there was a steadily increasing number of invalidly appointed cardinals among the electors; and so on. Computer scientists will recognize the cascading power of recursion here.

That's what this unusual bylaw is designed to fix. I wonder if similar rules would be useful in other types of organizations.

Permanent link to this entry

Simulating circuits with Multisim

I'm trying out the latest version of Electronics Workbench, which comprises Multisim circuit simulation software as well as other tools which will even lay out the printed circuit board for whatever you've designed.

The latest version has a free 45-day trial. You can download it here.

Definitely an interesting thing to try out! I like it particularly for designing tricky analog circuits such as active filters.

The realistically simulated Tektronix and Agilent oscilloscopes are especially fun. This software may actually teach our students how to use our test equipment.

[Note added May 20: Like much engineering software, Multisim currently violates standard Windows programming procedures in at least one way – it expects you to log in as Administrator. To enable ordinary users to run it, give them full control of its folder within Program Files, or perhaps specific files or subfolders within it.]

Permanent link to this entry


How not to advertise a jewelry store

This morning, everybody in my office got a phone call or voicemail from a prominent local jeweler advertising a free ring-cleaning offer. The receptionist got multiple copies of the voicemail from all the telephones that roll over to hers.

I called the store to find out what was going on. They were autodialing, and playing a recording to, "random" phone numbers that were checked against the Do-Not-Call List. Of course, our phones, being state government, are not on the list. Frankly, until today it didn't occur to us that they needed to be.

Why is it a bad idea to autodial telephones without knowing whose they are? If you can't think of any reasons, let me suggest a few...

(1) Even if you're not violating it, remember why there's a Do-Not-Call List. People don't like to receive recorded commercials by telephone. It may be technically legal in some circumstances, but it doesn't win customers.

I wonder if the store owner's home telephone is on the Do-Not-Call List. Probably so. "Do unto others..."!

(2) This ad campaign looked, at first sight, like a deliberate attempt to disrupt the University's voicemail system by filling up its memory with a vast number of copies of the same message. That would be a type of denial-of-service attack. Our security people are still looking into that possibility.

(3) The University has telephones in conference rooms and other places where incoming calls will disturb people. Their numbers are unlisted, but if you dial "random" numbers, you'll hit them, possibly interrupting a meeting or a class.

(4) What about police departments, fire departments, and hospitals? In particular, if you're calling all the unlisted numbers in town, you'll ring every bedside telephone in the whole hospital, waking up every patient. That is not a nice thing to do! (And autodialing health-care facilities is illegal.)

In short: This particular jewelry store has not endeared itself to the University. I presume some unscrupulous ad agency sold them a bill of goods – assured them that this scheme would make money, that it was legal, and that nobody would mind. Falsch!

Permanent link to this entry

Multimeter madness

I have yet another multimeter, a B&K 2706A, picked up on eBay because it was too cheap not to. (About $32 delivered, as I recall.) It's like the 2704A with a temperature scale added; its sensor is a standard Type K thermocouple.

And in response to popular demand, I've put the schematic of the Beckman Circuitmate DM25L on line even though my own DM25L is beyond repair.

Permanent link to this entry


Two ecclesiastical notes

Jaroslav Pelikan, the great church historian who converted from Lutheran to Eastern Orthodox at an advanced age, and who was Dean of the Yale Graduate School when I was a student there, has died at age 81.

He will be missed. I have his five-volume magnum opus on my shelf and refer to it regularly. His big message to his fellow Protestants was that you cannot jump straight from the Bible to the modern era without considering the development of Christianity during the intervening centuries.

Sample some of Dr. Pelikan's erudite writing here.

Meanwhile, Christian financier Gary Moore sounds two warnings about the "Christian Money Management" movement (of which I have also been somewhat critical):

(1) It does not make sense to demand that people always be "debt-free" since we have everything on loan from God in the first place. The compulsion to be "debt-free" can stem from a kind of selfishness that is contrary to Christian values, the desire to be "independent" even of the Creator. It is also a needless burden on people who could prosper by borrowing money or other resources. Borrowing, says Moore, is Biblical; Jesus even rode into Jerusalem on a borrowed donkey.

(2) A Christian cannot be a perfect laissez-faire capitalist. The free market is good for a lot of things, but it's no substitute for good sense. The free market does not authorize you to ignore the known effects of your actions.

Permanent link to this entry


Con game of the day: Fake checks

The other day somebody sent to Covington Innovations a check for $3.25, with the "memo" field filled in as "Computer Software" and a fake purchase order number.

