Unbecoming Levity

Pies Pies Pies Yeah

Abacquer — Sun, 15 Mar 2009 14:30:51 -0400

So I celebrated Pi Day with some friends yesterday and had a nice time. The pies and treats were delicious. Sadly I didn't bring a cherry pie for Julie, I will have to remember to do that next year. The blueberry pie was great, the chicken pot pie was delicious, and though the glutin-free lemon meringue was very runny, it still tasted fine.

We played games (I had a nice long game of Carcassonne, and other people played Bohnanza), made up silly jokes and songs, and just had a very nice afternoon. I got to meet some new friends, and spend a little time with some old ones. And I ate WAY too much pie.

This morning my idle mind wrote a little song parody based on the day which I decided to share here:

You made me blueberry,
Apple and pecan;
Wet meringue with lemon goo
And pot pie tastes so grand.

(The bigger, the better)
Our tummies fully crammed...
Collected from around the world
we'll eat them if we can!

Pies pies pies yeah!
Pies pies pies yeah!
Pies pies pies yeah!
Pies pies pies yeah!

The pound cake was delicious,
And whoopie pies were fine.
But poor Julie got no cherry,
What a waste of time.

Big ones and small ones,
Look lovely to our eyes,
Tantalizing on the plate,
Not so much on our thighs...

Pies pies pies yeah!
Pies pies pies yeah!
Pies pies pies yeah!
Pies pies pies yeah!

Go back to your larder,
Fetch me another plate,
My pancreas is still alive,
And I'm not the sort to wait.

The bigger, the better,
And some are glutin free,
So what if you don't eat your share,
More for me and me and me and me and me and

Pies pies pies yeah! (We're gonna eat you.)
Pies pies pies yeah! (Not too discreet to.)
Pies pies pies yeah! (Fat like a B-2.)
Pies pies pies yeah!

Oh you know I know...

Pies pies pies yeah!
Pies pies pies yeah!
Pies pies pies yeah!
Pies pies pies yeah!

Oh you know I know...

Pies pies pies yeah!
Pies pies pies yeah!
Pies pies pies yeah!
Pies pies pies yeah!

Attention Executive and Legislative Branches -- I WANT SINGLE PAYER HEALTH CARE

Abacquer — Wed, 11 Mar 2009 09:08:30 -0400

In a single payer system, the cost of healthcare drops dramatically. Yes it's true taxes go up, but at this point? They're going up anyway. The economic stimulus our country needs is going to have to be paid for, and we are all going to pay for it. Combine this with the 30 years this country has frittered away when it should have been building a new energy grid, green technologies, maintaining infrastructure, and so on, and it's clear, the party is over.

And by "the party is over" I'm referring to those artificially low taxes that our conservative friends still think are too high.

I am sick and tired of listening to gloom-and-doom naysayers fiddling about long lines, delays, and taxes while Rome burns. I have friends who are suffering right now simply because they cannot get the help they need due to lack of coverage.

Single-payer health care works in many countries around the world, some better than others. I have enough faith in American ingenuity that we can make it work here. But it does mean (AFAIC) the end (or serious curtailing) of health insurance companies that have risen to wealth on their two core tenets--1. collect premiums, and 2. don't pay benefits.

For those who quake in fear at the notion? We tried it your way for decades and it sucks. I'm ready for something new.

Peter Schiff on FOX News in 2006 and 2007

Abacquer — Thu, 04 Dec 2008 06:37:35 -0500

YouTube video of loons at Faux News laughing at Peter Schiff as he tries to warn of impending financial disaster in the 07-08 timeframe:

Peter Schiff Was Right 2006 - 2007 (2nd Edition)

What stock picks were the FOX tools recommending? WaMu, Merril Lynch, Goldman Sachs... financials basically.

NYT on Nate

Abacquer — Mon, 10 Nov 2008 11:57:37 -0500

Here's a nice article in the New York Times about Nate Silver of FiveThirtyEight who's predictions about the 2008 presidential and senate races were so amazingly accurate. Hat tip to my friend Bruce who brought the article to my attention.

Looking Iffy

Abacquer — Mon, 27 Oct 2008 13:03:19 -0400

Yeesh. Things are starting to look a little iffy in my industry. So far my work has been unaffected by the instability in the markets, but I am beginning to hear from other developers that potential clients are becoming nervy about starting new development projects. Hopefully that won't hit us, but frankly I've been expecting it to. Usually software houses take a hit when the market looks uncertain.

Anyway, I have a lot to do and I am heads down on it. This post (and my previous 3) have been posted via "moblogging"--I send an e-mail to the blog provider and they turn it into a post. Still planning on doing a blogathon on Tuesday November 4 to cover the election. I encourage all my friends to drop me an e-mail with their experiences at polling stations.

Designing the Fun Away

Abacquer — Thu, 24 Jul 2008 10:52:38 -0400

Ever have a challenging pastime to which you applied analytical skills and ended up making it so easy it ceased to be fun? This happens to me occasionally. A couple years back we had a sort of "puzzle craze" in the United Stated over "sudoku", a sort of numeric logic puzzle originating from Japan.

The basic puzzle is a 9x9 box of squares in which some of the squares have been filled in with numbers. Your assignment is to fill in the rest of the squares by following 3 simple rules: (1) each number can appear only once in each row, (2) each number can appear only once in each column, (3) each number can appear only once in each "major square". The major squares are nine 3x3 boxes superimposed over the 9x9 puzzle in a non-overlapping fashion (so the first major square is formed by columns 1, 2, and 3 where they pass through rows 1, 2, and 3, the next at the intersection of columns 4, 5, and 6 with rows 1, 2, and 3, and so on.) Since every row contains 9 squares, and every column contains 9 squares, and every major square contains 9 squares, the only numbers you will ever add to the puzzle are the numbers 1 to 9.

There are larger and more complex variants on the basic sudoku, but the above is the most popular and familiar variety. I did sudokus for awhile but eventually tired of them and moved on to new pastimes.

One reason I tired of them is that a large part of sudoku is simple, repetitive, busy work. You eliminate the possibilities for each square until a solution pops out for one square and then you fill that in and go back and reevaluate the possibilities for the other squares. For example, a given square can be any number between 1 and 9. But if you look at the column that square is in and find 1, 5, and 6 are already used, those are no longer possibilities for that square. If the row that square is in contains 2, 3, and 7, then those possibilities are also eliminated. Further if the major square it is in already contains 4 and 9, then those possibilities are eliminated too, leaving only 8, so in this case the square must be 8.

It's much more interesting when you start getting into the more complicated bits of reasoning that involve multiple squares. Like for instance if a major square is filled in except for two squares, then those two squares have only two possibilities. Let's say the possibilities are 5 and 9. Further assume that these two squares are in the same row (or column). Immediately we can say of all the other squares in that row (or column) that none can be 5 or 9 since those two options have to be used up by those two squares. But getting to that point involves getting through a period of tedious repetition.

But when you are stuck in a cabin in the rain for a day, away from the amenities you typically would enjoy, even a tedious pastime is welcome. However, I've found one thing consistently true among software engineers: they generally hate doing tasks by hand which would be better done by a simple computer program. So I was plonking along with sudoku puzzles, eliminating options by hand when I said "this is stupid, I could make a simple spreadsheet macro to do this for me."

So I built a spreadsheet that would show me the remaining possibilities for each row and column as a row or column header. Then when I was eliminating possibilities for major squares I said "this is stupid, I can use the same macro to do this for me." So I added a bit to my spreadsheet to show me the possibilities remaining for every major square. So now for a given cell I could compare the possibilities for its row, to the possibilities for its column, and the possibilities for its major square and the intersection of those three would represent the possibilities for the cell. (For example, if the cell is in a row with 4,5,6,7 available, and a column with 2,3,5,7,8 available, and a major square with 1,5,7,9 available, the possibilities for that cell are only 5 and 7, because that is the intersection of the three sets of possibilities.)

I suppose you can guess what happened next. "This is stupid, I can write a simple function to take the intersection of these possibilities."

One other basic sudoku-solving technique is to look for "loners". A loner is the only square in a row (or column or major square) that includes a particular number as one of its possibilities. For example if you have four unfilled squares in a major square and their possibilites are (2,3,4), (2,3,5), (3,5), and (2,5) you can say the first one has to be 4. Even though 2 and 3 are possibilities for this square, since it is the only square that can be 4, then it must be 4. Finding loners is also tedious, and you can imagine what I did as a result. "This is stupid, yadda yadda yadda..."

