Apple is a different situation. Instead of openly announcing that they were turning the product line over to another manufacturer, Apple announced that they were obsoleting all legacy code by switching to a completely different CPU architecture. They did that once before, but with a significant difference: all old code would continue to run unmodified, using a 68K emulator that still ships as part of the "Classic" mode of OSX. Rosetta, the dynamic translator Apple reportedly plans to offer, does not do 68K "Classic" code, and Intel chips are too slow to make serious PowerPC code emulation usable. Almost everything I do is on application programs that run in 68K emulation -- and the vendor (Apple!) declines to provide OSX code that comes even close to the power of these venerable products.
"The soul of a Mac is its operating system," Jobs is quoted from his WWDC keynote. Right, unix -- properly pronounced "eunuchs," which are people missing a vital organ, so they cannot perform; the system is aptly named. The real MacOS, the only commercially viable WYSIWYG operating system that ever existed, sold its soul for a mess of pottage. It's dead. Unix is a 35-year-old system; OSX puts a heavy layer of pancake makeup on to make it look modern, but underneath it's still the same aging, text-file-oriented, command-line driven, geeky system it was in 1970.
What will happen when Intel 86 clones are the only game in town? There will be only one choice: Windows or unix. We have that today: Microsoft takes 95% of the market, Linux 3%, and Apple and Sun duke it out for the other 2%. Sun is (slowly) moving to Linux. Microsoft will hold their 95% at least until the end of the decade -- basically until another Federal government takes up its anti-trust litigation again (obviously not on this President's watch). Linux will continue to get better and hold their 3%. A few rich geeks will choose the (Apple) unix that is slightly easier to use, but unless Apple offers a dual-boot (Windows+unix) computer, they are going to need more than an ugly "Intel inside" sticker to grow their market.
One blogger "leander" argues that the real reason is that "Apple wants Intel's new Pentium D chips" with hardware digital rights protection. Unfortunately, Apple is playing catch-up in this game; Microsoft is way out in front. Microsoft is not in the habit of letting other players into their game. Apple will need to invent something as radical as Mac (the 1984 version, not the current unix in drag) or iPod to stay in business.
Me, I think I will sit this dance out. Apple lost me when they killed
the only system that gave me an order of magnitude productivity leap --
and it still beats out all its imitators and successors. Linux sucks, but
if all the alternatives are locked up and unmaintainable, it may be the
only game in town. sigh
What Colson does not say is that passing a marriage amendment is futile. At most it will slow the Court down, but eventually -- for all the same reasons Colson details -- the Court will simply nullify the amendment even if duly passed. Don't get me wrong, I think that it is important to pass the amendment, and to do so quickly, before the Court takes pre-emptive action. However the recent Court has shown an increasing willingness to ignore the express will of the American people, and to look for precedent in other countries, not just any countries, nor the average of all of them, but rather to those particular countries whose novel laws support the personal agenda of the justices. The Constitution no longer means anything at all.
When -- not if, but when -- that happens to marriage, I think the churches in America should rise up and refuse to participate in anything to do with this newly redefined label. We should boycott the whole idea of government-sanctioned marriage. We can still have weddings in the church, but as a religious ceremony only, where the pastors simply refuse to sign any state marriage certificate. It would be meaningless anyway. If people want tax benefits (what benefits? there is a marriage penalty tax) or hospital access or inheritance rights, they can sign up for "domestic partnerships." Leave "marriage" to the homosexuals -- they don't want it anyway, as can be seen in Norway, where this has already happened.
We would need a new name for our religious union -- how about "religious union" -- to distinguish it from the Court-enforced counterfeit, but eventually the whole counterfeit idea will disappear, and we can go back to doing marriage God's way.
In the mean time, support the Federal Marriage Amendment. It probably won't stop the Court, but there's a chance it could.
20.. 30.. 40 minutes later, I lay there in the dark, counting the seconds between flash and thunderclap... There! Only a block away! Another! I wondered how long I should wait before calling the electric company about getting the power restored. I wondered if I would be able to get through to the appropriate emergency responders if lightning should make a direct strike.
Suddenly the whole room filled with bright light and startled me out of my reverie -- the power had just come back on.
Did you ever notice how much our economy depends on reliable services,
particularly electric power? I think it's pretty amazing that everything
works as well as it does.
I picked up a C.S.Lewis book over the weekend, God In the Dock. It's been a while since I read any of his writings, and I had forgotten what a clear thinker he is, compared to the muddled mush that passes for Christian writing today. This book is a posthumous collection of his essays on various topics, to which the editor gave a very British title. In America the corresponding phrase would be "God on Trial." The whole point is that God is not on trial, He is the Judge, not the accused. And yet in our feeble foolishness we arrogate to ourselves the presumption to interrogate and accuse.
I have for a long time recognized the utter foolishness of atheism, the supposition that there neither is nor can be any God or gods, that as Carl Sagan so famously put it, "The Cosmos is all there is and was and will be." A more subtle form of atheism is taught in the public schools (with government funding, contrary to the Constitution!) and practised by many who call themselves Christian, which is based on the supposition that Nature and the real world we live in is the only reality that makes any difference -- especially in science, where it is called "methodological naturalism."
