© Norman Sperling, March 7, 2012
I would like to read a comprehensive book about Y2K, especially a competent description and analysis of the aftermath. I haven't been able to find such a book. Does one exist?
Broadening to the big issue of legacy software would generalize it from a single event to an ongoing situation. Legacy code is a real issue for many companies because a lot of original code was not optimal:
* it was written as a first try,
* under great pressure,
* in an under-funded company,
* thinking months ahead, not decades.
Inelegance is the least of its problems.
A lot has been learned about superior ways to do things since then, but later editions all have to work with the original. This weighs down products from many big companies.
A software engineer who had worked at Oracle told me that Oracle did indeed find and fix what would have failed.
review © Norman Sperling, February 7, 2012
This new book tells nothing new, and offers many examples of no value. Ostensibly celebrating the pirate economy, the author neither self-publishes nor finds street sellers. Instead, he contracts with a name-brand publisher, copyrights his tales of piracy, and repeatedly invalidates his own premises.
The underground and pirate economy is not rising, it's always been around. This "informal economy" is older and far more entrenched than the formal one. In several places the book admits that, but immediately reverts to the fantasy that working "off the books", on the street, on the margins, or not fully licensed, is new, or increasing.
What's newer, and growing far more vigorously, is the formal economy that earns confidence, enforces inspections, builds brands, and does things right. Several times, the narrative brushes up against the roughly-direct relationship between an enterprise's degree of formality (for which the author selects the odd proxies of being licensed, registered, and taxpaying) and its degree of trustworthiness. Trust and confidence are critical in transactions.
That's why customers graduate to more formal levels of the economy as soon as they can. They get better quality and therefore better value: the things they buy are closer to "real" and "working" and "sturdy" and "supported", and therefore worth the higher price. This generates valuable repeat-business, compared to street-hawkers who always need to drum up yet more new customers. Of this, the book gives only the slightest mention.
The author offers several sighs over capitalist misbehavior, while citing far more examples (without sighs) of pirate misbehavior. Almost all the misbehavior is just plain short-sighted: taking an immediate advantage and ignoring its (bigger) long-term consequences. Undermining value, as several chapters on piracy celebrate, undermines confidence. Folks who can't afford the most-trustworthy goods, and therefore take less-trustworthy, discounted street-goods, often live to regret it. Frequently-cheated customers are less eager to buy, which slows the 'speed of money', whose rate tracks the health of economies.
Save your time and money: skip this book. To improve the economy, earn as much confidence as you can (in reality, not just "licenses and registrations"), and do business with others who also earn confidence.
© Norman Sperling, January 25, 2012
Norman W. Edmund founded Edmund Scientific Company on a card table in his home in 1942. When he retired in the mid-1970s, it had over 200 employees. He died at the age of 95 last week in Fort Lauderdale, Florida, to which he had retired.
I vividly remember devouring every new issue of the Edmund catalog while I was growing up in the 1950s and '60s. The catalog always had a lot of "tutorial" segments - several paragraphs each, usually with diagrams, so the users could understand the technicalities of the equipment. They weren't particularly slanted toward Edmund products, and they taught a great many people a lot about their hobby and its hardware. Only a few catalogs (like Orion) continue to do that, though it's absolutely the best policy and should be fostered. Tutorials are NOT waste-space, and they foster brand loyalty: I trust the company that makes the effort to tell me the straight information.
I met Norm several times in the 1970s, while I consulted for his son Robert. In those years Norm kept his desk in the main office, kept a bunch of neat science-thingies around, and had appropriate input. But I also sensed that he kept his distance from daily operations, carefully avoiding stepping on toes.
What always impressed me was how nice he was. Plain, no affectations, no flaunting. And he passed all that on to the rest of his family, several of whom I met. They're all nice. They treat people well. They treated me very well. It wasn't just a put-on performance, it was genuine.
