Jeff Duntemann's Contrapositive Diary Rotating Header Image

Daybook

Descriptions of what I did recently; what most people think of when they imagine a “diary entry.”

LED Strips for Workshop Lighting

60W Bulb Wafer - 500 wide.jpg

We replaced nearly all of the incadescent bulbs in the house with LED bulbs shortly after we bought it in late summer 2015. That came to 27 60W bulbs and ten or twelve 75W floods for ceiling can lights. You can’t get incandescent bulbs in the commoner sizes anymore, and the power savings is considerable. LED bulbs have come a long way in the last few years, and the “softer” light quality levels (3000K and below, most of which were 2700K for us) approximate incandescent light close enough for what we do.

There are a couple of issues with LED bulbs. We bought Feit bulbs in bulk when we re-bulbed all the ceiling fan lights. Two years on, and the Feit bulbs are dropping like flies. I have nine dead but intact Feit 60-watters on my workbench, and two more that I dismantled to see what’s in them. What’s in them isn’t much: A voltage doubler/rectifier board that converts 117VAC to 236 VDC. There’s a disk holding nine LEDs, with two wires down to the power supply. The photo above shows what it looks like when you hacksaw the plastic globe off a Feit 60W equivalent LED bulb. It’s unclear to me whether the nine LEDs are single diodes or blocks of several fabricated together. It seems like a stretch to put 26 volts on a single LED and not see it emit plastic instead of light.

I’ve been taking bulbs apart to try and see why the Feit bulbs die so young. I’ve pulled two apart: One had a bad power supply. The other had a bad wafer. The wafer, however, was intermittent: Tap it with a screwdriver and it flickers, and sometimes comes up to full brightness. The dead power supply was just dead, without any indication of why. It stayed dead when tapped, unlike the intermittent wafer. Cheap crappy manufacturing, I’d guess. This is one Feit I will gladly walk away from.

LED lighting is problematic in ham radio work because of the broadband noise generated by the bulb power supplies. Fluorescent tubes have the same general problem. My notion is to create a separate lighting system in my workshop/ham station using LED lamps run straight-on at 12VDC. I had hoped that the wafers ran at 12VDC or 24VDC, so that I could harvest them from bulbs with failed power supplies. Not an option.

So I started sniffing around to see what sort of lamps are available for 12VDC. I bought a couple at a hamfest to play around with, and talked to the techs at a lighting store up in Kierland. They had an interesting display: Assemble-it-yourself LED strip lights. These consist of an anodized aluminum U-channel, with a self-adhesive strip of LEDs stuck to the bottom of the channel, with a linear plastic lens that snaps into the U-channel. The LED strips take 12VDC, and use 2.88W per foot.

LEDTapePackage - 500 wide.jpg

The LED tape strips come on a reel, with 16.4 feet on the reel and 3 LEDs per inch. The reel isn’t cheap; I paid $235 for it, and that’s about what they go for on Amazon. What I did was buy an 8′ section of channel/lens and put down just under 8′ of LED strip. The strips have clearly marked points every inch where you can cut them with an ordinary scissors, leaving solder pads on each side of the cut.

LEDTapeCloseup-500 Wide.jpg

The strip comes with a .215″ female barrel connector on the end of 6′ of 2-conductor cord. But again, there are solder pads every inch, so you can wire them to a power source any way you want to. What I did (at least for testing) is dig around in my Big Box o’ Wall Warts and found a Dell APAC-1 power supply, which came with a Dell matrix printer someone gave me ten or twelve years ago. It’s labeled for 12VDC @ 2A, and the strip draws 1.92A, so the supply is working pretty hard. (It squeals, way up there where Carol can hear it and I can’t.)

What I did with the strip is mount the four clips that came with the U-channel under the front edge of the first shelf above my workbench, and then clipped the U-channel to the shelf. The lower surface of the shelf is 11″ above the bench, so I’d have to do some bending and stooping to actually see the LEDs. And the intensity of the light is marvelous, shining down right where I do the sort of close work that electronics requires, which these days is real damned close.

Illimuminated Bench - 500 Wide.jpg

Because the light comes from a strip of 288 LEDs, there are basically no shadows anywhere under the strip, unless you completely block the light. This means that you can work in good light without worrying as much about your hands casting shadows on what you’re doing. It’s almost like the project photos in old QSTs, where all parts of a complicated hand-wired chassis are completely illuminated. The photo below is of a high-voltage power supply placed on the bench beneath the LED strip, and is literally what you’d see were you sitting at the bench poking at the supply.

Lit Chassis - 500 wide.jpg

This was only a first step. The strip won’t be powered by the APAC-1 long-term. I have a 30A 12VDC supply that I built out of an old minicomputer power supply somebody gave me. That will be the power source for all the lighting in my workshop/shack once the whole system is complete. I’m looking into overhead lighting now, and may simply use the remaining 8′ of LED strip on the reel with a white diffuser lens, clipped to the drywall ceiling.