So now I had a spreadsheet that would eliminate all the impossible values for a square based on simple cross checking and which would identify loners for me. But I found as I was entering puzzles that this solved so much of the puzzle automatically, there was very little left for me to do. The spreadsheet would notify me when I caused an error (forced the possibilities for a cell to dwindle to nothing) and in those cases where I needed to make a guess to explore one possible solution, the spreadsheet would allow me to snapshot the puzzle at that point so I could get back to it if the puzzle didn't work out. With these features available puzzles rated "difficult" were taking under 2 minutes to solve. After 5 or 6 of them I shut Excel done and was done with sudoku for the day. Why bother? It was too easy now.

The pastime had become writing the code to perform the tedious operations. Once that was accomplished I was no longer interested in performing those operations. Go figure.

Silent Sky - A Thought Experiment

Abacquer — Wed, 28 May 2008 12:57:30 -0400

This is an interesting thought experiment and it involves statistics and some figuring, so I welcome any of my smart friends or anybody who cares to think about it to chime in here.

Assume that sentient civilizations with the ability to transmit some form of radiant communications signals (radio, light, whatever) started appearing in our galaxy as early as 4.5 billion years ago (when the Earth was forming), and assume that they continue to appear (each on its own evolutionary trajectory) for another 4.5 billion years. Further assume that over this timespan of 9 billion years a total of one million different intelligent civilizations begin transmitting radiant communications. After a random time period of 100 to 1000 years each civilization stops transmitting radiant signals, either because they move on to a better technology that does not require radiant communication (pretty much necessary for interstellar communications) or the civilization dies out. Finally given that our galaxy's diameter is about 100,000 light years, assume that any of these civilizations when transmitting would be at some random distance from us which does not exceed 100,000 light years (yes I know that's a little big since we aren't at the very edge of the galaxy, but bear with me.)

Now, here we are, smack-dab in the middle of that timescale, halfway between year 1 and year 9 billion. We turn an array of radio telescopes to the sky and listen in all directions simultaneously, constantly, for a period of 1,000 years.

Is our sky silent? Or is it noisy? With a million civilizations out there at some point or another, all sending signals out at different times, what are the chances we would hear none of them at all even if we listen for 1000 years?

I threw together a crude simulation last night to try and answer this question, and I have a result which you might find surprising, but I'm curious to know what other people think first. Care to venture a guess? Or perhaps create your own sim and see what results you get? I'll post my results tomorrow.

Lens Cults and The Isoceles Field

Abacquer — Fri, 28 Dec 2007 19:14:59 -0500

So early this morning I couldn't sleep and I ended up writing an article on flickr in one of the many "what lens should I buy" discussions that goes on there. I have noticed, in my days on flickr, that certain pieces of equipment and certain techniques have a following, and often get recommended simply because of the following rather than because the equipment/technique is actually suited to the purpose of the person asking. In an effort to supply a counterbalancing opinion, I found myself in need of trigonometry.

The Argument

One such "cult" item is the "nifty fifty" (the EF 50mm f/1.8) lens made for Canon EOS cameras. It is very sharp, very fast, and very cheap ($80). If you are on a tight budget (or even if you aren't) it makes sense to have one for your EOS camera unless you have a better 50mm prime, or don't need a 50mm prime.

While I will not argue that it is probably one of the best value-for-money lenses, it is not versatile at all, and yet it seems to get hailed as a magic-bullet lens. I regularly see people making claims like "it never comes off my camera". And after having used it myself, I can only conclude that these people shoot one type of thing and one type of thing only, or it never comes off because they don't own any other lenses.

There is a certain love affair with the 50mm focal length because it was the standard focal length for 35mm film for decades. But in the age of digital SLR cameras, things are different for the less expensive consumer DSLRs. These DSLR's tend to use an image sensor that is smaller than 35mm film. The APS-C style sensor, or crop sensor, does not render the entire image cast by a standard lens, but only a smaller piece in the center. This results in an apparent magnification factor of 1.6. Hence if you put a 50mm lens on a crop-sensor camera, it's like working with an 80mm lens (50 x 1.6 = 80). The end result is a smaller-than-expected "field of view" (FOV).

On an old Canon 35mm film camera, a 50mm lens has a FOV of 46°. But on a crop-sensor camera the FOV is a hair under 29°. This loss of over a third of the FOV means that on crop-sensor cameras the EF 50mm f/1.8 lens has distinct limitations as to how much you can fit in the frame.

Fortunately, on my EOS 5D the 50mm behaves as expected. Because the 5D is a full frame camera, its sensor is the same size as a 35mm film frame. So I get 46° out of my EF 50mm, just as nature intended.

The "nifty fifty" on crop-sensor cameras is often described as a "portrait lens". With the crop factor, the 50mm lens behaves like an 80mm lens, and 80mm is ideal for portraits. But if you want to shoot anything larger than a head-and-shoulders portrait with the EF 50 1.8 on your Rebel XT or 30D, you'd better have a lot of room behind you, because you are going to need to back up... a lot.

But how much?

The Trigonometry

Well that's where the trig comes in (you can skip this section if you don't want to see how I figured it out). In order for me to say how much, I needed to be able to reliably compute the distance necessary to view an object of a given width. But how? I started by drawing a diagram like this one:

V is my viewing angle. Okay it's not 29° (or 28.98333° which is the actual FOV of the nifty fifty on a crop sensor), but close enough. The legs of the triangle extending out from V represent the edges of my FOV as the distance to the subject (marked by the dashed line, d) grows. The base of the triangle (marked as w) is the width of the field of view at the distance d. Basically this is a representation of the wedge or cone of that falls within a particular FOV, in this case 30°.

I can pick any distance I want for d, but what I really need is a way to say what d should be to accomodate a subject of a certain width. In other words, to fit a subject 10 feet wide in my viewfinder, how far back do I need to stand with the nifty fifty on my EOS 30D camera? I supposed that given a formula for that, I could solve the formula for the width so that one could also compute the maximum width viewable given a distance.

The triangle depicted above is an isoceles triangle, as both the legs are the same length, and consequently the angles where the legs meet the base is also to the same. I spent some time looking online for computations for isoceles triangles, but what I was looking for didn't appear (namely, given the length of the base, and the angle of the peak, what is the height or altitude of an isoceles triangle?)

I studied trig over 20 years ago so I remember very little of it, but I did remember there were a lot of simple equivalences for right triangles (that is, triangles where one of the angles is 90°). And I realized while looking at my diagram that the line I had drawn to represent the distance, bisected V and split the triangle into 2 right triangles, each of which looked like this:

Bisecting V gives me a 15° angle (V'), and a base width exactly half of what it was before (w'). So if I could take a given distance d and come up with a formula for w', then I should be able to solve that formula for either d or w', keeping in mind that V' is V/2 and w' is w/2.

Doing a quick check online I found the two rudimentary trignometric equivalences for right triangles: for either of the angles other than the 90° one, the sin of that angle is equal to the length of the opposite side divided by the length of the hypoteneuse, and the cos of that angle is equal to the length of the adjacent side divided by the length of the hypoteneuse. Here are those equivalences for the right triangle above:

Sin V' and cos V' I can get with a pocket calculator, and I'm going to pick a value for either d or w' and solve for the other. I can solve the equation on the left for w' [w' = (sin V') * h] and I can solve the equation on the right for d [d = (cos V') * h], but both of these solutions require me to know what the hypoteneuse of this triangle is.

But in order to get w' from d or d from w' I need to do more work, mostly because I am not going to know what the hypoteneuse is. I'm only going to be starting with either V' and w' or V' and d. So what I need to do is solve one of the equations for h, and then plug that into the other equation. That should give me a formula I that I can use to solve for either d in terms of w' and V' or w' in terms of d and V'. So I picked the equation on the right. Solving that for h gives h = d / (cos V').

So I should be able to substitute d / (cos V') in the equation on the left, like so:

Now I'm good. I know what V' is, I can get sin V' or cos V' from my calculator, and I am going to pick either d or w'. So now I can solve for either one, like so:

Done, right? Well, yes, if I want to know what the appropriate distance is for half the width of my subject using a lens with half the field of view. Now I want to substitute in the equivalences that w' = w/2 and V' = V/2. In the equation on the right that will put w/2 on the left of the equal sign, so I will multiply both sides by 2 to solve the equation for w. That gives me:

Okay they probably aren't the cleanest formulas in the world, but they work and let you get the height of an isoceles triangle from its base width and peak angle, or vice versa. Using these formulas I could handily compute the needed distance for a given width in a given field of view, and this allowed me to present something more concrete than "gee whiz, that EF 50mm 1.8 is awfully confining on a crop sensor camera."

Back to the Argument

So how confining is that nifty fifty?