To Norman and Robert, "treating people nicely" is business policy as well as personal. While it's true that being nice to people is good customer service and good business, I think they are nice to people simply because they think that is the right way to be. I learned a lot from that.
They didn't outsource service. Callers were transferred to people who knew the technicalities they needed. Customers could get replacements and refunds.
Robert once told me "Customers will always complain. They'll complain about price, or they'll complain about quality. As long as I'm president, they aren't going to complain about quality." Which is to say, the stuff he designed, produced, and marketed would actually work well. And it did. Sure, humans aren't perfect and hardware isn't perfect, but when problems cropped up, the company tried hard to fix them, and usually succeeded.
Norman Edmund was well-respected as a leader in science business, an advocate of science education, a business leader of Greater Philadelphia, an expert fisherman, and a gentleman who "lived long and prospered". I'm really glad I knew him.
© Norman Sperling, January 12, 2012
Computer and software competitors come up with more and more features, but here's one that users would appreciate, that shouldn't be hard, and is decades overdue:
When the user tells the computer to "delete" something, the machine should totally delete it. Not recoverable, not traceable ... not there anymore. At all. Really. That's what most people intend, most of the time, anyway.
I recently read that the only way to completely erase something from a computer is to physically destroy the hard drive, and I think that's contrary to every user's intent. I want to "really delete" stuff, using my judgement. I own the computer, I bought the software, I put stuff on, so I should be just as able to take stuff off.
To not really delete constitutes misbehavior, and perhaps fraud, on the part of the software maker. If some software engineer has secretly decided otherwise, that's a huge disservice to the customers.
My computer seems to have created about 200,000 files without my telling it to. I can't tell what's in them. Were those things I deleted, or would want to? Obsolete versions of everything? Every website URL or contents ever viewed? Software I've replaced? Clear those out of my computer permanently! I want my computer to be fast again, and I (rather than the software maker) know what I don't want to keep.
I've heard that stuff that common users, like me, think we've deleted, but that the computer actually has kept, is sometimes used to sue people. That's bogus!! If a user has instructed the computer to delete something, and that user lacks the technical skills to recover it, it's functionally deleted and the law should recognize that.
I've also heard that such "data recovery" is used to destroy the privacy with which many people view things meant to be private, and meant to be kept private. If so, the software design and marketing is more guilty than the customer suckered into using it.
© Norman Sperling, October 13, 2011
For many years, I assigned my college students to write a 5- to 10-page term paper as part of their course. I wouldn't approve any topic unless I thought there were good enough sources to produce that much content.
After doing their research, some students had a lot of information and a lot to say about it. They often found rich resources I didn't know about, and sometimes contacted scientists directly. These students often wrote very thorough papers, but to cram everything into a mere 10 pages they used type so narrow it was hard to read, a point-size too small to read, and the thinnest margins their printer would allow. They also edited out not only fat but meat, and the resulting paper suffered.
Other students didn't find much, and had only a little to say. Sometimes they found 8 books but they all quoted the same original research. Such students padded their narratives beyond reason, used the fattest, biggest type they could get away with, with very wide margins and only a thin column of type down the middle, just barely dripping a couple of lines onto page 5. The resulting paper suffered.
So I changed my directive. While 5 to 10 pages was the initial target, when they had finished diligent research, I told them to "write it for what it's worth": include everything that ought to be included, and then stop. Don't leave out anything useful, but don't pad either.
The result is papers of a far wider range of quantity, but a significantly higher quality. These papers aren't artificially stretched or compressed. They feel comfortable because the writers weren't compelled to distort them out of all proportion. The average grade went up noticeably simply because all the papers could be right-sized for whatever the writers found. The students are happier because they aren't squeezed, and get better grades. And I'm much happier because the papers I read are well-proportioned, which makes for better reading.
So, when I took over JIR, I made this the rule for the magazine, too. We receive submissions in an extreme variety of lengths, from one-liners up. I only rejected one submission purely for length (it was 54 pages long, and our whole magazine is just 36 pages per issue). But all the others are pretty much the right length for what they attempt, and don't feel cramped or puffed out. That improves the quality of the magazine for readers, eases the constraints on the writers, and improves my reading and editing. Win-win-win.