I do make software, and $3.25 would pay for about 90 seconds of my consulting time.

On the back of the check, in very fine print, was a statement that by endorsing this check I am signing up to pay $27.50 per month to be listed in YP.COM.

(Note: I have no proof that the management of YP.COM was behind this. It may have been done for/to them by an ad agency or even an impostor. Just like spam, deceptive advertising should not be assumed to come from where it claims to come from.)

In my opinion, it ought to be illegal to put a contract into the endorsement of a check (other than maybe an agreement that the check is payment for a specific debt).

But even if it's technically legal, this check went over the line by including the false purchase order information under "memo."

The sender clearly hoped that someone at Covington Innovations would deposit it without reading the fine print on the back.

It was drawn on a tiny state-chartered bank in Minnesota.

I sent it to the Federal Reserve System, from which it will filter down to the Federal Reserve Bank of Minneapolis and then the appropriate regulators. From the Fed's point of view, of course, the issue is the bank's complicity in an apparently deceptive practice. I suspect they ended up with a tiny bank, far from home, because the big banks wouldn't play the game.

It's even imaginable that the whole bank is fake. Such things have happened!

My other reason for contacting the Fed was of course to suggest a new regulation. I cannot think of any legitimate reason to put subscription contracts or the like into the endorsements of checks. Deception seems to be their main purpose. Why not make life easier for everybody by banning them?

Permanent link to this entry


My best pen yet?

This hasty snapshot doesn't do justice to the fountain pen that Melody's father, Jim Mauldin, made as a surprise present for me a few days ago. Its barrel and cap are not metal but turned ebony wood. The cap screws onto either end of the barrel, and when the pen is capped, it's sealed with an O-ring. Clearly, the maker of this pen is a machinist at heart.

The nib is European, has an iridium point, and takes international cartridges. To write well, it needed to be opened up just a bit by inserting a razor blade about 1/16 of an inch between the tines. In my experience, most pens need this. When the nibs leave the factory, they are deliberately left very tight so that they can be used by people who bear down as hard as if they were using a ballpoint.

You, too, can get a pen custom-made. Call Jim Mauldin at 770-867-5944 or e-mail me and I'll pass the request along. He gets requests regularly from my web site.

Permanent link to this entry

Amazing comet photo (not mine)

Have a look at this amazing picture of part of Comet 73P/Schwassmann-Wachmann passing almost in front of the Ring Nebula.

The brightest fragments (B and C) are now in Cygnus and have brightened to sixth magnitude. We've had cloudy weather, and I've only had one quick look at them lately. That was last night. At 1:30 a.m., Babbage, whom Melody has dubbed "the astro-dog," came and woke me up, demanding to go outside. I grabbed my binoculars and got a good look at Fragment B while Babbage ate clover or something. Within fifteen minutes, the sky was cloudy; he somehow knew just when to get me up, in the short interval between the comet's rising and the arrival of the clouds.

Some astrophotos of mine from more than a week ago are going to appear here soon, including an image of the comet through the telescope.

Permanent link to this entry


A treasury of semiconductor and IC data sheets

DatasheetArchive.com has greatly expanded its coverage.

That's a free archive of data sheets on semiconductors and ICs past and present. They make their money by trying to link you to distributors who still have obsolete parts available.

In the 1980s, everybody had a bookshelf of IC data sheets. (I still do; I got a good bit of my technical education from the National Semiconductor "Linear" volume.)

Then, around 2000, most people threw them away. From now on, data will be on the Internet. Right?

Partly right. The moment a part is discontinued, its data sheet disappears from the manufacturer's web site, even though the parts continue to be available, in small quantities, for 20 years or more.

There's now a wave of discontinuations going on. Use of lead solder in consumer electronics is now prohibited in Europe. So all the slow-selling ICs that have lead in them are going to be discontinued and not replaced with lead-free versions.

This is a boon for experimenters like me, in America, who will be able to buy the lead-coated parts cheaply – if we can still get data sheets to tell us how to use them.

And of course if you're fixing an older piece of equipment, you need data on what's in it.

That's why DatasheetArchive is indispensable.

Permanent link to this entry


Oops, we forgot to put the help files on the CD!

Today's dubious achievement award goes to Adobe Illustrator CS2 for arriving incomplete. To use it properly, you have to download 30 megabytes of things from Adobe: the help files (a ZIP file that you put in the appropriate place under Adobe Help Center in Program Files), the update from 12.0.0 to 12.0.1, and the Security Patcher, which updates Help Center and does a few other small things.