Atheism is logically foolish and self-contradictory, because the only way you can know there is no deity is to have searched the entire universe -- all at once, lest a shy god might be dodging about where you are not at the moment looking -- and if you succeeded in that, you would be God. The more intelligent atheists, realizing this folly, typically retreat to a nominal agnosticism, admitting that we neither know nor can prove that there is no God. But honest agnosticism must honestly search for God, for if there really is a Creator, then we His creatures owe Him fealty. It is no different from when I write a computer program, the program is mine and I have the right to insist that it perform the calculations I intend. If you merely buy (not manufacture) a car, you as owner have the right to insist that the car go where you are driving it. That idea of ownership is repugnant to the modern mind, but distaste does not make it false.
In his first essay, Lewis brings another proof against naturalism, especially the Darwinian variety. If your thought processes are merely the product of random natural chemical reactions, what makes you think that you are thinking correctly? Suppose the chemical reactions only give the illusion of correct reasoning? An article in one of the computer magazines I read discussed a new, more virulent kind of computer virus or spyware now making the rounds, which changes the operation of the system such that the anti-virus programs report that there is no infection -- and if an anti-virus program should try to remove it, the virus prevents that from happening but reports that it succeeded. The only remedy is to completely erase the hard drive and re-install the whole system. Unless we have some external guarantee of logical thinking, any chemical basis of thought is like that virus, reporting (through introspection) that all systems are normal and rational -- when in fact they are corrupt and confused.
The Kansas board of education has been making news lately by trying
to bring some balance and reason to the teaching of biological origins.
The atheists are unwilling to give up their position of honor as the government-funded
Established religion. I think that they, like the Muslims, know that their
religion cannot stand on its own two feet in the face of alternative views
with compelling truth claims such as Christianity, so they must
exclude the infidels (Christians) by force. The Christian God is a God
of Truth, and the real (in this case, Christian) Truth is so powerful that
it wins on its own merits without compulsion. To their shame not all Christians
have believed that, but the atheists and the Muslims sure know it. That
is the reason why modern democracy developed in a Christian culture, and
not under Islam nor Soviet atheism nor any other religion. The woman in
Kansas who reportedly feared "theocracy" was actually projecting her own
religious values: we are already in a theocracy as hostile to the opposition
as any she imagines -- except that her (non)god is currently the dominant
religion.
Dear Mr.Gates:
I recently noticed in a news item that your company abandoned its previous prudent decision to remain neutral on political advocacy issues unrelated to the corporate core business, and capitulated to the shrill demands of a tiny minority of left-wing heterophobic bigots. You have the right in America to advocate any position you choose, but you might consider the relative proportion of your market represented by the 1.5% whose politics you have sided with, compared to the unprecedented numbers who showed up at the polls last November to vote against it.
As an information technology professional, people often ask my opinion on matters related to computers. Although it is not always practical in a monopoly-dominated market, I will nonetheless now advise them to seek alternatives to Microsoft software and game box products.
I believe it is still not too late for you to forge a corporate policy not quite so hostile to your customers' interests, and more in line with the public health needs of the country. Please let me know if your company again chooses that more enlightened way.
Tom Pittman
The Federal income tax code is based on the 16th Amendment to the US Constitution, and actually quotes from it for its authority. And the US Constitution is the supreme law of the land, right? Not really. A few years ago a political cartoon showed the Justices cutting paper dolls out of the Constitution. It's almost that bad, except they don't use scissors.
The 16th Amendment authorizes Congress to "collect taxes on incomes, from whatever source derived." That sounds like Congress has no remaining Constitutional limits on what kinds of income they can tax, doesn't it? The Court placed a limit -- but only one, and only to protect their own wallet. See my essay for the details and links to the relevant decisions. However, because they did that, Larken Rose can truthfully talk about "Constitutional" (actually Court-imposed) limits on the tax, and he is eager to cite the decisions where they mention such limits -- except he does not cite the one decision which explains the only limit, nor does he mention that the Court was divided in rendering that decision (a significant minority argued that the words should mean what they say).
From a legal and precedent perspective, all bets are off; the Court can rule any darn way they want, and their decisions override what the laws actually say. From a practical perspective, Rose's 861 argument is a lost cause.
Do you remember the 2000 Presidential election? It came down to a few disputed ballots in the State of Florida, and the Supreme Court stepped in and called it for the Republicans. When the recount finally completed, it turned out they had guessed correctly, but the real eye-opener was that for all the various Justices' philosophical preferences, every one of them voted against their own traditional logic and for their political preference. When Larken Rose's case comes up for trial, he will lose, no question about it. The Justices (and lower judges) will rule their own selfish wallets, just as they did in Evans v. Gore to protect their own salary from taxation, only now -- knowing full well that they are paid from the public treasury funded by the income tax -- they will rule to uphold the tax. Ever since 1913, and only with very tiny exceptions that are not generally applicable, the Court has upheld the income tax applied to all incomes from whatever source derived. Including domestic income. Despite Rose's serpentine argumentation and out-of-context quotations, the law and the Constitution are both pretty clear; the courts will have no difficulty finding against him.
But even if it were difficult to convict him, even if Rose were technically
correct, he would lose anyway. The government cannot afford to lose the
revenue generated by the income tax, and the judges know that. It pays
their own salary.