Ultimately I want to work up a solar panel system on the roof of the garage, charging a 12V battery of some sort to supply lights and radios. The landscapers are going to auger a hole next to the shack for an engineered ground system tomorrow morning, so I have other work to do before I get on the air. It was a fun project, and will make working on projects a great deal easier. Good light is not optional!

A Tall Tree in a Tight Spot

IMG_2201 - 500 Wide.jpg

We had the big sumac tree by the front door cut down this morning. “Big” is no exaggeration, either: It was forty feet high, and two feet thick at the ground. (Look carefully and see Carol standing behind it.) It was a bad place for a tree that size, for several reasons. It was messy, and dropped seeds and leaves almost continuously between April and August. That was annoying, but what worried me was triggered by what happened to the guy right next door to the east of us: He had a biggish (but not even that big) mesquite tree snap in half in a windstorm and destroy the pergola over his back patio. I looked at the sumac and calculated what would happen if it lost structural integrity in any direction. If it fell to the west (toward me in the photo) well, ok. Any other direction, and it would take out one or both of two gates, part of our block wall, some or most of the guest room, and some or most of the front entrance, including our stained-glass encrusted front door.

That was a thick tree, probably as old as the original house, which is now 52 years. Some parts of the two main trunks were well over a foot thick, up higher than our roof line. I agonized over the decision, because it was a healthy tree that looked solid as a rock. But it was too close to the house, and even closer to the front gate. So we had a landscape company we knew and trusted come out and take it down. We also had them take down a much smaller mulberry tree that was not healthy. “Not healthy” is putting it mildly. See the photo of the mulberry’s main trunk, below.

20180130_075945 - Cropped - 500 Wide.jpg

Well, that had certainly been the right call.

The mulberry was quick; they had it down in twenty minutes. The sumac took the rest of the morning. The crew knew very well what was at stake, so it probably took more time than it might have, had it been growing in the middle of the back yard. It came down a chunk at a time, with each chunk tied on a stout rope and steered expertly down to terra firma. Some of the chunks were impressive.

Down, down, down. Then: Six or seven feet above the base of the trunk, the cross-section started to change. Take a look:

20180130_103838 - Cropped 500 Wide.jpg

Egad. The damned thing was hollower than the sickly mulberry. After I took this shot, I dug into the chocolate-colored stuff surrounding the void and tore huge chunks out with my fingernails. There were probably three inches of actual wood–sometimes less–forming a 15″ trunk. I wanted to yell into the hole: “Hey, any elves in there? The chipper’s at the curb. Last chance to come out, guys. Bring cookies.”

Any regrets we had taking down the tree vanished the instant we saw this. Sure, there was solid wood all the way around. But consider the lever-arm torque on the tree trunk if a really bad west wind hit the tree’s canopy. Crunch! We could have been out our front entranceway.

There’s a downside to losing that tree: It provided considerable shade to the house in the worst of the summer. My electric bills are probably going to go up.

The major lesson in all this is that we assume trees are immortal, but they’re not. Trees live for some period of time, and then they die. The typical lifespan of a sumac like ours is 30-50 years. We were already past that. The rot at its core was nothing worse than old age. I remember when I was a kid, and the cottonwood trees in the parkway on Clarence Avenue all started to die at once. The city had planted them, six to a block on both sides of the streets, when they platted the neighborhood in 1929. But once the market crashed, nobody wanted to build homes there until the last of the 1940s, when the trees were already twenty years old. By 1960, the cottonwoods were over thirty years old, which is pretty much end-of-life for that species of tree. Just about every one was hollow enough to hold a whole bakery’s worth of elves, including a few really fat ones. My sister remembers that one of them on another street fell on a house and did some serious damage, and since the parkways belonged to the city, wham! Hundreds of cottonwoods vanished in a couple of years.

A postscript: Our cottonwood was the last one on the street to go. When we saw the logs stacked up, we realized that it was solid to the core. So trees have bell curves too. Bummer.

Anyway. We have plenty of other trees, none of them (thankfully) quite that close to the house. We have a gorgeous Aleppo pine in the front yard, outside the wall, that may exceed fifty feet high. Google tells me Aleppo pines typically live for 150 years. If I ever feel the need to hug a tree, well, I’m going with that one.