5 feet wide = 9.7 feet away
10 feet wide = 19.3 feet away
15 feet wide = 29 feet away
20 feet wide = 38.7 feet away
25 feet wide = 48.4 feet away

Pretty confining! If you are trying to capture 3 people sitting on a couch which is 8 feet long all in one shot, you need to stand 15 feet 6 inches away. Better have a big living room, or one where there isn't a TV 10 feet from the couch. Or maybe if you moved the couch outside... that would be cool for an album cover, but for Aunt Bea, Uncle Joe, and Granny, it is probably less so.

Working with the EF 50mm f/1.8 is a good exercise though for learning how to push a lens to do what you need, and it's plain old good exercise, because you're going to be backing up a lot. You can get that 8 foot couch in shot if you shoot from an angle, but then you will need to stop your aperture down to widen up the depth of field so that everyone will be in focus... which means you can't shoot low light anymore so you might need lamps or a flash. Or you could give up on that shot and shoot the people individually.

Or, you could simply not get the EF 50mm 1.8 in the first place, if you are not planning to shoot primarily portraits. If you want to consider the traditional FOV that the great 35mm film artists shot with, you need a lens that gives a FOV on a crop-sensor camera similar to a 50mm lens on a 35mm film camera (i.e. 46°). The closest bet would be a 28mm lens, like the EF 28mm f/1.8 or EF 28mm f/2.8. These have a crop-sensor FOV of 47.25°. With one of these lenses the width to distance figures look like this:

5 feet wide = 5.7 feet away
10 feet wide = 11.4 feet away
15 feet wide = 17.1 feet away
20 feet wide = 22.8 feet away
25 feet wide = 28.6 feet away

Much more reasonable. And quite interesting how the distance to subject is almost the same as the width of the subject. No surpise that the 50mm lens became the standard on the cameras of old.

Here We Go Again...

Abacquer — Tue, 02 Oct 2007 14:51:46 -0400

Definition of "undercutting":

Me:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=120163294055
Starting time: Sep-27-07 19:45:00 PDT (scheduled, so as to begin after other auctions ended)
Starting bid: US $49.99
Duration: 7-day listing

The other guy:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=170153862318
Starting time: Sep-27-07 20:27:48 PDT
Starting bid: US $0.89
Duration: 7-day listing

From Bulk Comes Bilk and EBay Wants Your Money

Abacquer — Wed, 19 Sep 2007 19:21:12 -0400

I've been so busy with work and trying to sell cards that I haven't had time to do much photography. I did get some great pix in Hyannis a couple weeks ago and some more good stuff at a reception for a christening I went to last weekend, but I haven't had time to photoshop the photos, clean them up, organize them, and so forth. It's just been either work or eBay the last couple weeks. Here's one picture I took (at the office, of couse). Somebody in my building went out and got one of those "Dream Cruisers". Isn't it gorgeous? Wow.

On the eBay front I continue to sell and sell. Part of my unlimited set is gone now, and I've gotten the right price for it, and I continue to identify rarities or obscurities in my collection and put them up for sale. I am happy to report that I am at 90% of stage one and expect to make it there by the end of the week. Cool beans!

But I have learned one important lesson about selling. Selling in bulk may allow you to ship more product, but it definitely costs you money. I sold a set of 8 revised dual lands back on the tenth of September for $122.50. In preparation for the end of that auction I got another 7 dual lands together and was prepared to sell them as a batch. But I decided not to. As an experiment I decided to sell the 7 duals as individual auctions (here they are 1, 2, 3, 4, 5, 6, 7). They went for a total of $139.60.

Sold as a batch, they went for $15.31 apiece. Sold individually? $19.94. That's 30% difference!!! Viewed from that perspective it makes sense to sell individually unless you have a very small market (say you are selling on a table on your front lawn... maybe a couple dozen people swing by and look at the items, chances of selling them all individually are remote.) But on eBay the market is huge, millions of potential buyers, so go for it right?

Well there *are* the eBay fees to consider. eBay will charge you for absolutely anything you can think of and plenty of things that wouldn't even occur to you. Back in the early days of eBay there was a listing fee and a sale fee. The listing fee used to be 30 cents. That was it. Then if the item sold there was a sale fee which was this sliding scale thing that was very complicated but generally worked out to about 4% to 5% of your sale. Nowadays everything costs money. No longer is the listing fee flat, but it is based on the minimum bid of your auction. If your minimum bid is a dollar, the listing fee is 20 cents. If your minimum bid is $49.99 your listing fee is $1.20, and if your minimum bid is $59.99 the fee is $2.40. As you can see the fee is determined inside some sort of bracketed structure. Want to add a reserve price? You'll be charged 1% of your reserve price. Want to add a buy-it-now price? There's a fee for that. Want to schedule your auction to start later in the day or later in the week? There's a fee for that. And so on and so on and so on. Sale fees aside, eBay takes a hefty chunk out of you up front just for listing the item. To the point where it makes no sense at all to sell anything for under say, 5 dollars. Ebay will simply eat so much of it that it becomes pointless to sell it.

And of course once your item sells, eBay charges you for a percentage of the sale price, this you would pay no matter how many items you are selling, but the listing fees are paid for each item. My listing fees for the first dual land auction was 30 cents (I had no minimum bid, no buy it now). The 7 individual auctions cost 40 cents apiece to list (probably because I set a minimum bid of $9.99, otherwise it would have been 30 cents). But that is $2.80 to list 7 individual items instead of 30 cents to list one.

Then come the PayPal fees, which generally work out to about 3.5% of the sale price but which have overhead costs causing them to be at least 30 cents as a minimum. I read an amusing story about a guy who overcharged his customer about 90 cents for shipping, and offered the customer a refund via PayPal. So the customer got sixty cents. Ouch. For items worth about $20 each, the overhead is not really all that relevant, but it would be for inexpensive items. The PayPal fees would be easier to stomach if PayPal wasn't owned by eBay. Yes, that's right, you pay eBay to list your item, you pay eBay to sell your item, and you pay eBay to collect the payment for the sale of your item. Yeesh. And, to make it even more annoying, PayPal fees are instantaneously applied to any cash transferred, but eBay fees are simply billed to you at the end of the month. That way, PayPal can charge you a percentage off of the $100 you were paid, instead of the $100 you were paid minus the eBay listing and sale fees. I hasten to point out again that PayPal *is* eBay. Double dipping anyone?

After correcting for eBay listing fees, eBay sale fees, and eBay PayPal fees, the group lot went for $14.20 per card, and the individual lot for $17.80. That's still a 25% improvement in price, even after being ~~robbed~~ charged for eBay's fine services. Altogether the 8 auctions sold for $262.10, which isn't bad, even if eBay took $23.92 in total fees. My state government charges 5% in sales tax. eBay? 9.13%. Yikes!!!

So from now on I'm going to sell individually and try not to think about the fact that I might as well just set every tenth item on fire since I'm not going to end up with any money for it.

eBayers Drive Me Crazy!!

Abacquer — Wed, 05 Sep 2007 10:32:46 -0400

Looking down the notes/status on my selling/sold auctions, they read like something out of "How to be Obnoxious -- A Practitioner's Guide". Let's run through them, shall we?

Item 1 is a very rare set which did not sell because some rather clueless sellers all decided to run auctions for the same item at the same time as mine, and undercut me. As a result they all hurt each other and nobody made what the set was worth. One guy had an extremely low minimum bid and got no bids. After his auction ended, there was only one auction left to bid on, it shot way up over the first guy's minimum despite having poorer quality cards. Thinking about that makes my brain shrivel. Now I have to wait around for all of these shmoes to finish relisting and selling their sets so I can get a fair price for mine. And they appear to be waiting for me. Greaaaaaaat.

Item 2 is my most valuable set, wouldn't sell for 80% of the retail price. I received the most ridiculous offers, promises from people who then disappeared, and got strung along by people who eventually just backed out as if surprised when I repeated stuff to them which was clearly stated in the auction description. Meanwhile the auction has had hundreds of views and dozens of people watching it. I've now relisted (cha-ching! extra ebay fees!) lowered the price to 70% and it still isn't selling. Jesus. I'm going to have to split it up and sell it in pieces (cha-ching! more fees!)... and I'll end up making way more than the current asking price. Had one buyer offer the full price if I would ship to Spain (auction says shipping to USA only.) He assured me up and down that it was perfectly safe and he does it all the time. Then I told him that he would have to assume the shipping risk (i.e. item goes missing, he has to wait for the UPS refund, up to 6 months). And suddenly it's no longer something he feels comfortable doing. (Guess he doesn't have as much faith in the Spanish courier services after all.) Currently I have one offer from a guy who "is trying to get the money together". We'll see... past history is not a good indicator.