Somewhat more relevant, in the same column, Nisley reports on a reported intrusion to Linux, and how hard it is to catch such things. He concludes, "this level of involvement far exceeds the abilities or interests of most PC users." This, mind you, is not the known swiss cheese security of Windows, but the supposedly much higher security of Linux. The whole Unix way of doing things -- which for internet purposes is the same on the Windows platform -- is broken. The PC I bought to develop software on/for, I won't let it anywhere near the internet. It's networked to the Mac, which is my primary computer and becoming safer all the time, but when I'm on the internet, the PC is turned off. Even if I were to forget, the Mac does not know how to bridge between the dialup internet connection and the ethernet connection to the PC. Everything I read tells me how wise that decision has been. No spyware, no viruses, no adware, nothing gets through.
The next column in the back-page series is Jerry Pournelle bemoaning his own spyware infection, and how long and complex it was to get rid of it. He concludes (wrongly, as we just saw), "it wouldn't have happened if I'd been doing all this on a Mac [meaning Unix, nobody but me uses the real MacOS any more] or a Linux box."
The amazing thing to notice here is that these are the experts. The internet is doomed.
If I didn't have my plate full of another project, I certainly would spend the effort to engineer an internet connection that worked right, and was thus impervious to all these malicious attacks. Fortunately, this Mac still has a lot of life left in it -- hopefully enough to finish the current project and get my OS project running before the hardware here needs replacement. And fortunately also, I don't have to spend a lot of time and effort fighting off malicious spyware and viruses (or repairing the damage).
For all the technical sophistication of the magazine, the DDJ web site is exceedingly intrusive, so I don't recommend you expose your system to it. Most web sites want to set a couple or at most a dozen cookies, but DDJ kept trying continuously for hundreds of attempts. I don't allow cookies on my system at all, but I want to know what scum is trying to load them into my computer. There are no cookie attempts at all from my web site here, none. No viruses, no adware, no obnoxious flashing animated banners, nothing but pure information. I don't collect private information from you or your computer. What you tell me in email is personal email, it doesn't magically turn into spam.
And you spammers, I know who you are, and when my new OS is up and I
can control such things, your spam will bounce (no such address).
I'd do it now, but internet protocols was what I planned to learn the third
year at the university -- and there was no third year. Oh well.
One of the regular columns in the current issue of Computer magazine describes the phenomenon he calls "pecking order." Not content with just describing (and deprecating) the social behavior, Colwell also tries to give it a religious ("Where did we come from?") explanation. By the end of the essay he grudgingly is willing to admit that social hierarchy serves a useful purpose.
Thinking about Colwell's religious ("evolutionary") aside, I realized that I would also be inclined to give this observation a religious explanation -- except of course my religion is Christian, so I would argue that there is an inherent heirarchy in God's created order, with God Himself at the top. Jesus implicitly affirmed this view of the universe when he agreed to heal the centurion's servant remotely [Matt.8:9,10]. Therefore I would start out with the same conclusion which Colwell grudgingly reaches at the end, and I without feeling the need to deprecate it. Actually, Darwinism supports social hierarchy too (in a rather brutal manner: "red in tooth and claw"), but most Darwinists cannot truly live by their own religion.
Because religion answers the important questions of life, religious wars are essentially unavoidable, diplomacy being only "war waged by other means." I wish no greater battle with Bob Colwell than to observe how so few people can actually live their religion, but I often find myself at the receiving end of religious attack.
Case in point: three weeks ago I posted a remark to this blog on the logical shortcomings of the tax protesters. One of the group's more prominant defenders, perceiving (I would claim) the bankruptcy of their position in the face of logic, devoted a significant part of his critique to ridiculing my religious postings. It comes with the territory. An experienced trial lawyer (whose name I cannot remember) once reportedly advised his younger colleague: "When the facts are on your side, argue the facts. When the law is on your side, argue the law. If neither is on your side, just argue." The "861" tax protesters have neither the law nor the facts on their side, but oh my can they argue.
Bob Colwell mostly argues the facts, which makes his column a good read.
There is another small but apparently growing class of citizens who also don't like long lines -- at least not the line on April 15 -- which they solve by not mailing anything at all. They don't like the term, but the authorities call them "tax protestors." They claim the U.S.Constitution and the Internal Revenue Code does not require citizens to pay tax on domestic income, based on a misreading of Section 861 of the tax code. They are wrong.
I won't embarrass by name the fellow who brought these folks to my attention, but my immediate reaction was from what I call my "BS Detector." It usually lets me know I'm hearing a crock of baloney long before my slow cognitive processes get around to figuring out why. Their fatal flaw was surprisingly easy to find. Some of this analysis made it into my Tax Essay.
I suspect the same holistic, right-brain thinking stood me well in high school and college national math tests. A few days ago a family member called with a math problem, something like this:
A man spent 1/6th of his life as a boy, then after another 1/12 he grew a beard, and after another 1/7th he got married. 5 years later his son was born, but the son lived only half as long as his father, who died 4 years after his son died. How old was he?I think they expect you to write an equation
x/6 + x/12 + x/7 + 5 + x/2 + 4 = xand then solve for x. Me, I see all these fractions, and I realize that a good math problem comes out even, so the man's age must be divisible by both 12 and 7, the product of which is 84. I leave it to the reader to verify the correctness of this solution.