So I recommend that the same rule be adapted as widely as practical. Try it, you'll like it.
© Norman Sperling, August 28, 2011
"Calling the Roll" has been a standard part of class in practically every school, worldwide, for centuries. But someone with an awful sense of proportion now fantasizes it to be an "invasion of privacy". It never was and still isn't. Nevertheless, an education system has recently prohibited its instructors from using students' names in class - calling the roll, handing back tests, and so on. This inhibits the actual conduct of classes, and reduces teachers' opportunities for learning students' names.
Privacy for students' names inside the classroom is a bizarre concept. I can picture some instructors resorting to student numbers or row-and-seat numbers. Treating people as numbers instead of names would be far more offensive. In all my decades as a classmate and instructor of thousands of students, the only problems I've seen with student names stem from pronunciation, never privacy. (My son heard of a student with a name so exotic that, when he saw a teacher squint and stall while calling the roll, he responded "present" before they even tried to pronounce it.)
Privacy for students' grades is desirable and achievable. Mark the grade on a part of the paper that is concealed, perhaps by folding, when handing the paper back. Students often react loudly, saying "Hey, I got a B+!" but let it be the student who tells it out loud, not the instructor.
Decades ago, Steve Wozniak dropped out of college because he was gaining fame and fortune with Apple Computer. Years later he wanted to switch to school teaching, which required a college degree. He returned to the University of California, Berkeley, under a pseudonym, with the administration's approval. He blended in well, made friends by being friendly instead of rich, earned his degree, and enjoyed his new profession. Colleges may not even know which names aren't genuine. To achieve privacy in the exceedingly rare cases where using a true name would violate it, use pseudonyms.
To cope with the situation, I tell my students to "Pick your 'public name' to be called by in class. You may use some version of your real name if you wish. You may create a pseudonym for any reason, such as privacy or humor (and you don't have to tell why). Use that 'public name' on your quizzes. If it doesn't obviously relate to a name on my official roster, privately tell me what 'public name' corresponds to what 'roster name'."
So far, no student has shown any need for privacy. One made up a different amusing name each week (like "Ty Gur"). Another assumed the name of a fiction hero but soon switched to his own. What will students do this semester?
© Norman Sperling, August 4, 2011
3 articles in 3 days have exposed hoaxes and scams.
A bizarre story claiming that users of Microsoft's Internet Explorer browser are a lot dumber than users of Firefox, Chrome, and Opera, was unmasked in a day or 2. Wired's Epicenter reveals the hoax and sparks its perpetrator to claim it was a joke.
The horrifying "collar bomb" in Sydney, Australia, was a hoax. Who concocted it?
For 140 years, Scots have been proud of their unbelievably-loyal dog, Greyfriars Bobby. Reuters reports that it was a "scam to lure tourists".
Do the media you read tell you the initial claim, but not that it was a hoax? Time to smarten your news sources.
Reviewed by © Norman Sperling, July 21, 2011. Published in The Journal of Irreproducible Results, v51 #4, August 2011.
The Stray Shopping Carts of Eastern North America: A Guide to Field Identification, by Julian Montague. Published by Abrams Image, New York, 2006. www.hnabooks.com . 0-8109-5520-2. $17.95
Scientific classification principles can be applied very widely. Artist Julian Montague applies them, with droll irony, to the situations in which stray shopping carts are found around Buffalo. He classifies their condition, their origin and distance from it, and how they apparently came to the places where he found them. Montague's shopping carts progress through categories as weather, vandals, and snowplows batter them. Every example is photographed, with the author's classifications and occasional brief comment.
Shopping carts typically stray to the grimier parts of town, so the setting is often along railroad tracks and creeks, amid graffiti-covered walls, tires, underbrush, trash, and snow. Montague systematically excludes humans from his photos - only 1 or 2 can be discerned in distant backgrounds. This casts an "abandoned" feel over Buffalo.