Item 3, a set so rare that only a handful exist in the world. I set a reserve and have people who expect me to sell it to them for one tenth of the reserve. One bidder offered to buy it for the reserve price, but insists I cancel the auction and relist it with a buy it now option first!! (For those wondering, it cost about $9 to list it the first time, and would cost another $9 to list it again.) Guy refuses to just bid the goddamned reserve amount. I refused to end the auction for him.

Item 4, fixed price/best offer auction. I accepted an offer of $150, only to discover that the buyer was from Japan (you don't get buyer's location until you accept their offer.) Auction description says shipping to USA only. Genius. I recalculating the shipping charges, and they tripled. Sent buyer an invoice and pointed out that he should have contacted me first before he bid because that's exactly what the auction description says. I hope he doesn't back out, but I suspect he will (and I lose $5 in listing fees... cha-ching!)

Item 5, sold Aug-31. No contact from buyer. Invoice sent 9/1. Reminder sent 9/4. Still no contact. Standard eBay grace period is 3 days from end of auction. Going to have to send buyer a warning tomorrow and if he doesn't pay within a day after that it will be negative feedback and relist the item (cha-ching! more listing fees...)

Item 6, shipping to USA only. Bidder from Germany pleaded with me to ship to him, and eventually I agreed and let him bid. Then after winning the item he argued with me about the shipping costs and insisted I ship USPS instead of UPS.

Item 7, buyer asked repeatedly about card condition despite pictures of the cards being in auction description, and then took his sweet time getting his payment together after he won (took 5 days to make an instant electronic payment.) At least he apologized for taking so long. He'd be the first.

What next???

I See How It Is...

Abacquer — Tue, 28 Aug 2007 14:35:11 -0400

As you know I am selling some of my MTG collection (the oldest and the bestest stuff) to finance photographic gear. I spent awhile researching prices so my stuff could be competitively priced--which was a little challenging because for some of the hotter stuff (the Unlimited and Arabian Nights sets) there hasn't been an auction on ebay for months on these items. I finally figured out why. Potential sellers are waiting for a price to beat. After months with very little in the way of complete sets of Unlimited/Arabians, within a day of me listing my sets for auction, a bunch of other sellers crawled out of the woodwork and listed their sets for auction, all undercutting me by small amounts (typically about $50), or by setting high reserves and no minimum bid. Since all these auctions are now running simultaneously, there is suddenly a lot of options for the prospective buyer, and since mine is ending before the others, the lower-priced auctions are going to see all the action.

So at this point I don't expect the Arabians or Unlimited sets to sell, and now I'm going to have to sit around like the other campers and wait for some other seller to come along so I can undercut him or her.

Pain in the ass. The amusing thing is, sets this rare sell so infrequently that if a seller waited until my auction was over, he'd end up doing better than he would if they all try to sell at the same time and undercut each other. Oh well... I guess the buyers win!

Wedding Gear...

Abacquer — Fri, 10 Aug 2007 09:00:00 -0400

So while sitting around yesterday trying to work from home, aching and dealing with my swollen face, I considered the sort of kit I would need to take quality wedding/portraiture photos.

It occurred to me that if I had 10 grand kicking around, I could put together an amazing kit for doing wedding photography and portrait work. In a moment of idle dreaming I listed it out on Amazon using their "Listmania" service. You can see it in great detail here. I don't see how I could come up with that kind of dough unless I sold off my collection of Magic: the Gathering trading cards. It probably wouldn't get me all the way there, but it might get me far enough along to build a useful subset of the gear.

A serious wedding photographer brings a spare camera body on a job--both because the primary camera could break down, and also because it can make switching between lenses much quicker, so I would want a more serious body to be my primary camera and my XTi would be my backup body. Then I would need a couple high quality zoom lenses and a flash for the new camera. That would be the basic wedding kit.

Basic Wedding Kit
EOS 5D 12.8 mpx Camera	$2,643
EF 24-70mm f/2.8L lens	$1,139
EF 70-200mm f/2.8L IS lens	$1,699
580 EXII Speedlite	$430
TOTAL:	$5,911

In order to do serious portraiture these lenses plus those I already own would probably suffice, but there is a little more equipment I would need for portraiture, and there is at least one additional lens that would be nice to have for the wedding work, and some flash enhancements that would be good for both:

Stage Two
EF 16-35mm f/2.8L lens	$1,449
Botero #035 Black Muslin	$129
Impact Support System	$99
Pony Spring Clamps	$8
OC-E3 Flash Cable	$75
Gary Fong Clear LightSphere II	$49
TOTAL:	$1,809

The 16-35 lens gives me serious wide angle capability for large family shots (not unusual for weddings) and I can use it for landscape work when I am shooting for pleasure. The muslin backdrop, support system, and clamps would be necessary for decent portrait shots, and ultimately I would probably want a few varieties of muslin backdrop, but for starting out, basic black would be fine. The OC-E3 would allow me to separate the flash from the camera and hold it overhead or mount it nearby, and the LightSphere would let me diffuse the flash for softer lighting. Stage two brings the total pricetag to $7,720. This would be a solid wedding/portrait kit.

The only thing lacking is an even longer lens for very special sorts of situations, and a quality backup lens for the XTi body if I am ever shooting with an assistant and we both need to be in the most common shooting range (20's-70's) at the same time. That's where stage three comes in:

Stage Three
EF 100-400mm f/4-5.6L IS lens	$1,410
Tamron 28-75mm f/2.8 lens	$379
TOTAL:	$1,789

The 400mm L glass with image stabilization should provide the last conceivable bit of reach necessary for weddings and I suspect for most weddings it wouldn't be necessary. However it would be a solid lens for wildlife when I am not shooting weddings. The 28-75mm Tamron gets a lot of respect despite the brand and the price, and as a backup lens on the XTi body, that would be fine. This brings the total pricetag to $9,509 and gives me enough glass to do almost anything I want. Yes, there's no extreme length lenses (> 400 mm) but I can't conceive of a use for such lenses that could justify the expense.

The glass listed would cover me for macro, wildlife, weddings, landscape, low light, sports, and portraits (especially when you include the 28mm, 50mm, and 90mm primes I already own). Looking toward specialty landscape and architectural photography, the only other thing I can think of that would be "nice to have" would be a fisheye lens and a perspective-correcting lens. That would be the "bonus stage":

Bonus Stage
EF 15mm f/2.8 Fisheye lens	$580
TS-E 24mm f/3.5L Tilt-shift lens	$1,099
TOTAL:	$1,679

The fisheye lens lets you squeeze a lot of lanscape into a single shot by giving the shot a spherized look... here's an example taken with the EF 15mm fisheye. A tilt-shift lens is a strictly manual focus lens that allows you shift the focal plane in order to correct the perspective and distortion that often results from shooting tall structures from relatively nearby with short focal length lenses. Here's a sample shot with the TS-E 24mm. Anyway this would put the pricetag at $11,188, but what a killer kit this would be!! Ah well... if I win the lottery someday... fun to think about I suppose.

If I sell off my MTG collection and save up my pennies for awhile, the basic wedding kit becomes a distinct possibility. I'll be devoting some serious thought to that over the coming days.

Say Goodbye to Cellulite

Abacquer — Thu, 24 May 2007 16:00:06 -0400

Personally, I am sick of online cosmetics ads that show two side by side "before and after" photos of the miraculous effect of the product on wrinkles or age spots or whatever. Usually both pictures are photoshopped, or at least the "after" photo is. Today I saw one that at least had the decency to announce in (extremely small type) that these were "simulated images, not actual photos".

It strikes me that these annoying adverts do a better job of advertising Photoshop than they do of advertising the beauty cream or whatever. So to that end, I made this parody of a webvertisement I saw for "BodyShape" while I was reading a news article today:

So folks, go out and get some Photoshop by Adobe, and you can have a creamy smooth ass too ... at least in your digital photo album.

Wallet Say "Urk!"

Abacquer — Tue, 24 Apr 2007 10:17:28 -0400

One of the nice things about new cars is that it takes a long time before you need to shell out big bucks to fix them. For the first couple years or so, there are generally no major problems. Well Vanessa is about 4 years old now, and that time is gone. She's been running hot lately, the engine fan is always on, she has a check-engine light coming on, and she is overdue for a manual transmission fluid change. The dealership where I purchased her is no longer in business, so I took her to a different dealership yesterday to get these issues looked at.

The check-engine light is on because there's a certain emissions-related component which has failed. Replacing that component will cost $250.

The car has been running hot because the water pump is going, the water pump is covered by the extended powertrain warranty with a $100 deductible, so that will cost another $100.

The manual transmission fluid change will run another $150. (That is a lot, but apparently the job is difficult to do on this model. I had previously taken it to Jiffy Lube and they tried but could not do the work.)