The progressive tax code we have in the USA is a good and sensible way to collect more taxes from the people who can afford it more. We are a rich country, and we have far more left over after paying those taxes than 99% of the people who ever lived, including almost everybody else in the world. The tax code may stop working as more and more people grow up without the Puritan Ethic which made this country and its technological superiority possible, but given that today the tax code is sensible, Larken Rose and his followers somehow had to be wrong. It only remained to find where. It wasn't that hard. Reading laws is something like reading computer programs, except that there is no computer to debug them, so there are more undiscovered bugs.
I'm glad I'm not a lawyer. sigh
I looked at what the local medical facility put together for an "Advance Health Care Directive" but it seems mostly about when and how to kill the poor schmuck who signed it. So I wrote my own.
Let me know what you think.
Jesus Loves me, This I know,The university bell tower -- not really bells, it's all electronic fakery -- are playing Sunday-School children's songs as I write this; when the wind is right, I can hear it. The nice thing about electronic fakery is that it can be programmed once and then ignored, to repeat its script mindlessly over and over like the robot that it is. When I taught on campus, I never noticed any repetitiveness (they mostly played hymns, which I enjoyed singing along with), but they lost a number of faculty the year they fired me, among them apparently the kappelmeister. Shortly after the spring semester ended, the bell tower started playing Negro Spirituals -- the same tunes every day, all summer long, and then all fall, all the way into early December when I guess somebody noticed that they weren't Christmassy enough and put up a program of Christmas carols which played over and over all month. In January they replaced the program once again, now these children's songs over and over for three months and counting. Do you think it will last until Christmas?
For the Bible tells me so.
The famous theologian Karl Barth reportedly replied to a question, "What is the gospel?" with the first two lines of the children's song quoted above. I don't know why he should have been convinced of Jesus love by finding it in the Bible; so much of his teaching denied the simple truth of what the Bible says, but this line is particularly interesting because in many decades of looking, I have yet to find exactly where in the Bible it actually says that "Jesus loves me."
The Bible is pretty stingy about whom God -- and especially Jesus -- loves. Three particular people, most often the writer of the 4th Gospel, but also Martha and the rich man who went away sorrowful and never became a follower, received that designation. Once in John 14 Jesus promises to love (in the future) whoever keeps his commandments, and twice in the next chapter he tells the disciples that he loved them (in the past) as an encouragement to keep those same commandments. Twice more in chapter 5 of his so-called Ephesian letter Paul mentions "Christ loved us" (again past tense) as exemplar of the kind of love we (and especially husbands) ought to have. That's all. Nothing at all in the present tense "Jesus loves me [now]," and even the love that the Bible does tell us about seems to be conditioned on our keeping his commandments. Wow.
On another occasion [Matt.25:32] Jesus described how he plans to separate the Good Guys from the Damned, and it isn't on the basis of their "faith" nor their correct theology, nor even Jesus' love for them, but rather on whether they went about doing good as Jesus did on earth. The blood of Jesus on the Cross surely wipes away the karma of past sins, but if you don't repent and start doing good instead of sin, all the faith in the world won't do you much good on Judgment Day. Jesus' love does not extend to those who willfully continue in sin.
This I know, For the Bible tells me so.
Bookmark this item
Additional comments
A couple years ago, partly in connection with teaching a VisualBasic class at the university, I developed some useful programs in that tool. Dot-Net was still pretty new, so I based my teaching and work on VB6. While not as easy to use nor powerful as Apple's HyperCard (used to be, another vendor-killed viable product), VB6 sure beats C/C++ or even Java. It's also much faster than either one.
Last year, when my access to the university computers and software base was about to disappear, I bought my own PC and specified that it should run VB6. No way Jose, it was not available. The vendor sold me VB-Dot-Net, which is completely incompatible with no upgrade path at all. It has a tool apparently intended to do the job, but it failed to import a single line of a 3,200 line VB6 program that used nothing but standard VB6 components.
My conclusion: Taking their non-support of VB6 as a predictor of the future, I am forced to consider Dot-Net also to be totally unsupported and inappropriate for mission-critical software development. A great pity.
Obviously, I am not alone. The InfoWorld article has the developers
pressuring Microsoft into announcing additional upgrade tools to help users
migrate onto Dot-Net. Nice try, but too little and too late. I plan never
again to let third-party proprietary products hold my software hostage.
"Fool me once, shame on you. Fool me twice, shame on me." Apple suckered
me the first time with the Mac and then with HyperCard, both outstanding
products that outperformed all their successors -- and then killed them.
Then VB6, in a much smaller way. Quoth
the Raven...
I don't have a TV, but I had an opportunity to borrow some DVD movies recently, on the foolish supposition that there's a DVD player in this computer. Well, there is, but it's every bit as stupid as the standalone players, maybe worse. It refuses to run at all in my normal software development configuration, I have to reconfigure the whole computer and reboot a few times, which among other things messes up my window configuration. Ah the price of watching a free movie! As author Robert Heinlein once put it, TANSTAAFL (There Ain't No Such Thing As A Free Lunch). The player software also got stuck in that start screen, but the same idiot who programmed the standalone play did this program, so the repeated sequence of Play-Stop-Play-Stop three or four times broke it out of its catatonic autism.