Tuesday Night’s Super Blue Blood Lunar Eclipse

Very late Tuesday night (in fact, just before dawn on Wednesday morning) we’re going to have us a super blue blood lunar eclipse. If that sounds peculiar, it’s because it is–and rare. The last one, in fact, happened in 1866. The key facts about the eclipse are these:

  • It’s a “blue moon” eclipse. A blue moon isn’t about color; it’s about having more than one full moon within a single month. This happens…once in a blue moon. Most months are several days longer than the Moon’s 29.5-day run around the Earth, so if you get a full Moon in the first couple of days of a month (February being the usual exception) you’ll also get a full Moon in the last few days of the months. This second full Moon in one month is a blue Moon.
  • It’s (almost) a supermoon eclipse. A day before the eclipse, the Moon will be at perigee in its orbit; that is, at its closest approach to Earth. This is called a “supermoon.” Although the eclipse occurs a day after lunar perigee, most people say, wotthehell, and call it a supermoon eclipse.
  • There’s nothing particularly bloody about this lunar eclipse, compared to other lunar eclipses. (Because the Moon will be close to perigee, the eclipsed Moon will appear larger in the sky.) The Moon will be in Earth’s shadow, and the Earth’s atmosphere scatters the light that passes through the atmosphere, with blue wavelengths scattered more than red wavelengths. This is why most lunar eclipses leave the Moon looking brown or orange. Much depends on how clear and steady the Earth’s atmosphere is at the time of the eclipse. A very occasional eclipse occurs when the atmosphere is unusually calm and clean; such eclipses are very dark, and the Moon almost black. There’s really no reliable way to predict the color of the Earth’s shadow before the eclipse actually happens.
  • Alas for Americans, this eclipse will occur while the Moon is setting, just before dawn on 1/31. If you’re on the west coast, you’ll see almost all of it. If you’re on the east coast, you’ll see relatively little. The best view (and this is sheer chance) is over the central Pacific. Alaska, Hawaii, Japan, eastern China, and (most of) Australia are the places to be. Western Europe, most of Africa, and most of South America are pretty much out of luck.

So. When precisely will the eclipse occur? That depends completely on where you are. Your best bet is to look at the timeanddate.com eclipse page, which has a feature that will display local times for the eclipse if you search on a particular city. Look for the search box, titled “Find Eclipses in Your City,” in the right-hand margin of the page. Most cities of reasonable size will be there. If your town is too small for the search box, search for the closest larger town until you find one.

Here’s an example, for Scottsdale AZ. You’ll get local time for the eclipse begin, end, and maximum. The summary below the umbra/penumbra animation will tell you if a particular phase of the eclipse is visible at your location. In Scottsdale, approximately the last 90 minutes occur after the Moon sets, and thus can’t be seen. In Colorado Springs, the last two hours of the eclipse occur after moonset. Make sure you run the animation, which will give you a very clear picture of how the Moon’s disk crosses the umbra and penumbra of the Earth’s shadow.

Carol and I generally wake up at six or a little after, so we’ll either get up half an hour earlier than usual (no big trick for us larks) or just miss the first few minutes of totality. Now, for the owls in my audience, the better path may be to wait until January 21, 2019. That won’t be a supermoon, nor a blue moon, but eclipse visibility will be almost centered over North and South America, which means that anybody in the US will be able to see it in its entirety.

In truth, total lunar eclipses are common (the Moon is small, and the Earth is big) and in my life I’ve seen dozens. I need to emphasize that Wednesday’s eclipse is remarkable not for how it will look but for its flukes of orbital dynamics that coincide only rarely.

Still, lunar eclipses, irrespective of their orbital dynamics, are celebrations of the beautiful and extravagant universe we live in. Take a look, especially if you’ve never seen one before. Compared to a total solar eclipse, it’s long, slow, and almost meditative. Seeing it in a photo or on TV just doesn’t do it justice. Go out there and use your own eyes. You won’t regret it!

Egg++ and My USB Microscope

Something a little peculiar happened this morning. I cracked an egg into a (white) bowl for scrambling, and the albumen looked a little pink rather than clear. Blood, fersure, though I already knew (I don’t know why) that blood in an egg doesn’t necessarily mean that the egg was fertilized. However…next to the yolk was a little brown thing about 3/8″ long. It was about the right shape for an embryo, but it was too small to pick out any details. So…

…I cranked up my new USB microscope, which I got from Carol for my 65th birthday this summer. Worked like a champ:

minisee0002-500 Wide.jpg

I’m no expert in chicken embryology, so this is still a guess, but I’ve never seen anything like it in an egg before. The pink in the albumen suggests blood, after all.

Here’s the setup I used to take the photo, which will show you the microscope and its focusing stage:

microscope-500 wide.jpg

Like any reasonable optical microscope (we have one from Carol’s college years in biology) it has a coarse focus (the knurled column attached to the metal base) and a fine focus on the end of the camera tube itself. It plugs into any USB port and draws whatever power it needs from the port.

The device shown above costs $77.95 from Amazon.

If you don’t think squicky blobby things do the instrument justice, here’s something on the hardware side. This is a surface-mount LM386 audio amp, measuring just a hair over 3/16″ long:

LM386SM0002.jpg

The imaging software I’m using (free) is called MiniSee, and it works tolerably well. Other packages exist, and as time allows I’m going to try them.