The techinician who took the car for a ride said that the engine fan is much louder than normal, which he takes as an indication that the fan is going and will ultimately bind up. When that happens the car will overheat. The problem is that the component is extremely expensive ($500+) and installation isn't cheap either. To replace the fan will run me another $820.

It took 66,000 miles to get to this point, but yeesh, that's a whole lot of damage, and financially speaking, life sucks for me and mine right now. So, I told them to take care of the first 3 issues but leave the fan for another day. With the waterpump functioning properly, the fan won't need to be on as much, and hopefully I can get by without having to replace it for awhile.

I hope they finish fixing the car soon. I miss it already.

A Vista Rant and a Ray of Hope

Abacquer — Fri, 20 Apr 2007 17:13:51 -0400

It bugs me that in the last few weeks of his life my father spent an inordinate amount of time trying to get his new "Vista" computer from Gateway to run his existing software, and to play nicely with the two other computers on his network (one XP and the other Windows-98). My mom described him as regularly infuriated by the new "improvements" to Vista, such as drastically and mystifyingly reworking the interfaces of the major office applications so that you no longer know where any of the functionality is anymore, and doing away with meaningful text menus so that everything is an icon. (Which of course requires you to hover over every freaking picture and wait for a tool tip to come up and let you know if you found what you were looking for.)

This is not something a 67 year old man with high-blood-pressure and heart problems should have to deal with from a computer which is supposed to make his life easier. It is still utterly mind-boggling to me that some genius over at Microsoft thought people were going to go so loopy with joy over the new interface that nobody would ever want the option of say, switching to a "classic mode" so that they could continue to be productive. Go eff yourself, Microsoft.

Now that Dad has passed away and my Mom's old Windows-98 machine is becoming (more) unreliable, we decided to help her move her business on to Dad's new computer. The very first thing we did was back off Dad's files, and then pack his computer into a box with all of its peripherals and Dad's old disks, and take it to a computer store near my house. Dad had Win-98 install disks, an XP home edition upgrade disk, and Office-2000 from the older computers which were going to be disconnected and shut down once the new box was functional. I told the guy at the store: "Wipe it, and install Win98, then upgrade to XP, install office 2000, download and install all the service packs for XP and Office, get AOL installed and functional, then locate, download, and install all the drivers so that these peripherals (monitor, scanner, printer) will work with the XP OS."

It took the store 2 days to get the computer back to me. Total cost, $90. When we brought the computer back to Mom's house it took all of a couple hours to have it up and running with all of her software and fully networked with the other machines so that she could transfer her records to it. Poor Dad had spent weeks wrestling with the stupid thing. He was upset when it arrived because he didn't realize that it would be coming with Vista and that practically everything he did on a daily basis had been turned upside down and obfuscated by the new interface. The new computer didn't even end up in his office--he had to unpack and put his old computer back together again because the new one wouldn't work the way he wanted it to--so he ended up putting the new one on a tiny desk adjacent to his bedroom.

Stories like my Dad's are sadly not uncommon (minus the tragic ending, of course). But home users everywhere, excited about buying new hardware and updating to the latest and greatest new OS, are discovering that latest is not necessarily greatest. In fact it might just suck royally. The end result, tens of thousands of angry customers phoning up computer retailers and saying "WTF? Why can't I buy a box with XP on it? I don't want this screw-head OS on my new box, I want something I can actually use!"

Well, computer retailers are finally getting the message. Dell has just announced that customers buying home systems can opt to have the system shipped with XP and not Vista. From a Computer World article:

Dell Inc. yesterday bowed to pressure from customers and dumped its Vista-only policy for consumer PCs by adding Windows XP as an operating option on half a dozen machines...

..."We heard you loud and clear on bringing the Windows XP option back to our Dell consumer PC offerings," Dell said on its Ideas in Action page, a companion to its IdeaStorm site...

...Earlier this month, Dell decided to add Windows XP on systems sold to its small business customers. At the time, Tom West, Dell's director of small business marketing, said the older operating systems wouldn't be offered to consumers. "Dell does not have plans to launch Windows XP for home users as the preference and demand is for the 'latest and greatest' technology, which includes Windows Vista," he said on the Direct2Dell blog.

Yesterday, Lionel Menchaca, Dell's manager of digital media, noted the reversal in another blog entry. "After Tom West's post that said Dell would offer Windows XP on systems for small business customers, many Direct2Dell readers have commented or sent e-mails asking us to bring back Windows XP for home users as well. Today, we're doing that."...

...Rival Hewlett-Packard Co. was not available for comment on whether it would answer Dell with a Windows XP option for its consumer customers. HP's small business systems can currently be ordered with either Windows XP Home or XP Professional, as well as Vista Business or Home Basic; the company's consumer PCs, however, are available only with Vista...

How long do you think HP is going to go on having home users refuse to buy computers from them unless they get XP knowing that Dell is making that option available? Interestingly the XP-equipped computers from Dell are cheaper. If Vista was so easy to set up and use, wouldn't the XP machines be more expensive? According to Microsoft they should be, but then Microsoft isn't in the same reality as its customers.

Rule #1 when upgrading a user interface: do not make using the application HARDER for your installed base. Rule #2: if you are making a fundamental change to the interface which will require a learning curve from users either make the new interface optional, or introduce it in slow steps.

There's already a company out there offering some sort of patch that attempts to put menus back on your office applications if you were unfortunate enough to step in Vista without realizing what you were getting into. And of course, they want you to pay for this. Yeahhhh, I'm tired of being part of Microsoft's world-wide division of beta testers known as "customers", and I don't think I should have to shell out more money to get back to productivity, so I think I'll stick with XP for now. When Microsoft finally pulls its head out of its butt and releases a service pack so that users can shut off the silly ribbon and get their hierarchical menus back, then I'll be looking at Vista--till then, no way, and Microsoft can lips my donkey.

Recently, MS slyly slipped Internet Explorer 7 into the "critical updates list" for XP. Which means if you have your XP set up to update itself automatically, you may soon sit down to your machine and find your browser has been quietly replaced by a different version which has a different interface... most notably, the browser tool bar is now at the top, but some of the icons have been moved down to a second line on either side of the tabs, and the menu bar is conspicuously missing (screen shot). Why? Because IE7 is basically the "Vista" version of IE... so if you were thinking of upgrading to IE7 that's something to keep in mind.

In IE7, you can at least turn the menu bar back on, but you can't put it at the top--no it must sit below the toolbar now because MS thinks that is more important. So instead it gets sandwiched between the two rows of icons, which is a bizarre place to put your menus.

Personally I barely use the toolbar at all in IE, the only buttons I use there are the forward and back buttons. I use the menu bar all the time on the other hand, mostly to access favorites, save copies of pages, print, and view the source of pages. Here's a thought, get rid of all these unnecessary icons, and put the next and back buttons ON THE MENU BAR. There's even enough room to fit the URL box there too!

It's becoming a recurring theme that MS comes up with some dippy interface idea and is flabbergasted when the community at large says "WTF were you thinking? How do I shut this shit off? Why is it turned on by default?"

Anyone remember "Clippy"? That dip shit paper-clip cartoon that suddenly started getting in the way of accessing application help in the manner that you were used to? People HATED Clippy with a passion. I don't know anyone who hasn't turned the stupid "office assistant" off, ANYONE. Except perhaps for people who couldn't figure out how to turn the annoying thing off. I already know how to work with application help files, I don't need to be forced to start entering "English questions" into the numbtastic assistant so that it can screw up and take me to the wrong page.

"The Ribbon" is the new office assistant, once again Microsoft runs off the god-damned end of the Earth with the fool notion that we want their help. Please, Microsoft, please stop trying to help me and just give me an application that I only have to learn ONCE.

Wolfowitz? Now What Has He Done?

Abacquer — Fri, 13 Apr 2007 09:17:23 -0400

Remember that guy? Patron of the "topple Saddam, democratize Iraq, (magic happens), democracy spreads across the Middle East" creed? Handed the presidency of the world bank when the ill-fated Iraq policy imploded? Yeah, that guy. Not sure why this would come as any surprise, but the World Bank isn't too happy with him right now either.

Apparently it has something to do with him making arrangements to get ridiculously huge salary for his girlfriend at the World Bank. Hard to blame him. His time of ascendancy came with the Bush administration, and so cronyism probably sounds like the right way to do things to Mr. Wolfowitz. He himself has basically said that he didn't realize it was wrong at the time.

Okay, I'm kidding, it's very easy to blame him. What a sleaze.