I noticed as I took the disk out of its carrier, that it was slightly scratched. Oh-oh. A VCR player knows to do just exactly the right thing when the tape is scratched or wrinkled: it keeps on playing, albeit with snow or crackling sound or perhaps some tearing of the image. And it keeps going. DVD players are so stupid, they just stop. For small scratches, the image freezes, but the sound continues -- or maybe drops out for a few seconds like digital cell phones (as with analog VCRs, analog cell phones keep working when the signal degrades, just with some extra static; digital cells just plain stop working completely). As I got a little farther into the movie, I guess I got to a bigger scratch, and the whole player just froze. After about a minute it finally let me pull down the play menu to tell it to skip to the next "chapter" (fast-forward still froze). Further into the scratched disk it just hung the computer. Nothing would recover. I could hear the drive seeking back and forth endlessly, "braap, braap" and doing nothing. I pressed the restart button on the computer. Fortunately this computer still has one of those buttons; lesser hardware (running much lesser systems like Linux, which normally cannot recover from a crash) don't have that escape, and you must pull the plug out of the wall. Even the power switch on those computers only politely but futilely asks the dead computer to shut itself off.
Come to think of it, cable TV must have gone digital also. When I was still watching it (last year) right at the show's climax, all the action decoding evidently froze up the satelite signal scramblers or something, because right at the climax, without fail, the image would freeze and the sound would break up for five or ten seconds. Then it would resume. Those computers are (slightly, not a lot) smarter than DVD players.
I'm thinking of becoming a Luddite.
Go figure.
Like so much of American business, the lawyers have put a stop to anything legitimate that people might want to legally do with their clients' products. Most people just ignore these unenforceable "licenses" and use the software the way the vendors' ads seem to imply is their purpose (and which the EULA always explicitly denies). Of course they can't actually catch anybody violating these terms unless they come prying into everybody's bedroom and den. Oh, but the EULA also gives them that right at your expense. Figure two corporate lawyers at $500 per hour each, including travel time from the other side of country. You didn't know that, did you? Most people don't even read that garbage, and well they shouldn't; it's unnecessary and abusive. Me, if I agree to something, then I abide by it. It's the only honest thing to do. I guess they assume there aren't very many honest people. I guess they might be right. Otherwise they couldn't sell their products.
Therefore, don't bother sending any ShockWave movies my way unless you can figure out how in good conscience to send me a player that I can in good conscience install and run on my computer.
And maybe some day, somebody will actually challenge these "agreements"
in court, and get them back to what they should be. But I'm not holding
my breath waiting.
A mutual friend recommended Raskin's book to me, which I carefully read then engaged in a long email dialog with the author. I'm afraid I am not easily persuaded by dismissive remarks which assume I do not understand the author's point. I usually get this particular kind of criticism when I do understand the author and some of the implications that the author missed. No matter, Jef is now gone, and nobody will be quite as persuasive of his ideas second-hand as he was. On the other hand, maybe they can be less self-absorbed and therefore get a better hearing.
There's a lesson in there someplace, I think.
Full disclosure: I think the Macintosh that Apple actually sold to the public was far superior both to the idea pictured on Raskin's website obituary and described in his book, and also better than the 30-year-old Unix system with a thick layer of pancake makup which Apple currently sells under the same moniker. So I'm writing my own Mac-like OS; I guess that puts me in competition with the Raskin project -- except he has funding and I don't.
Bookmark this item
Additional remarks on Raskin's "Leap Key"
The very first microprocessor only processed data in 4-bit chunks. That's enough for one decimal digit, not even a whole alphabetic letter. Shortly after that, we got 8-bit micros. Bytes are usable, one letter or other printable graphic each, and a pair of them will address 65,536 bytes, which was more than anybody could afford to put into a microprocessor system. For about a year. Program size very quickly outgrew that boundary, and the IBM PC took one approach to expanded memory, in 64K "segments." Motorola had a better idea: 32-bit registers, which could (in principle, although the first implementation was more limited) individually address more than 4 billion bytes. Motorola's better idea was a couple months later than Intel's goofy segments, so IBM went with Intel, and we all continue to pay the price for the next 25 years and perhaps beyond.
We still today do not have any programs 4 billion bytes in size, and not very many people want to pay for that much RAM. There are a small number of programs that need to access that much data randomly, so they have a problem with 32-bit registers, which any good programmer could solve with an imperceptable performance cost. Oh wait, there aren't any good programmers, they all write in C/C++. OK, most database programs really do run somewhat slower when enlarged to handle more than 4 billion bytes of data. Those few server applications will run faster in a computer with 64-bit registers. Everybody else... sÄ q†ƒ´@ÿ
Real programs dealing with actual data, well, how big are their numbers, anyway? Text data is one byte each; 64-bit registers hold eight of them, but you mostly can't manipulate 8 text characters. You get one character, see if it's the one you want, then go to the next. Numeric data, ordinary numbers in ordinary programs mostly are less than 10,000 and easily fit into 16-bit registers. An accounting program might want to handle numbers up to $10,000,000.00 (in integer cents, so there's no round-off error), which fits easily in 32 bits. More than that and you are not running the data on your desktop computer, you have an accountant do it. But mostly we just do not operate on billions of individual objects; the human mind boggles at numbers like that.