The real challenge with the microscope is lighting. Lighting makes a huge difference in the quality of the image coming in from the sensor. There are eight white LEDs in a circle around the sensor, with a brightness control built into the USB cable. These work well for looking into dark places (like the back of my mouth) but don’t do well with objects lying on the metal stage. A flat black background is useful, especially for metallic objects. I intuit that some sort of small gooseneck desk lamp would do the trick, and I’m looking.

The instrument comes with a number of plastic probe tips for looking at your ear canal, up your nose, and, well, where the sun don’t shine. The mini-CD wouldn’t spin up on my quadcore, and as it turns out I didn’t need it, given MiniSee. (One of the reviewers on Amazon claims it’s all in Chinese, anyway.)

Overall, I’m more than pleased, especially for something in the $75 price class. There may be better ones. I see quite a few on Amazon. But this one will do.

Oh…I scrambled the egg and ate it, once I picked out the embryo. What’s a little chicken blood when dinner generally hits the table medium-rare and still dripping?

A Tappy Kind of Life, Re-Examined

I did a really dumb thing a few days ago: I was hosing off the pool deck, and fell in. With the water at 83 degrees and outside temps at 106 that would ordinarily have been a welcome break…except that my Samsung Note 4 smartphone was in my gym shorts pocket.

I tried hard not to hit the water, and bruised up my left arm a little in the process. However, the phone was underwater for a few seconds (more than five, less than ten) and has not yet come back to life, even after several days in a ziplock bag with all the dessicant packs I could scrounge around the house. This is a serious bummer. I liked that phone. Carol has one too, and in a number of ways, it changed the way we live.

We bought the Note 4s in November 2015, and came to love them almost immediately. They were part of the process of moving from Colorado Springs to Phoenix. We’d had Droid X2 phones since 2011, and used them as…phones. They were good workaday phones, granting that we had a landline in Colorado and used it for talking to relatives or any time a conversation was expected to take longer than a couple of minutes. Although we’d expected to get a landline in Phoenix, a few weeks of using the Note 4s showed them to be so effective that we just didn’t bother. Carol bought a Bluetooth headset for long conversations with her sister, and mostly I just put it on speaker. The fidelity was superb, and there was a lot less packet-loss than with the Droids.

What startled me about the Note 4s was how much else they could do. I’d tried texting on the DroidX, but the screen was too small and my fingers too big. On the phablet-sized, stylus-equipped Note 4, no problem. I had tried reading ebooks on the DroidX, and again, it just wasn’t big enough to be comfortable. I marvel at how well the Note 4 handles the Kindle app. I have a Kindle Paperwhite with a bigger display, but because it’s another slab, I mostly use it at home. If I’m waiting in a doctor’s office or somewhere, the Note 4 serves spectacularly.

Then we started trying some apps. Two that we use a lot are Raindar and Weather Underground. Raindar shows where the rain is, how hard it’s falling, and which way it’s moving. Period. That’s all we wanted, and that’s all we got. Win! Weather Underground is more complicated: It’s a formerly independent weather geek site (founded in 1995) that was bought by The Weather Channel and somehow hasn’t yet been turned into a global warming shill operation. Its magic lies in its architecture, as a network of “hyperlocal” weather stations. The app can determine which one is closest to your house, and when you bring up the app, it will show you data from that station. Phoenix alone has thirty or forty such stations, a couple within a mile of our house. I was a bit surprised at how different the readings were from one station to another, but I tested the closest stations against my own thermometers, and chose the one that was the best match. We use it to find out when to open the windows on summer nights, and when to close them again in the morning. We use Raindar to see when there’s a lull in a thunderstorm long enough to let the Pack out to potty. Weather matters.

After getting stuck in traffic once too often in Denver, I searched for and discovered Waze, which crowdsources data on traffic conditions and lays it all out on a map. It knows (from the phone’s GPS logic) how fast you’re going, and it plots bad traffic in different colors. Users report construction, accidents, vehicles on the shoulder, and speed traps, all of which also appear on the map. We did a lot of driving between here and Colorado Springs from 2015 to 2017, and Waze was surprisingly helpful.

The local classical music station, KBAQ, has an app that will stream their audio if you have an Internet connection. That’s useful. However, what’s even more useful (especially since we listen to KBAQ on a real stereo system when we’re at home) is that they tell you what’s currently playing, so if an unfamiliar piece comes on, I can yank out the phone, tap up the app, and see what it is, following the links on the composers if they’re new to me. I’ve had no better education in classical music since the course I took with Dr. Raymond Wilding-White in 1973.

The Note 4s came with Flipboard, a news aggregator app that I doubt I would have sought out on my own. We don’t have cable TV anymore, and although I’m not a news hound, I generally like to know how close the fires and riots are. Alas, Flipboard seems to emphasize UK news sources, and has an almost inexplicable obsession with celebrity trivia, particularly celebrity women showing off their baby bumps or going topless somewhere. Most of these celebrities are people I’ve never heard of, and even when a genuine celebrity appears, the context is, as often as not, banal. There have been a few reasonable science and tech stories on Flipboard, but mostly it’s catch as catch can. I check it most mornings to make sure that the world still exists. (Given the stuff they post, sometimes it’s a little hard to tell.)