Having a girlfriend who receives her salary from the organization you are President of is probably fine as long as she is qualified, and she has worked for the World Bank for eight years, so there's no doubt she is qualified for the work. But her salary is wayyy out of scale with the salary awarded to other employees working in similar positions at the World Bank. From a BBC article:

...When Mr Wolfowitz took over at the bank in mid-2005, Ms Riza - then a World Bank employee for eight years - was transferred to work for the US state department, to avoid any conflict of interest.

But rapid rises in her tax-free World Bank salary to about $193,000 - more than the $186,000 Secretary of State Condoleezza Rice receives before tax - have aroused ire among other bank employees...

So we knew Wolfowitz was delusional because of the cockeyed Iraq policy, and now we know he's crooked as well. One thing is certain, this guy needs to stop working in positions where he gets to give orders or advice. He's clearly unfit to manage. A manager has to have a firm grasp on the pragmatic--pipe dreams and money schemes do not a great manager make.

And the World Bank is considering just that as the board is meeting to decide what to do with him. I don't think they'll ask him to step down, but certainly some sort of reprimand must be coming. Stay tuned...

Don'tcha wanna haggle?

Abacquer — Thu, 12 Apr 2007 11:42:10 -0400

So I spent the last hour plus on the phone with my new dentist. I was trying to figure out exactly what I needed to pay today. I know the cost of the temporary upper denture was $1575, and I might possibly want a temporary lower denture, which the dentist offered to discount to $800. (With no lower denture, I have nothing to chew on.) The 7 little incisors in the front won't cut it, or rather they will cut it, but they won't chew or grind it. The oral surgeon offered us pamphlets on companies that pay medical bills and offer you a payment plan, so we needed to know if the dentist would work with these companies.

Phone call number one to the dentist made clear that they would not. They don't work with any such companies and would be unwilling to do so. Adding a middle man complicates the payment process. Oh well, at least the oral surgeon will work with these companies, and the oral surgery has been estimated at $4700, so we're going to need the help. Too bad the dentist won't go that route, but I understand their reasons.

So the initial bill with the dentist is either $1575 or $2375 (if I get the lower temporary), no argument there. The question is, how much am I supposed to bring to the office with me? The lady at the office said, at first, you will need to bring your co-pay. Okay, so I'm thinking like, $25 or something, and then they bill me for whatever the insurance doesn't cover. Well no, the lady explained, the co-pay is whatever the insurance won't cover. She said insurance would probably pay half the bill. So I said, "so then I need to be able to pay the other half of the bill... today?" (somewhere between like $800 and $1400.) And she said, "yes." But then she said, "well when the co-pay is that high, we usually only charge half the co-pay at the time of the visit, and then bill you for the other half." So I would need either $400 or $700 today? She said, "yes." But then she said, "let me look at your chart and check with your dentist and then call you back just to make sure."

Several minutes pass.

Finally she calls back. It turns out that for temporary dentures, they don't submit bills to insurance companies at all. They only bill the insurance for the permanent denture. So today's bill falls entirely on my shoulders, and is not a co-pay... it's just a "pay." So I will need to cover the entire payment for the temporary dentures today, which is either $1575 or $2375, and I am thinking $2375 because I can't imagine going 2-4 months with no lower back teeth. As it is I have only one usable lower back tooth (broken, but usable), but I use it to chew everything I eat.

Needless to say that's a large chunk of change to come up with, and I would have appreciated it if someone could have explained to me in advance that I was going to have to be able to pay the total amount today. I was expecting insurance to kick in $1000 today. I'll just have to make it work, and I can swing it but it isn't going to be a cakewalk.

Teeth are expensive. The temporary dentures will be $2375, the surgery will be $4700 (assuming no unexpected work), and the final dentures will be about $3150. That's 10 grand folks. Insurance? They'll pay $1000. Period. So I get to come up with $9 grand. Can I get a "yikes" anyone?

The IRS Wants a Piece of Your eBay Earnings

Abacquer — Mon, 26 Feb 2007 08:36:37 -0500

It's been years since I sold anything on eBay, and when I was selling, I made a pittance. But some folks make their living via eBay, and the IRS knows it. Unlike full time employment, there are no W-2's from eBay, and it is not up to eBay to report your earnings. Yet...

According to an article on ars Technica, the IRS is attempting to force eBay to submit documentation to the IRS on the yearly earnings of it's sellers. eBay for its part is resisting at present, by rightly pointing out that there are many online auction/sales houses, and if eBay is singled out, the customers will simply go elsewhere.

So a word to eBay sellers out there... if you aren't reporting your earnings, you might want to start.

Weelanders Show How Evolution Leads to Extinction

Abacquer — Sun, 18 Feb 2007 22:12:45 -0500

Okay this is going to be a lengthy article with a lot of graphics, so I am going to use an excerpt today. Basically it's a summary of the first run of my Genetic Factoring sim, and how it demonstrates that evolution can lead to extinction. It also includes a summary of the most efficient (and robust) genome to be generated by random mutation and natural selection, and how it compares to the original genome. Let's begin with a graph...

Genetic Factoring -- Weelanders Run Amok

Abacquer — Sun, 18 Feb 2007 10:00:00 -0500

I'm sure you remember the Weelanders... artificial life forms which lived on a grid and helped demonstrate that evolution through random mutation causes natural selection to kick in something fierce. At the time I originally ran tests with the Weelanders, I hypothesized that the Weelander concept could be adapted to do more than just demonstrate natural selection.

And, a few months ago, I did just that. I built a new version of the Genome application called "GeneticFactoring", and stripped out a ton of Weelander support code. I wanted to make Weelanders that factored numbers and then reproduced based on how successful they were at factoring. As a result they didn't need to search for and consume food, they didn't need sexual reproduction, they didn't need the ability to move, heck they didn't need chromosomes in pairs, as a result, all of the "critter emulation" code was unnecessary, resulting in a simpler Weelander support system.

What they did need was the ability to compute square roots, find primes, and factor numbers... this was the new code that needed to be added, and it was all going to be written in the native Weelander instruction set. It took me awhile to code the genome "Facto-f", but eventually I had it working. Then I modified the container and world code to invoke the Weelander factoring code in the following manner:

Populate: at the beginning of time, populate the grid so that it is completely full of the starting Weelander genome "f".
Setup: build a list of 50 numbers to factor, then factor each number using internal container code and store the results.
Execute:
1. for each critter on the grid
2. run through the 50 numbers
3. call the Weelander's Execute method and pass it the number to factor.
4. The Weelander factors the number and returns the results.
5. If the Weelander goes into an infinite loop without returning results, terminate it.
6. Otherwise, compare the Weelander's results to the original results.
7. If they do not match, terminate the Weelander for inaccuracy.
8. If they do match, note how many execution steps the Weelander needed to compute the result, and proceed to the next number.
9. Once all the numbers have been factored, compute an efficiency value for this critter (total number of steps divided by total number of tests).
10. Proceed to the next critter and repeat.
Cull: examine all the living critters on the grid and gather the top ten percent based on efficiency, and terminate all the others.
Repopulate: for the most efficient critters, allow each one to reproduce asexually (possibly with mutation). If after processing them all, the grid is not filled, go back and run through them again. Keep doing this until the grid is filled.
Loop: Jump to step 3.

An immediate problem which cropped up was that Weelanders would mutate in ways that made them faster at factoring this specific set of test numbers. So if by random chance, none of the test numbers was a multiple of say, 7, the Weelanders might make themselves faster by throwing away the "divide by 7" test somehow. Which is great until you hand them a multiple of 7 at which point they die from inaccuracy.

I solved this problem somewhat by changing step 6 to jump to step 2 instead of 3. Thus on each "tick" of the world, a new set of test numbers would be created and then all the Weelanders would be tested against that set. Makes sense really, and automatically weeds out Weelanders that made the grade last tick by "cheating". I improved this further by making the first 10 test numbers a fixed set that didn't change and which exercised a lot of the conditions I wanted to make sure were covered. The remaining 40 were random.

But another problem appeared, one that completely flummoxed me because the system appeared to drift toward less efficient creatures over time, instead of more efficient ones. How could that be possible? Natural selection should continue to hold true, and the creatures should be *more* efficient over time, not less efficient.

But nothing prepared me for the worst problem at all... hardware. My computer, wonderful though it is, is experiencing a hardware problem. The cooling system is no longer operating, or not operating well. As a result, any operation that pins the CPU usage at 100% for a period longer than 10 minutes or so will cause an audible heat warning alarm to begin issuing. It's an alternating two-tone klaxon which doesn't come from the computer speakers, it comes from somewhere inside the computer case. And it is, needless to say, very alarming. I like my CPU. I don't want to cook it.