What about pixel data in image processing? Or weather data? Yup, there it helps to do a lot of crunching in parallel. But we already do that in existing 32-bit computers. The pixel data is already being processed in 64- and sometimes 128-bit chunks, nevermind how wide the native register size is. Making the registers bigger will have absolutely no effect on those programs. But 64-bit processors do have 64-bit memory width, so we get data from memory twice as fast. Twice as fast as what? The 32-bit PowerPC ten years ago already pulled up memory data in 8-byte chunks. Again, the register size is irrelevant; the chip makers can give us the same performance advantage in a 32-bit processor.
The test results are starting to come in: 64-bit computers are slightly
slower
than comparable 32-bit computers. I could have told you that in advance.
I did say almost exactly that, a decade ago. The biggest bottleneck in
modern computers is memory bandwidth, getting data out of RAM and into
the processor. Memory often runs ten or twenty times slower than the CPU
clock. Cache helps, but not a lot -- especially if your program is written
in bloated C++ and has a lot of data. What happens when you double the
register size? All your numbers are now wastefully stored in 64-bit memory
words instead of 32-bit words (which are already too big most of the time),
so it takes twice as much RAM and twice as long to get it into the CPU.
And the program runs slower.
I'm not afraid of the truth, but dishonest people are. There was another person who told the truth. It got him crucified. My employment situation is trivial -- almost comical -- by comparison.
This is the condemnation, that light is come into the world, but people love darkness rather than light because their deeds are evil. -- John 3:19Let your light so shine... -- Matt.5:16
Case in point: a few years back, after it became clear that the MacOS was headed for extinction, I spent some effort trying to learn what I could about its survivors, Unix and Windows. Unix was hopeless: Apple didn't seem to want to make their version available, and Linux was (still is) beyond comprehension. But I succeeded in ordering a 10-year-old used 5-volume set of Win32 Programmer's Reference manuals. These document the system calls for developing programs to run on Win95 and later, but they don't tell you how to use them. A year ago, preparing to teach a class on programming in C, I ordered an evaluation copy of Charles Petzold's Programming Windows, 5th edition (1998). The Microsoft documentation is so bad, there's a great market for 1500-page tomes like this. It's that "job security" thing again: by providing full employment for people like Petzold, Microsoft ensures their own success (selling computer systems that force programmers to invest a huge cognitive capital on getting started). I had the same problem with VisualBasic two years ago. Petzold's book is nowhere near complete, but at least it gives you a basic understanding of how the system works, so you can switch over to the reference manuals to do what is needed. I got my virtual machine up and running in native Win32 in just a few days.
An important part of the BibleTrans program (which is where I'm going with all this) is DragonDrop, the direct manipulation of data by clicking and dragging. This has been working on the Mac for more than 6 years; I need to make it work substantially the same on the PC for it to be usable. Unfortunately, there is nothing about DragonDrop in either Petzold nor the reference manuals. I found a couple dozen entries in the search index for VisualStudio, all irrelevant (FoxPro and C# and VB and Dot-Net). Browsing the "windows.h" header files provided by the compiler, I was able to determine that it probably did support the feature, but there was no way to get access to the documentation. Remember that "job security" thing? Better products don't sell. So I Googled "Win32 drag drop" and turned up a number of clearly written articles. Some were inaccessible (I refuse to turn on the virus enablers in my web browser), but James Brown's Catch22 website was just exactly what I needed -- and his code is all in the public domain. I can now do DragonDrop in my own code on the PC, and it even runs in Win95.
I don't know how to thank James Brown for his generous contribution
to the betterment of humanity -- there is no email address on his web site,
perhaps because he runs Windows himself, which is notoriously too dangerous.
I tried sending something to <JamesBrown@catch22.net> but it
bounced, "no such domain". Oh well, Thank you anyway, James Brown, wherever
you are!
2. What is the most important attribute of God?
If you answered "Love" to the second question, guess again. When you measure importance by how much space devoted to it in the Bible, or by how much time Jesus spent teaching it, or how much attention his early disciples gave it in the beginnings of the Church, love is so far down the list as to be completely off the radar most of the time. It's there, but it doesn't get the attention given to it by modern Americans.
My concordance lists a little more than 500 verses (out of a total 31,218 -- about 1.5%) in the Old and New Testaments with the word "love" in them. Looking at just the Greek word for divine love, agaph, there are 272 verses in all its forms. Less than 10% of these are about God's love, half of them in just five chapters from the pen of one Disciple John: two from his gospel and three from his short first epistle. The word does not appear at all in any form in the book of Acts, which is what the Twelve -- make that Eleven -- Apostles did and taught when the church was starting up. By way of comparison, there are about the same number of verses mentioning some form of "righteous" and again the same number mentioning "holy" but a much larger percentage of those verses are about God's righteousness, and almost all of the "holy" verses are about God or things related to God.