Samsung’s S-Health app does a lot of different things, the most useful of which is to track steps, pedometer-style, and present step data in various reports. It also uses the camera to test blood oxygen levels, but the software is fussy and my cheapo pulsox does the job much better. As a pedometer, though, it’s first-rate.

I like GPS Test, especially since it can tell me my current altitude. It also works as a compass. The SoundHound app is much less useful, and I’m being charitable. I tried and generally dumped a number of games, most of them puzzle games. The screen is a little small for Mah Jongg, given the complexity of the patterns on the tiles. However, Ultimate Jewel (a Bejeweled clone) is a sort of software fidget-spinner that handily gets my mind off of vexatious people and their damfool drama.

I have a flashlight app that’s been useful a time or two. The camera is decent, though it has nothing on my Canon G-16. It’s a reasonable photo viewer, especially with a 128GB Micro SD card inside to hold my photobase. (The lack of an SD slot is primarily what kept us from buying Note 5s.)

However, all that said, the biggest single use that Carol and I put the Note 4s to is voice search. We had voice search on the Droid X2s, but somehow it just didn’t comprehend us as cleanly as the Notes. Now, if we’re sitting around talking about something and an unfamiliar concept or person comes up, one or both of us grabs our phones, taps “Google Voice Search” and speaks the search terms. It’s still a little astonishing how reliably the app understands spoken search terms. Granted, a 5 1/4″ diagonal display is on the small side for doing research, but for quick orientation, well, it’s like nothing else.

Five days since I went swimming with my Note 4, I miss it terribly. This has nothing to do with Facebook or Twitter, which by agreement with myself I only use on my desktop. The strange part of the adventure is that I integrated the Note 4 with my life so slowly that I never fully grasped how important it had become. Some things come at you fast. Others sneak up on you when you aren’t looking. A very few just have this talent of dissolving into the background noise of ordinary life, where you never miss them until they go away. So it was with the Samsung Note 4.

I ordered a new Note 4 from Amazon a few days ago, and it should be here by Monday or before. It’s white, which is a feature, since the two phones Carol and I bought are both black and physically identical. To know which it is, you have to hit the button and parse the wallpaper. Now we can tell from across the great room.

There are rumors that Samsung is preparing a Note 8 for release later this year, with an SD card socket if not a replaceable battery. I’ll give it a fair hearing, but in truth, if it has no strong advantage over the Note 4, I may give it a pass, especially if the rumors are true that it will cost $800. I promised Carol I would no longer skim the pool with a $400 phone in my pocket. An $800 phone? No promise necessary.

Ten Gentle Opportunities in Trade Paperback

10-Gentle-Opportunities-Revision-Final-Adjusted-With Type-500 Wide.jpg

I’ve been promising to do a trade paperback edition of Ten Gentle Opportunities for over a year now. Printed books are always good to have around for promo purposes, but I’ve gotten eight or ten explicit requests for paperbacks since the ebook edition was first released in January 2016. Why disappoint customers?

Buy Ten Gentle Opportunities from the CreateSpace store.

Buy Ten Gentle Opportunities from the Amazon store.

Sorry it took so long, guys.

Anyway. Why two sales links? It’s yet another peculiar kink in the increasingly kink-y world of independent publishing. Simply put: I make significantly more money per sale on books ordered from the CreateSpace store than from the conventional Amazon store. I’ll lay it out for you, though you can calculate it yourself using the CreateSpace royalty calculator, with a detailed explanation of how it all works on their Understanding Royalties page.

The book’s specs are these:

  • Black and white interior
  • 6″ X 9″ trim size
  • 310 pages
  • $12.95 Cover price

Basically, my share of the book’s cover price is the cover price minus the portion that CreateSpace takes. Their share is the sum of three things:

  • The sales channel percentage
  • A fixed per-book charge
  • A per-page charge

The sales channel percentage is basically the retailer’s discount. There are four sales channels available through CreateSpace, each with an associated discount:

  • Amazon US: 40% of cover price
  • Amazon Europe: 40% of cover price
  • The CreateSpace store: 20% of cover price
  • Expanded distribution: 60% of cover price

Expanded distribution is basically retail wholesaling to B&M stores and libraries through distributors like Ingram, Baker & Taylor, and NACSCORP. As you can see, orders coming in from Amazon take twice the amount off the top as orders coming in from the CreateSpace store. I get so little from each expanded distribution sale that I decided not to both with expanded distribution. Sure, it would be cool to see the book on the shelves at bookstores…but the chances of that happening at all are pretty slight.

The fixed per-book charge is a sort of minimum charge for manufacturing the book. For b/w books having 110-828 pages, the fixed charge is $0.85 per book.