Montague does not classify or give any taxonomy to the carts themselves. His classification deals with where they are found, not their inherent characteristics. In doing so, the book resembles astronomer J. Allen Hynek's attempt to categorize reports of encounters with extraterrestrials. "Close Encounters of the Third Kind" made a splendid title for a good movie. But it was never scientifically useful because it did not classify extraterrestrials, which was what we wanted to learn about, but rather how far they were from humans at the time of encounter, which is far less interesting and often accidental.
Montague's book can be used to demonstrate principles of classification in an amusing way, without getting tangled in Latin, Greek, or scientific technicalities.
© Norman Sperling, July 13, 2011
Marty Halpern, another editor, has blogged in More Red Ink about a time when he and I disagreed about stylebooks, among other things, while serving on a panel at the BayCon science fiction convention. The Journal of Irreproducible Results does indeed use different styles than most other publications. Contributors don't have to conform; if we accept a contribution, we will handle that hassle.
Not following the Chicago Manual of Style is NOT an error! The Chicago Manual is hardly the best way to present humor - it's dull and sober and stuffy, the very antithesis of humor. Many editors detest that stuffy antique. Its followers seem like sheeple who mindlessly obey what emperors dictate, even though they can recognize clothing if they see it.
Here are some of our style standards, with some of the reasoning. We welcome other publications and writers adopting any parts of these that appeal to them.
Body type: 11-point Bookman Old Style.
Captions, By-lines, and Sub-heads: 16-point Century Gothic.
Our own advertising: Rockwell.
Bookman, Century Gothic, and Rockwell are the most-readable fonts we have. We use them because we want people to actually read our magazine. Semi-condensed fonts such as Times are harder to read. They cram more text onto the paper, but savings from the printer come at a cost to the reader, and we think the reader is more important. We particularly note that many readers are elderly, and as we age we sympathize with their vision difficulties more and more.
When there is just one table or figure, call it "the table" or "the figure", not "Table 1" or "Figure 1".
Digits are far easier to read than the words for them, and the principal point is ease of reading. Numbers are as tall as capital letters. Spell out "one" except when it is used mathematically as a digit. But all higher numbers should be expressed as digits, even if beginning a sentence.
0 can be ambiguous. If it's clearly the digit, use the digit. If in danger of being mis-read as the letter 'oh', would "zero" work more clearly?
"20th Century", "17th Century", and so on sound stilted, require a mental calculation to subtract to get the dates ... and are often misunderstood, especially by non-Western people. Almost always, they don't mean the specific, technical inventory of years starting with '01 and ending with '00. Almost always, they just handwavingly refer to a century-or-so. It's far clearer and simpler to say "the 1900s" or "the 1600s".
Punctuation in Quotation Marks
Punctuation that is part of what's being quoted goes inside quotation marks. Punctuation that is not part of what's being quoted goes outside of quotation marks. That way you know what's being quoted.
One contributor notes that JIR people seem to have more letters after their names than in them. For JIR's college-educated and technically-oriented audience, 100% understand "%" and are therefore slowed down by seeing it written out as "percent". For people with so many degrees, the same goes for the degree sign.
NASA, US, PM, etc.: full capitals, no periods. Styles that put them "down" were meant to save expensive labor on Mergenthaler linotype machines ... which nobody has used for decades. Instead, let's save clarity.
Cities which are very well known and unambiguous need not be followed by their state, province, or country.
Almost all capitals, and major-league cities (in major sports) are that well known and unambiguous: New York, Los Angeles, Chicago, even Green Bay. Nobody thinks those are anywhere but the big place.