So why this emphasis on love? I don't know, but I suspect it's a modern form of self-righteousness: "I'm pretty good, it must be so because God loves me." Or as the popular children's song puts it, "Jesus loves me this I know, For the Bible tells me so." Well, the Bible says no such thing.
Are you ready for the answer to #1? Those three words do occur in the Bible, but not as a statement of present fact. Paul uses the phrase in a conditional, "If I love you, ..." when he is scolding the Corinthian church. Jesus comes close: three times he says "I have loved you" (past tense) to his Disciples, and once "I will love you" (future), all in one evening, all in the context of urging them to remain true and faithful because he is about to be taken away. That's it.
In all fairness I should make it clear you can infer God's continuing love from the few verses that are there, just as you can infer the obscure and easily misunderstood doctrine of the Trinity from the (even fewer) verses that touch on it, but neither of them get the attention that repentance and doing the right thing get -- all through the Bible.
Nowhere, however, does the Bible even come close to teaching that people should tell other people how much they love them. It doesn't even teach that we should tell them how much God loves them. What we should teach, and what the early disciples taught, and what Jesus himself taught, is repentance. Stop sinning.
The interesting thing about Firefox according to this article, what made Firefox succeed and its parent Mozilla project fail, is that the developers made it small and simple. The author did not say it, but one of the things that gave Linux its jumpstart is that young Linus Torvalds made it small and simple. Linux is no longer small and simple, which is probably why it has pretty much stalled at 3% market share. The author of the article tries to claim that it's the open source nature of Firefox that made it happen, but if that's true, why did Mozilla fail? It was open source long before Blake Ross "forked" off the Firefox project.
The Macintosh operating system started out small and simple. It exploded into the personal computer industry and forever changed how people use their computers. Then it got big and clunky -- and lost its market share. Apple then switched over to a Unix-based system -- and continues to lose market share. They don't get it.
WindowsXP is big and clunky like everything else out there, but it manages to keep its market share. What else is there?
My new operating system is consciously
small and simple. The blurb on the Wired cover concludes, "Watch your back,
Bill Gates." Probably not mine, but maybe another 14-year-old without the
"creeping feature creature" envy of the software establishment will build
another "Keep It Simple Stupid" system that does what people want without
all the remote programming virus highways that make WinXP and Linux so
dangerous and hard to use. Watch your back, Bill Gates.
He describes me as "like a master carpenter who makes beautiful cabinets and does detailed wood work." When people ask me what I do, the story is a little different. Like my friend and probably everybody else in this industry, our jobs are hard to describe to outsiders. To the small-talk-not-really-interested inquirer, I say "I write computer programs," and that usually satisfies them. Most people know that computer programs go inside computers and make the computer do something. Some people have written a Basic program or two, so they can handle the more esoteric answer: "I write computer programs that translate computer programs from one artificial language (like Basic) into computer programs in another, even more artificial language (the ones and zeros that computers need to see in order to run)." That usually blows everybody away. But it's true. I'm not so much a cabinet maker as a tool maker. My specialty is building the toolbench that the cabinetmaker uses to make cabinets. Yes, I sometimes build my "hammers" more often than I buy them.
My friend sells "hammers" and I didn't buy his. But you do need hammers to make hammers, and I do from time to time buy software development tools. I might have bought the product he sells, except I took a long hard look at it and I couldn't figure it out. Or maybe I did figure it out, but it was more like a jackhammer with attached air compressor on an attached dumptruck with a front-loader scoop. It didn't fit in my garage shop, and it wasn't very useful for making cabinets nor workbenches. That's OK, because we need jackhammers and dumptrucks to do heavy construction. I'm a toolmaker and I could probably make those kinds of tools also -- but not alone, they are just too big. We have (very few) big factories to make those kinds of tools.
Pressing this hammer analogy just a little bit more, I could point out that ordinary hammers are made in (very few) big factories also. I bought one, but when I went to use it, I discovered the head was on cockeyed. All the hammers out of that factory are crooked the same way, and the carpenters who use them just figure that's the way hammers are. It dents the wood around the nails pretty bad unless they twist their wrists just so, but that gives them tendonitis so sometimes they just let the dents go. Oh, and I almost forgot to mention, the cockeyed head puts a lot of stress on the handle, so the tool simply breaks after a few months of use and they have to go buy another. I called it "job security" (for the toolmaker); replacing the dented wood products provides job security for the carpenters, and curing their tendonitis gives employment opportunities to doctors and nurses and physical therapists. 20 years ago I bought a really good "hammer" that never broke, never dented the wood, and didn't give the users tendonitis. I bought a second one as a spare and never used it, because the first one still works fine, 20 years later. Really. Nobody buys those hammers any more, because everybody who wanted one is still using the first one they bought, and the vendor stopped making them. Oh, and it doesn't work with Properly Crooked (PC) nails, which are the only kind you can get, because they work best with the cockeyed hammers everybody has.
There is such a thing as well-designed, well-built tools, and I have bought and used them. I still have them on this computer in front of me. But the computer in front of me is a Macintosh, not a PC, and I cannot write software on it that runs on a PC and is easy to install and use. So I bought a PC with the industry-standard tools for software development on a PC. It's on the desk here, right next to the Mac -- silent and dark most of the time. I cannot use those tools to write software for the PC that is easy to install and use, either. Maybe some programmers can, but I have never seen any of their programs. Do such programs exist? Maybe my friend has seen them. Or maybe everybody has been using cockeyed hammers so long that they only know about cockeyed hammers.