The per-page charge is the rest of the manufacturing cost, and depends on page count and whether the interior is b/w or color. For a b/w book in the 110-828 page count range, this charge is $.012 per page; i.e., 1.2 cents per page.

Turning the crank, it comes out like this:

  • $12.95 X 20% = $2.59, calculation of channel discount
  • $12.95 – $2.59 = $10.36, cover price minus channel discount
  • $10.36 – $0.85 = $9.51, minus per-book fixed charge
  • 310 pages X $0.012 = $3.72, calculation of per-page charge
  • $9.51 – $3.72 = $5.79

My share of each sale through the CreateSpace store is $5.72. For a sales through the Amazon store, the channel charge is 40%, or $5.18. With all else being the same, my share would be $12.95 – $5.18 – $0.85 – $3.72 = $3.20. So by ordering through the CreateSpace store, I get $5.72 rather than $3.20.

However….there is a significant gotcha: You have to set up an account with the CreateSpace store. Also, Amazon Prime shipping does not apply to CreateSpace sales. I recognize that these may be show-stoppers for some people. That’s ok; I won’t be annoyed if you order from the Amazon store.

Mostly, I wrote this entry to provide a little insight as to how authors are paid for paperback editions of books offered through CreateSpace. Because I don’t expect to sell a great many copies of the paperback, it’s a matter of no great importance. Like it or not, we’re hurtling toward an ebook future at most of the speed of light. The ebook is $2.99 and it’s delivered Right Damned Now rather than sometime next week. The ebook is selling well (considering I haven’t been pushing it much) and I’m happy with the money I’m making. Even $3.20 per copy is about par for royalties I’ve received on traditionally published technical books, and this is fiction.

If you still like printed books, I’d be honored if you’d buy a copy. And on that note, I’m going back to writing my latest novel. There are worse ways to be retired than this!

A Two-Arm Monitor Stand for the Raspberry Pi

Two Arm Mount on Steampunk Table-Cropped-500 Wide.jpg

If what I’ve heard is true, most Raspberry Pi installations consist of the naked circuit board lying atop a nest of wires on the desk behind a monitor. I think it’s true; that was certainly my Raspberry Pi installation for a long time. Now, I’ve decided to use my steampunk computer table as my Raspberry Pi 3 workstation. And I got an idea: Use one of those VESA-standard 2-arm monitor stands that clamps to the edge of a desk without any drilling into or other hacking-up of the desk. One arm holds the monitor, and the other holds the Raspberry Pi itself.

The trick is to buy one of several Raspberry Pi cases that includes a flange with VESA 75 or VESA 100 holes. VESA is a standard for TV and monitor mounting hardware. Its two smallest configurations are a 75mm square, and a 100mm square. Most modern flat-panel TVs and many monitors have threaded holes on their back faces arranged in one of the several VESA configurations. I’m pretty sure (having looked at a lot of monitors and TVs in the past few years) that the 100mm configuration is the commonest. It’s the one on the Dell 1907fp monitor that I’ve been using for Raspberry Pi boards since the beginning. VESA-compatible displays generally use metric screw threads in the mounting holes, with M4 the standard for the smaller configurations, including 100mm. M4 screws can be had at Ace Hardware, and probably also at Home Depot and Lowe’s. (I go to Ace first for such things.)

The Raspberry Pi case that I used is this one:

RPi VESA Case.jpg

It has four little wings with both the VESA 75 and VESA 100 holes. The holes in the wings aren’t threaded, and easily pass standard 8-32 machine screws, which I used to hold the case to the second arm of the monitor stand. I oriented the Raspberry Pi with its USB ports on top, so I can reach over the monitor and plug in peripherals or thumb drives easily.

This approach isn’t limited to the Raspberry Pi. There are VESA cases for the Intel NUC (Next Unit of Computing) boards, and most of the higher-end embedded boards like BeagleBone. On a small table like the one I made, there’s not a lot of flat space to park a case of any size, so whatever computer I’m going to be using on it should be able to hang on that second monitor arm. The arms on the unit I bought can hold up to 12 pounds each. Most of the small-form factor Dell machines I use are that weight or lighter. Dell’s Micro 3000 series has an optional VESA bracket, and brackets for other models may be available. And hey, you guys could rig something, right?

Dipping Back Into Delphi with List & Label 22

I haven’t done a lot of programming for the last couple of years, and I miss it. Interstate moves and oxygen starvation will do that to you. I’ve converted some of my old Delphi apps to Lazarus, which in truth wasn’t hard and probably can’t be called programming with a straight face. And I have a project that I need to get back to, even if it has to be written in Delphi 7, which is the most recent version that I have. (Turbo Delphi doesn’t count.) I no longer had a publishing company after Delphi 7 appeared, so post-2002 I dropped off their reviewers list. And $1,400 is a little steep for hobby programming–much less $4700 on the high end.