So Adobe gets low marks for the way the software is delivered.

Illustrator itself, however, strikes me as an excellent product. I bought it today to end a frustrating 15-year relationship with Corel Draw, which seems to have declined in quality over the years. What really bothers me about Corel is that it keeps getting more complicated, and they keep rearranging the menus so you have to learn a new set every time there's an update.

Illustrator is simple and well organized though powerful. I figure it will take me about half an hour to learn to use it productively. Of course, I use Visio for technical diagrams; Illustrator (or Corel) is for things that don't fit the Visio model very well.

Permanent link to this entry


Prophetic words...

The last sentence of yesterday's entry was predictive... The old Beckman DM25L, may it rest in peace, suffered some kind of catastrophic failure while I was cleaning its contacts yesterday and is now a random-number generator.

So there is a B&K 2704B, its nearest living relative, on the way to me now from Digi-Key.

The two meters complement each other. The 2704B measures current and transistor gain, both of which the Fluke 110 lacks, and measures capacitance over a wider range. The transistor-gain feature isn't for testing transistors so much as for identifying them and figuring out which pin is which.

I'll probably take another stab at fixing the Beckman. I have a schematic diagram, and all the parts are replaceable. Of course, a schematic diagram that revolves around a 50-position (?), 8- or 10-pole switch is not a simple thing!

Permanent link to this entry


Remove from server when deleted from 'Deleted Items'

I use Outlook Express to check e-mail from the same POP3 server on three different PCs.

Normally, Outlook Express would delete each piece of mail from the server when it downloads it. Obviously, I don't want that to happen. So I've told it to delete e-mail only when I delete it from the Deleted Items folder in Outlook Express (see picture).

That is, the mail should disappear from the server after I move it into Deleted Items and then empty the Deleted Items folder. Otherwise, it should stay there so that the next PC can also see it.

This feature hasn't been working as advertised, and yesterday, thanks to someone on a Microsoft newsgroup, I figured out why.

Very simply: After emptying Deleted Items, I need to check mail again. Outlook Express only communicates with the server when you're checking mail. In that respect, it differs from Mailwasher, which communicates with the server immediately when you tell it to do deletions.

Permanent link to this entry

Recommended multimeter

The other day I bought two Fluke 110 multimeters for my lab, then liked them so much I bought one to use at home.

Multimeters have come up in the world since I bought my trusty Beckman DM25L about twenty years ago. The Beckman has a big dial with about 50 settings; it doesn't auto-range.

For years, I distrusted auto-ranging meters because they seemed to spend all their time vacillating between ranges. Not the Fluke. It settles on the right range almost instantly. That may be partly because it starts with a range that is usually reasonable (e.g., 6 volts) rather than starting with its lowest range.

Notice that the Fluke 110 doesn't measure current (amps or milliamps). I chose it deliberately for that reason. If it measured amps, it would also have one or two more sockets, and then the students would plug the cables in the wrong place. Not only that, but ammeters have internal fuses which, in a student lab, are almost always blown from trying to use an ammeter as a voltmeter.

The Fluke counts up to 6599, not 1999, so about half the time, we get an extra digit of resolution. Checking a PC power supply, we can read 5.013 volts instead of just 5.01.

The AC ranges are RMS-responding. That's a good thing.

But what I really like is the frequency measurement function. One of my long-term hobbies is building controllers for AC telescope-drive motors. Being able to connect the meter and read 60.05 Hz is a great help.

My one quibble with the Fluke is that it reads capacitance in nanofarads. For years, it has been traditional not to use milli- or nano- with farads (only micro- and pico-). The reason is that capacitor values span a trillion-to-one range, and to avoid changing unit prefixes too much, we use only half of the full set.

I guess I'm going to have to join the newer generation and use nanofarads. The ubiquitous 0.1-μF capacitors are now going to be known as 100 nF.

The Fluke doesn't completely replace the old Beckman, which also reads amps, milliamps, and transistor hFE.

Today I'm going to fix up the old one (which has developed some quirks) and calibrate it. If unsuccessful, I may have to get a new one. The Beckman DM25L lived on for a while in the Wavetek product line, but it's hard to say what its nearest present-day equivalent is. Probably the B&K 2704B.

By popular demand I have put a circuit diagram of the DM25L on line. Click here.