Needless to say factoring numbers is CPU/FPU intensive. And I've got 230 Weelanders factoring 50 numbers each, over and over, ad infinitum. After 10 minutes or so of runtime, the CPU gets too hot, and I have to shut the simulation down or risk damaging my computer. This problem cropped up immediately after building the new GeneticFactoring code and caused me to terminate the experiment permanently months ago.

Then last night I broke the code out again and tried to figure out a way to get it to execute without pan-searing my CPU. Eventually I came up with a rather simple solution. The simulation executes for 20 ticks, pinning the CPU at 100% for about 60-90 seconds. Then the simulation pauses, basically issuing a call to Thread.sleep for sixty seconds of downtime. This drops CPU consumption to 0% or close to 0% for a minute, giving the CPU a chance to cool. Resulting in a usage pattern like this:

I'm happy to say that having made this change I've been running the simulation for 3 hours with no overheat alarms. At this point I feel safe walking away from the computer and letting it process the simulation.

Regarding the other problem, the problem of the population drifting toward inefficiency, I also managed to solve that. I observed the changes to the Weelander genome over time. After awhile I began to observe what I thought was the culprit. Creatures weren't being selected based on their efficiency at factoring *all* numbers, just the 50 test numbers they had to factor.

So let's say a mutant does something weird like tests factors out of order... instead of 2, 3, 5, 7, 11 it goes 5, 3, 2, 7, 11. It's still accurate, but it tests a different factor first. If more of the numbers in the test set are multiples of 5 than 3, and more are multiples of 3 than 2, then this critter will in the end rate more efficient than the basic Facto-f genome, which is suitable for factoring any value. As a result, genomes which aren't really more efficient at factoring any value, get selected ahead of the f genome, and over time, this exterminates the f genome. After which the population becomes less efficient over time as more and more test sets are produced.

That was my hypothesis anyway. To solve the problem, I decided to artificially inject the original genome into the list of critters during the repopulate stage, thus if the population ever drifted away from efficiency, there would always be some number of the original genome to tug it back. Basically I grab my ten percent most efficient, and then tack 10 copies of the original genome to that list before I repopulate. This makes sense since natural selection is only concerned about the environment you are in (i.e. the 50 numbers you happened to test) as opposed to my overall goal (more efficient factoring of ANY number.) Now the top ten percent must always compete with the baseline. If they drift away, the baseline will win and it will get selected for repopulation instead. The only way to stay alive is to consistently beat the baseline.

Makes sense doesn't it? Yes, yes, I know, brilliant.

It was right about that time I noticed that the code that was selecting genomes for propagation was selecting the least efficient ones instead of the most efficient ones. Y'know, it takes a special kind of mind to come up with a brilliant explanation like the one above and still be completely wrong. Needless to say selecting the least efficient members of the population will uh, tend to make the population drift toward inefficiency. It's amazing what happens to the code when you say "greater than" but you meant to say "less than". > versus <, baby... classic coding error.

Nonetheless, although the effect I hypothesized was ultimately not to blame for the drift, I still think the effect could occur, so I decided to keep the "baseline infusion" code in place.

So the factoring Weelanders are back in action, and I'm interested to see what they will do to become more efficient at factoring. There are a lot of simple things I can think of that would lead to slightly more efficient code, but the Weelanders are always surprising me.

What I can tell you is that given a test sample, the Facto-f genome generally requires (on average) 1500 to 2200 execution steps per number. After running the simulation for 3 hours, the top genomes are showing efficiencies on the order of 500 to 600 execution steps per number. Ok I'm impressed. Especially when you consider that these Weelanders have to produce accurate results to survive. If they fail any single factoring test, they're gone.

I'm going to let the sim run for awhile longer and then take a peek at these efficient Weelanders. I wonder what they are doing?

Babylonian Weelanders?

Abacquer — Sat, 25 Nov 2006 15:03:39 -0500

If you futz around with math like I do, you probably have to compute square roots from time to time. We all know how to do it. Basically you punch the SQR button on your calculator. Or you type SQRT(A1) in Excel or something like that. But how do you do it by hand? And moreover, why am I discussing this? Well, for the latter question, you'll have to wait a bit. First, square roots.

In school they taught me this: For a given number x choose a smaller number r and square r. If r squared is bigger than x, choose a smaller value for r, if r squared is smaller than x choose a larger value for r. Repeat until you converge on the approximate square root of the number.

So, if x is, say 1200 (for reference sqrt(1200)=34.641...), then I would pick an arbitary value for r. Because I am a computer geek I happen to know that 32^2 = 1024, so r=32 is too low because 1024 < 1200. So then let's set r=33. 33^2 = 1089, too low. If r=34, r^2=1156... closer, but still too low. If r=35, r^2=1225, which is too high. So I know that the square root of 1200 is between 34 and 35, and I can refine my estimate further by choosing fractional values between 34 and 35:

r=34.5, r^2=1190.25
r=34.75, r^2=1207.56
r=34.625, r^2=1198.891
r=34.6875, r^2=1203.22

And so forth. As you can see we've gotten pretty close to the square root now.

This is basically a watered down variant of the Babylonian Method of computing square roots. When I was in Junior High, the above method was easier to understand. But now that I am an adult, the actual Babylonian Method description is more accessible and more concise. Here it is from the wiki:

The most common method of square root calculation by hand is known as the "Babylonian method". It involves a simple algorithm, which will bring you closer and closer to the actual square root each time it is repeated. To find r, the square root of a real number x:

Start with an arbitrary positive start value r (the closer to the square root of x, the better).

Replace r by the average between r and x / r. (It is sufficient to take an approximate value of the average, not too close to the previous value of r and x / r in order to ensure convergence.)

Repeat steps 2 and 3.

Starting from my original M of 33, it takes only 4 iterations to come within 1/100,000th of the actual square root of 1200:

r=33.00000, r^2=1089.00000. Too low.
r=34.68182, r^2=1202.82851. Too high.
r=34.64104, r^2=1200.00166. Too high.
r=34.64102, r^2=1200.00000. Done.

Interestingly this method works surprisingly fast even if you pick a dreadful starting value for r. If in this example I start with r=1, it takes only 10 iterations to come within 1/100,000th of the square root of 1200.

r=1.00000, r^2=1.00000. Too low.
r=600.50000, r^2=360600.25000. Too high.
r=301.24917, r^2=90751.06084. Too high.
r=152.61629, r^2=23291.73210. Too high.
r=80.23957, r^2=6438.38915. Too high.
r=47.59739, r^2=2265.51190. Too high.
r=36.40443, r^2=1325.28246. Too high.
r=34.68373, r^2=1202.96082. Too high.
r=34.64104, r^2=1200.00182. Too high.
r=34.64102, r^2=1200.00000. Done.

Square Roots and Factorization

Now why am I going on about this? Well I've been thinking about the Weelanders again. I'd like to create a new generation of Weelanders who break numbers down into their prime factors. (As in 1200 = (2^4)(3)(5^2), or 2 x 2 x 2 x 2 x 3 x 5 x 5 = 1200). Because of this I need Weelanders to be able to compute square roots, and the method above will be the one they will use.

Why do I need to do square roots to factor a number?

One rudimentary way to find the factors of a number N is to divide N by values which are < SQRT(N)... if any of them evenly divides the number, then it is a factor. The SQRT portion is important, because it lets you know where you should stop trying factors, without having to keep going all the way up to the value of N itself. This makes sense because any value X<SQRT(N) which is a factor of N, necessarily multiplies by another factor Y>SQRT(N). The only exception to this is when X=SQRT(N). This means that by the time you've reached the SQRT(N), going further will simply yield factors you already could have computed by dividing N by an earlier value of X. We can show this with 1200. Here are all the values of X less than SQRT(1200) which are factors of 1200, and the corresponding factor you must multiply by (Y) to get 1200:

X=1, Y=1200.
X=2, Y=600.
X=3, Y=400.
X=4, Y=300.
X=5, Y=240.
X=6, Y=200.
X=8, Y=150.
X=10, Y=120.
X=12, Y=100.
X=15, Y=80.
X=16, Y=75.
X=20, Y=60.
X=24, Y=50.
X=25, Y=48.
X=30, Y=40.

Note how X and Y both converge toward the SQRT of N. This is why when you are looking for the factors of N, there's no reason to test values of X greater than SQRT(N). The biggest value of X in this case that divides 1200 evenly is 30 which is just a little less than SQRT(1200)--34.641. The factor paired with that X is 40, which is just a little larger than SQRT(1200).

So if for example N was 1201 instead of 1200. You'd know that 1201 was prime by the time you had finished testing 1201/34. Because 1201/X where X is an integer larger than 34 must yield a value Y which is less than 34 which you would have already found.