For some years I’ve been poking at the concept of a personal medical database. I’m old now (how did that happen??!?) and I take pills and get bloodwork and monitor various things to make sure none of my component parts are rusting out. I have Word documents full of notes, and scribbles on paper calendars, all of which really need to be pulled together into one searchable and reportable database. Some doctors won’t believe that my blood pressure does not respond to sodium. I have proof. I’ll bet, furthermore, that it will be a lot more convincing if it’s placed in their hands as a professional-looking report.

All of what I’ve done so far has been in Lazarus, and most of that has been small proof-of-concept lashups, none of them newer than 2012. However, a marvelous report generator product has crossed my desk, and I want to give it a shot with my medbase app. The product is List & Label 22, from Combit, a small firm in southern Germany. It has God’s own kitchen sink of features, many of them related to Web programming, which I simply don’t do. However, it has all conventional reporting options I’ve ever heard of well-covered, and it supports all versions of Delphi back to D6. (It supports Visual Studio and many other dev platforms as well.)

It doesn’t support Lazarus, alas. So I’ll be trying it out in D7.

The big win (for me at least) is that L&L 22 provides a report designer in VCL component format that drops on a form and becomes part of your application. This allows end users to design their own reports. Given that my end user is me, I don’t have to worry about end users doing gonzo things. I’ve always liked my software to exist as One Big Chunk (DLL hell, and all that) so this is right up my alley.

I don’t yet know precisely what reports I’ll want, and it may be the case that I won’t know until I actually need one for a specific purpose, like laying out my data indicating that salt is irrelevant to my blood pressure. Having a report designer right there in the app means that I can design the report that I need when I need it, and not try to anticipate every damn thing I’ll ever want while I’m building the program itself.

I should make it very clear here that I don’t dislike modern Delphi. I still love it, but it’s gotten enormously expensive, and the Starter Edition does not include database programming features. My other reason for using Lazarus is that I still intend to write intro-to-programming books using Pascal as the teaching language. Expecting students to pay even $250 for the Delphi Starter Edition is asking a lot, and worse, I intend to teach database work as well as conventional programming.

I’ll have more to say about List & Label as I learn it. Ditto the medical database itself, which is now a set of tables full of test data and a couple of conceptual UIs. Stay tuned.

Using Ancient Software Under Win7

I had a need to print out a few calendar pages today, and after thinking about it for a second, I realized that there was a calendar program of some sort on a pile of ancient CDs in the closet that I had not yet dumped. I had actually thrown out a lot of CDs already because they failed to run under Win7, including a few things that I had sorely missed at upgrade time, like the software that came with two of my three scanners. (Thanks to God and all the fates that there is a VueScan X64 that understands both scanners.)

I dug in a box and there it was: Broderbund’s Calendar Creator 7, copyright 1999. It indicated that the software was for Win9x and NT4. I never remember installing it, and honestly don’t recall how it came to me. I have accepted dead or dying computers from other people who had no idea where to take them and didn’t want to just put them out on the curb. Some of these machines came with cardboard boxes full of odd and often broken stuff, with an alluvial layer of software CDs on the bottom. I’m guessing that this was one of them.

So hey, wotthehell: I popped it into my Core 2 Duo/Win7 lab machine, fully expecting it to fail to either install, run, or both. It installed. And it ran.

Once.

However, during its first run it actually created a calendar for me, and printed several pages as a test. I edited the page layout a litle bit and printed the three months that I needed. Then I closed the program and went on to other things. A couple of hours later I realized I needed one more month, but when I tried to run it again, it croaked.

I would have shrugged and tossed it, except that it did run once, launched by the installer as soon as the installer had finished with it. Hmmm. Since installers have to have admin permissions to do their jobs, this made a certain amount of sense, and suggested that if I could run the app as admin, it should work. I gave it admin permissions. It worked.

The 16-bit colors look a little weird, but it runs full-screen at 1600 X 1200 with only one glitch: Print preview doesn’t quite reflect reality. Screens that big didn’t exist in 1999, so I can forgive it that much. It did the job I needed, didn’t cost anything, and as best I can tell didn’t mess anything else up. (That’s not universal, and it’s why I always install things first on a lab machine or a VM. At least do a restore point before you install weird stuff like this.)

If you ever find yourself in this situation, here’s how you run old software as admin:

  1. Right-click the shortcut to get the context menu.
  2. Select Properties.
  3. On the Shortcut Properties dialog, select the Shortcut tab.
  4. Click the Advanced button.
  5. Check the “Run as administrator” checkbox.
  6. Click OK.

Keep in mind that not all ancient software will be this cooperative. A lot of old stuff won’t run at all, or even install. However, it’s useful to try running it as admin before you flip the CD into the trash.

I may still go looking for a modern calendar program. However, it was a good memory jogger, and made me acknowledge that whoever wrote that thing did a very good job of anticipating the future. This is not universal programmer behavior, trust me, and I am not exempt: A DOS program listing utility I wrote in Turbo Pascal in 1985 or so would run in a DOS box in NT4 and Win2K…but after 1999 passed into history, it labeled the printouts as occurring in the year 19100. So it goes.