The same applies to intellectually-major-league towns: Ann Arbor, Bangalore, Berkeley, Boulder, Chapel Hill, Charlottesville, Corvallis, Eugene, Evanston, Huntsville, Ithaca, Laramie, Lawrence, Leiden, Los Alamos, Norman, Oak Ridge, Palo Alto, Pasadena, Pune, Princeton, Provo, Rolla, Stony Brook, Tempe, Tucson, Uppsala. [How many have you visited? How many have you spoken at?]
Well-known unique names of smaller places, too, need not be followed by a state name: Albuquerque, Altoona, Amarillo, Bar Harbor, Baton Rouge, Bemidji, Cape Town, Castelgondolfo, Chattanooga, Des Moines, Duluth, El Paso, Fresno, Frobisher Bay, Galveston, Kalamazoo, Kokomo, Little Rock, Macon, Mobile, Muncie, Nairobi, Olduvai, Omaha, Oshkosh, Paducah, Perth, Sacramento, Santa Fe, Saskatoon, Schenectady, Spokane, Tallahassee, Terre Haute, Thule, Timbuktu, Tulsa, Walla Walla, Yakima.
Places that are not well-enough known, regardless of how distinctive, must stipulate the state, province, or country. Faaa, Iquique, Kamloops, Kano, Pismo Beach. [How many of those can you place?] When in doubt, add the state or country name.
When ambiguous, stipulate the state or country name: Alexandria, Athens, Austin, Berlin, Cambridge, Hyderabad, Kansas City, London, Macedonia, Manchester, Moscow, Oakland, Oxford, Peoria, Portland, Rochester, San Jose, Santiago, Springfield, Valparaiso, Wilmington. [How many of those have you been in 2 of? How many Springfields?]
Universities and other institutions which name their state should avoid repeating the state name after the city: "University of Oklahoma, Norman"; we don't need to say "Norman, Oklahoma" because we just said "Oklahoma".
For hyphenation at line breaks, the upper fragment of the word has to be pronounced pretty close to the way it is in the whole word. Fragments that are pronounced differently cause discordance in the reader, badly interrupting the content.
Usually capitalized, when meant as names of major, important fields: Science, Nature.
Usually capitalized, when meant as names of specific celestial places: Moon, Earth, Sun, Universe. Earth is the proper name of this planet, not merely a handful of dirt. Capitalize it the same way you must capitalize Venus and Mars, the planets on either side of it. I'm an astronomer so I can state that authoritatively. Lower-casing the name of this planet just because it's the home of the Chicago Manual of Style is a great insult to the 6 billion humans here, including all of our customers, most of whom have grown rather fond of Earth.
© Norman Sperling, June 21, 2011
I was talking business with another proprietor of a decades-old science business. We both have loyal customers. For both,
* most pay by check, instead of credit card or PayPal
* a few fill out their checks by typewriter (yes, in 2011)
* a few don't like the prices' .95 and round their checks up to .00
* none of their checks bounce.
The "typewriter" aspect no doubt marks people who have not fully adopted computers, because hardly anyone else keeps a typewriter handy. Customers who have been loyal for decades are, by definition, older, so this is no surprise.
The preference for checks is not just a failure to adopt newer technologies, since credit cards became very common by the 1960s. Some feel less secure about giving out their credit card numbers.
Their checks are always good. They want the product, and they'll want the next one too. They're quite content to transfer the money. They don't begrudge the price. It is tempting to read into this the high ethics of science, too.
Paying the rounded dollar instead of the .95 shows these customers' rationality overpowering their emotional reaction to the price. I've never heard of this happening outside of science businesses. Lots of business have thought of charging rounded prices, and many have tried it. Sales slump horribly. Customers only buy when prices end in .95 or .97 or .98 or .99. Only merchants that end their prices that way survive. I once raised the price of an item from $4.35 (which was determined by standard pricing formulae) to $4.95, and thereby markedly increased sales. That's the way customers want things ... except for this extreme intellectual fringe, who are so repulsed by that practice that they send $27.00 for a $26.95 product. They have no way of knowing what a pricing formula would actually call for - the formulaic price might actually end in .31 or .78 or anything else.