Permanent link to this entry


No, the comet is NOT going to hit us

A self-proclaimed "psychic" has been e-mailing everybody in the world saying a fragment of Comet Schwassmann-Wachmann is going to hit Earth on May 25. He uses the name of NASA in his rather confusing press release.

Well... He's wrong.

The comet is staying well away from Earth. Here's what NASA actually says. The comet will stay more than 20 times farther away than the Moon.

Bear in mind that it doesn't move under its own power, so its movements are easy to predict. Also, although it is breaking up into fragments, it's not exploding – the fragments are staying close together, scattered along the same orbit. None of them has moved violently in some other direction.

If you don't trust NASA, maybe you'll trust me. I've been observing the comet myself, and it's definitely not approaching us.

Permanent link to this entry


My debut as an artist

Last night, I needed the following diagram (a picture of some transistors with their pin connections labeled) and couldn't find anything suitable in a book. Anyhow, I needed a diagram that wasn't taken from any copyrighted source.

So I laid the transistors down on the table, photographed them, imported the image into Corel Draw, and drew over it, then deleted the photographic image. As you can see, it was a great success.

This is a good way to create realistic line drawings even if you're a better artist than I am. The idea is to combine the precision of the drafting tools (in my case Corel Draw, in an earlier era rulers and triangles) with the accuracy of the photograph. Such techniques, using film photography, were promoted in Thinking with a Pencil many years ago.

Permanent link to this entry



When I plugged my Lexar JumpDrive Pro into my office computer this afternoon, the computer announced the device as a DEPAR BUEPDRAVE PRG USB Device and then found it unreadable; it treated it as an unformatted disk.

After several tries, I eventually got the JumpDrive to work properly. I copied all the files off of it, reformatted it (as FAT16; reformatting it as FAT32 the previous night may have brought on the problem); and copied the files back.

Have you cracked the code yet?

DEPAR BUEPDRAVE PRG is what you get if you take LEXAR JUMPDRIVE PRO in ASCII code and change bit 3 of each letter (the 4th-from-lowest bit) so that it is always 0. This changes some of the letters but leaves others unchanged.

Should I be worried? I don't know. It sounds like a rather deep communication problem between the JumpDrive memory and its USB port, not a disk formatting problem. This is a battle-weary JumpDrive but I don't know if I still trust it.

Permanent link to this entry


What goes on inside an illuminated rocker switch

One of today's main accomplishments – at work, no less – was figuring out what goes on inside a Carlingswitch VJD1 illuminated rocker switch. Click on the link to see.

It isn't simple; there are two light bulbs, one of which you never see! (Unless the switch has a different cover on it than ours did.) With various kinds of labeling on the cover, this is a common switch in automotive, aircraft, and boat electrical systems. Manufacturer's data here, not always accessible; I had to make some tests of my own to actually figure everything out.

Permanent link to this entry

Calendrical checkpoint

Don't forget that tonight at 3:02:01 A.M. UT (11:02:01 p.m. EDT), the date and time will be

06/05/04 03:02:01

at least if you write the date and time with the largest units first, as I do in scientific work.

Permanent link to this entry


Practical Artificial Intelligence Workshop

No news today; I'm attending this conference.

Permanent link to this entry


Cosmic rays

Last month I got a bumper crop of astronomical photographs, the best of which are going to appear here. But we're still in the end-of-semester crunch, so it's going to take a while. Here are a few.

First, here's a single frame from a video recording of Saturn. Normally, I record about 3000 frames and use Registax to select and align the best. On this occasion – April 13 – the air wasn't very steady and my video didn't yield much. But one frame of it did have these bright spots:

That's what happens when cosmic rays hit the CCD sensor. Since these three all arrived during the same 1/25-second exposure, and there were no others all evening, we're probably looking at three particles generated by the impact of a single higher-energy particle on an atomic nucleus. More about this process here.

Permanent link to this entry

Comet 73P/Schwassmann-Wachmann

The big news among astronomers is of course that Comet 73P/Schwassmann-Wachmann has broken up. You can see the two brightest fragments in this picture, taken April 23:

The background is the constellation Corona Borealis; the picture was taken with a 105-mm f/2.8 lens on my Digital Rebel. Switching to a 300-mm lens, I got this closeup of the brightest fragment:

The stars look like streaks because two exposures were superimposed, lined up on the comet, which was moving relative to the stars.

Permanent link to this entry

If what you are looking for is not here, please look at previous months.


Search Covington Innovations   Search Web