Doing the Numbers on CreateSpace POD

TGO Sample Layout Page - 500 Wide.jpg

I’m hard at work on a print edition of Ten Gentle Opportunities. Several people have asked for one, and it’s something I’ve been meaning to do for the last six months or so. On the surface it’s easy enough; I’ve done many print books in the past. This time I got seriously tangled up in a critical issue: How many words should I attempt to put on a page?

It’s a critical issue that doesn’t come up at all in ebook layout, where fixed-length pages don’t really exist. (That is, unless you’re distributing PDF files, which almost no one does for fiction anymore.) The problem is that there is a fixed cost per page for POD books, so the bigger the type, the greater the page count, the higher the unit cost, and the smaller your profit margins. The page shown above may look dense, but it’s about par for trade paperback fiction from traditional publishing houses. Bigger type or greater leading would mean a longer book and a higher unit cost. In this entry I’ll try and explain how that calculation is done and what it means to your bottom line.

I’m not done with the layout yet, but a castoff (length projection) falls somewhere close to 300-310 pages. Unit costs add up this way: CreateSpace (Amazon’s POD division) charges $0.012 per page, plus $0.85 per copy, making the unit cost $4.57 for a 310-page book. As best I know, the unit cost doesn’t vary depending on the page size. More on this later.

Now, that’s just for the unit cost. There’s another factor that isn’t present in all POD systems, particularly Lulu.com, where most of my POD titles are currently hosted. This is the sales channel charge, which amounts to Amazon’s profit margin on the title. Adding to the confusion is that there are two different percentages for the sales channel charge, depending on how the customer ordered the POD book:

  • When customers order the book through Amazon.com, the charge is 40% of cover price.
  • When customers order the book through the CreateSpace e-store, the charge is 20% of cover price.

The CreateSpace e-store provides a page for each book. You basically earn the smaller sales channel percentage by driving buyer traffic to the book’s link on the e-store. I’ve never tried this so I don’t know how many sales I can steer to the e-store. I guess I’ll soon find out.

In terms of knowing how much you earn for each copy, then, you need to set a cover price and then calculate the channel sales charge for Amazon vs. the CreateSpace e-store. Let’s use $12.99 as a cover price example here:

  • For Amazon, you multiply 12.99 X 0.4 = $5.20. Knock $5.20 off the cover price and you get $7.79. Out of that value comes the unit cost of the book: $7.99 – $4.57 = $3.22 as the money you clear on each sale.
  • For the e-store, you multiply $12.99 X 0.2 = $2.60. Knock $2.60 off the cover price and you get $10.39. Subtract the unit cost of the book: $10.39 – $4.57 = $5.82 as the money you clear on each sale.

You don’t have to do the math manually like this; CreateSpace has an online calculator. I just wanted to show you how the calculation works.

That’s a significant difference, and my guess is that Amazon is trying to provide an incentive for actively marketing your POD books. Keep in mind that you don’t choose one sales channel or the other. Your book is present on both stores at the outset, and your sales will be a mix of both. Your challenge is to get as many people as possible to order through the CreateSpace e-store.

The other way to boost your royalty value is to use a larger trim size. I’m laying the book out as a 6″ X 9″ trade book because that’s a very common size for fiction and it’s what I’ve used on all my other POD titles. Now, the unit cost doesn’t vary by trim size, but a larger trim size (holding the type size and leading constant) will hold more type per page and thus give you fewer pages and a (slightly) lower unit cost. I played around with this and decided that the minimal difference isn’t worth altering my standard layout template.

You could, of course, raise the cover price. Be careful: Readers who have come to expect ebooks to cost $4 or so might consider $12.99 off-putting. In fact, I consider $12.99 to be something like a maximum for a trade paperback novel by an unknown, and I may drop that to $11.99. Pricing is a black art, alas.

So there it is: You sell a POD novel for $12.99 and you get some mix of $3.22 and $5.82 per sale. That’s modestly more than you’d get for the Kindle ebook version priced at $3.99, and close to what you’d get for the same ebook at $4.99. (I don’t think this is an accident.) Is it worth the trouble? I don’t know. Indie authors I’ve talked to say they like having a physical book to show around, but they really don’t sell many compared to the ebook edition.

I’ll admit: I’m doing it because I enjoy book layout and I’m good at it. The schedule isn’t clear yet. I’m still wrapped up in house issues. (Health insurance too; right now my insurance agent tells me there are no individual policies for sale in Maricopa County, as bonkers as that sounds. There may be some by November. Nobody knows yet.) I’ll certainly launch the print edition here when it happens.

The key point is that if you can’t lay the print edition out yourself, you may lose money on it, and sticking with ebooks could be the most prudent choice financially. Do the math and sleep on it. This can be a very weird business.