For doing prime factorization (as opposed to full factorization, which is what you see above), you test even fewer factors (only the primes), and your target endpoint changes with each successful factorization. Consider 1200 again, as with full factorization my endpoint for testing is SQRT(1200). So I try the first prime 2. This works 1200/2 = 600. But that means now I'm no longer really factoring 1200, I'm factoring 600. Which means my endpoint is now really SQRT(600) or 24.5. That means: none of the primes between 24.5 and 34.6 evenly divide 1200 or 600. So now I don't have to test 29 and 31.

Divide by 2 again, and that gives us the result 300, which has a square root of 17.3. So there is no reason now to test beyond the prime 17, and so we no longer need to test 23 as a factor of 1200.

Divide by 2 again and we get 150, with square root 12.2. Which means that 13 and 17 also no longer need to be tested.

Divide by 2 one more time we get 75, with square root 8.66. Which means that 11 doesn't need to be tested. Look how much we figured out! By dividing 1200 by 2 successively we now know that all of its remaining prime factors are less than 8. And since we clearly can't divide by 2 again, all that is left to test is 3, 5, and 7! Isn't that cool?

If N were 1205, we'd find after dividing by 5 that the quotient 241 requires us to test no further than 15. So that would mean we'd test 5, 7, 11, and 13 with no luck, which would tell us that 241 is itself prime.

Prime Factorization and Weelanders

I'd like to recode the Weelanders to do basic square roots, and basic prime factorization, and then build a "rack" of 1000 Weelanders, give them all the same number to factor, and let them all factor it. Any Weelander who gets the wrong answer or no answer gets the axe, and of the remaining Weelanders, the fastest 10% (those who computed the results in the fewest number of steps), would be allowed to reproduce (sexually or asexually, haven't decided yet.) The remaining 90% would also get the axe to make way for the offspring of the top 10%. The top 10% would reproduce until the "rack" was filled again. And then it would be time for another factorization test.

I'd probably not want to select Weelanders based on the results of a single factorization. I'd probably want to do it based on 5 or 10 factorizations. (It's possible that a mutant Weelander might be exceptionally well suited to factor the number 5268, but be incapable of factoring other values at all.) Therefore you'd want to test them a bit and average the results. (But of course you would immediately cull any Weelander that couldn't solve the problem correctly.)

I seriously doubt that I can come up with any novel ways to factor numbers via this method, but I know there are a number of little "tricks" for factorization that I am not familiar with... maybe my Weelanders can teach me those tricks through natural selection? Wouldn't that be cool?

Photoshop on the Cheap

Abacquer — Fri, 24 Nov 2006 10:22:17 -0500

Interested in Photoshop? I've always been. But for a long time I went without Photoshop, and limped along using crappier products.

Buying Photoshop for amateur home use can be problematic since Adobe Photoshop CS2 retails at a price point over $500. Which is why I recommend that you should not buy the latest version from a retailer. If you are just a home user mucking about with photos, an older version of Photoshop purchased used on eBay will have more features than you are ever going to use and will probably save you a bundle. It's only illegal if you are buying a pirated copy. If someone has upgraded from Photoshop 7.0 to Photoshop CS2, or switched to a new product, or a different operating system, they often will sell their old disks, manuals, and so forth on eBay, which you should be able to legally register under your name. Just make sure, if you get it on eBay, that you buy it from a reputable seller and be sure to ask the seller if you can legally register it. If you're really concerned you can probably also find old versions from some retailers. Amazon carries Photoshop 6.0 starting from $200. That's pretty big savings!

The older you go in versions, the cheaper it will probably be. You could even go as old as Photoshop 5.5. I checked eBay completed auctions and found several where 5.5 went from anywhere between $50 and $150. Here's a copy of 5.5 that went for $76. Here's a copy of 6.0 that went for $111. Here's a copy of 7.0 that went for $260.

One other thing to keep in mind if you buy on eBay, is that there are many Photoshop-related products that in the end you don't need or may not be able to use. So make sure you are getting the right product from a seller with a high feedback rating who sounds knowledgeable. Things to be aware of:

Do not buy anything that says "elements" in the title. "Photoshop Elements" is a companion product to photoshop, you don't need it and without photoshop installed, I'm not sure you can do anything with it.
Do not buy anything that says "upgrade" in the title as in "Photoshop 6.0 Upgrade"... these are special upgrade packages that will only install photoshop if you already have an older version installed on your system. Adobe sells these at a much lower price to entice people who shelled out big bucks for the last version to buy the next version. Be very careful here. If someone seems to be asking a very low price to sell you Photoshop, be sure to ask them if it is the "full retail version" as opposed to the "upgrade version".
Do not buy anything that says "educational", "training", or "tutorial" in the title. This is either a special stripped down version of Photoshop price-pointed for college students, which you will not be able to legally register unless you are a college student, or a companion product that teaches you how to use Photoshop. The prices on these versions will be inordinately low. Remember if it seems too good to be true, it probably is.
Remember to verify what platform (what type of computer) the software is for. If the seller doesn't mention "Windows" or "Macintosh" then the product he is selling could be for either. Ask.
Try to get the "full boxed" product. This means you are getting everything, the manuals, the disks, the registration card, everything. Be wary of someone who is just selling a CD, or even a CD and a manual. If someone is selling an unopened boxed product still in the shrinkwrap that's your best bet... everything you need will be in there. If you don't get everything, you may not be able to register it.

In the end try to pick a seller with a high feedback. A high feedback seller will likely have sold lots of old versions of packages and will fill out her auction description in such a way that you won't need to ask any questions at all, as in "Adobe Photoshop 6.0 for Windows. This is the full retail version (not an upgrade, not an educational version), in the box with all the manuals, disks, and registration materials. This product is unregistered." If the auction isn't clear on all of those points, ask the seller to clarify. She won't mind, after all she wants you to be happy with what she is selling, so she'll probably respond relatively quickly. If you can't get enough information or the seller seems shifty or ignorant, let it go and get the next one, there will always be a next one.

Happy photoshopping!

Hot? Check! Fresh? Check! Crispy? Check! Stupid?

Abacquer — Tue, 14 Nov 2006 02:03:23 -0500

If this article were ever linked to from FARK, the given title would be something along the lines of "Having Solved All Other Problems, Chuck Discusses French Fry Containers". I can't help it... sometimes I see stuff and it just eats at my brain and I need to get it out (the stuff, not my brain). Burger King recently redesigned the cardboard container they serve fries in. One has to wonder who thought the public would give a shit, and further, how many millions went into this project. Here's the new "FRYPOD" (bet Apple computer LOVES that name) unfolded...

The IAVA Rates Your Elected Officials

Abacquer — Tue, 07 Nov 2006 14:26:12 -0500

The Iraq and Afghanistan Veterans of America Action Fund (IAVA) has released a set of letter grades for all US Senators and Representatives. The letter grades reflect the percentage of times that the senator or congressman cast a vote on a military-related issue that the IAVA agreed with. It's pretty interesting. Rick Santorum gets a D-, John Kerry gets a B...

Girls Costume Warehouse

Abacquer — Mon, 30 Oct 2006 07:39:00 -0500

Probably because of my recent rant on the proliferation of the "sexy" Halloween costume for women, my friend Dave sent me this funny video. Beware! There's foul language (and a lot of sexy costumes), so wear your headphones.

Thanks Dave. I assume the next time you call me because I haven't been around in a while, you'll say "Get da fuck down here, I got shit I'm tryna sell."

EDIT: link updated to yet another copy of the video. Sorry the last one was pulled. This one will probably get pulled eventually as well.

EDIT: Thanks James for providing the source link for the video. I've updated the above link to point there.

Costume Trends

Abacquer — Fri, 20 Oct 2006 12:15:03 -0400

I just read an article over on Skepchick where she ponders the current proliferation of the "sexy" Halloween costume for women. It was a simultaneously amusing and disturbing read. I had noticed the same thing myself the other day when I found my daughter scanning the pages of a catalog we got in the mail advertising Halloween costumes...

Do NOT Click That Link -- Another E-Mail Scam

Abacquer — Thu, 07 Sep 2006 01:20:52 -0400

Well I just got a scary letter from PayPal.com... the subject was "Account compromised: billing information moved or changed". So I cracked it open and it said...

Sleepovers and Sudoku

Abacquer — Wed, 26 Jul 2006 14:33:23 -0400

Last night was somewhat hectic as Lynnea's cousins Samantha and Breanna slept over at our trailer...

New Experiences

Abacquer — Sun, 23 Jul 2006 12:06:24 -0400

So today I’m wearing boxer briefs. Now you may have one or two questions at this point, neither of which is probably "why are you wearing boxer briefs?" More likely you're asking "why are you telling me this?" or "why would I care?" Read on to see which of these questions I'm planning to answer...