My wrist hurts just thinking about it.
Back when there was such a thing as a Mac operating system, I used to say "The MacOS is like the US Government, way ahead of whatever is in second place, but all parties are diligently trying to close the gap." The US Government got some of their act together and turned around for a while, but Apple finally managed to pass up Microsoft in their relentless backward run. Not exactly a good horse to have my wagon hitched to, but here I am, a Mac expert in a world without Macs.
So now I diligently read PC-related magazines. I think there may still be one or two Mac--I mean unix-related mags out there, but why bother? Unix is where the industry was, 40 years ago. There are still a lot of unix geeks out there, many of them writing for and editing technical magazines, the magazines I'm now reading. The technology coming out of Microsoft is so bad, even unix looks good to the people who peer under the hood. And then there is the job security.
Windows is a crummy system, but you can teach office workers to use it, and they can do so with only a small amount of expert support every couple weeks to clean off the many viruses and fix the other many problems that the system invites. And there are a lot of experts who make a living doing that. That, my friends, is job security. Like going into health care because so many Americans are obese and doing other bad things to their bodies, so they will always need somebody to come patch them up, or going into police and detective work because we no longer teach "Thou shalt not steal" in schools.
Another kind of job security in the computer business comes from the obscure nature of the systems we build. The managers don't know what their programmers are doing, but they can easily see they are very busy typing away on their keyboards. What they cannot see is that all that clickety-click actually slows their workers down. Research (with a stopwatch) shows that mousing is faster than typing, but it's quieter and sounds less productive. Typing obscure commands also occupies more of the worker's attention, so they feel like they are working faster -- like the California commuters who regularly get off the 30-mph freeways to drive 15-mph on city streets because "it's faster." No, it only feels faster, because it takes more attention. Yes, they really do that there. Unix workers have this special insider kind of job security, because it takes so many more of them to do the job that fewer Windows users can do in the same time. I told my students that at the university; maybe that's why the department chair was so glad to see me gone.
The Mac was bad job security even when it was a viable system. It worked so well that users never needed to upgrade their software. The software vendors can't make much money in that kind of market. The median lifetime of a Mac was nearly five years, compared to a PC, which tended to get upgraded every year and a half. Can you imagine what would happen to the American economy if cars and clothing lasted five or ten years instead of needing replacement every year? Unemployment would be awful. But it wouldn't be so bad, because all those unemployed people would still be driving cars that run and wearing clothes that looked good. Unemployment is a funny thing: we could eliminate it entirely and very simply, just by outlawing tractors. The (former) Soviet Union had an unemployment policy that worked that way, but their people liked our way of life better. All parties are diligently trying to close the gap. Unlike Apple, we have much farther to go backwards as a nation before we pass up whoever is in second place.
Anyway, so here I am reading all these technical magazines that I never looked at for 15 years. It's amazing how far the technology can diverge in that time -- or maybe it's just the terminology. All these acronyms and technical terms that I have no clue what they mean. I think they are designed to obfuscate and enable the insider readers to feel smug about their inside knowledge. I have a 20-year-old PhD in computer science, and I can't get more than a half-page into these articles without getting hopelessly lost. Take "SOA" for example. That stands for "Service Oriented Architecture" (which name tells you nothing at all about what it means), but it really seems to be about writing a bunch of little programs that do just one thing -- kind of like the unix commands, which were invented 40 years ago. It's not new, but if you understood that, you wouldn't pay high prices for their technology. Job security. It also keeps the riff-raff (people like me, who can write a good program but don't know the buzz words) out, so there's less competition for their jobs. Job security again.
Is this the kind of world you want to live in? What are YOU doing to
make it better? Tell
me, I'm curious.
The ISP is easiest to explain. Most likely a consequence of the Federal "CAN-SPAM" law recently passed, the volume of unwanted and unsolicited email has increased exponentially. If Congress would give consumers the right to sue the spammers for punitive damages, instead of taking that right away (as they did), then spam would disappear. However, running for Congress is expensive, and disreputable businesses are willing to pay for it, while consumers are not. As a consequence, we have the finest government that money can buy. Ask your local Congressman who paid for their election.
Anyway, spam is up and the poor ISPs are at their wits end to block it. The local ISP here bought some new software intended to do that. But when they installed it, it stopped accepting outgoing email from me and my mother. The tech support at the ISP blamed it on the client software on my computer. Internet protocols are not something I know a lot about, but I know a lot more this week than last. One of the things I had to do in trying to track this problem down was completely hack my email client, so I fully understood its preferences settings -- if only to prove that the problem wasn't on my end. It wasn't. I have been using Netscape on this computer without any problem at all since I got the computer, almost a decade ago. It only failed for this ISP after they told us they were changing their software. Netscape on the PC seemed to have a workaround, and we got my mother back online. In my case -- well, the tech support person plainly told me that they don't support Netscape. What else is there? Microsoft. Microsoft products are the cause and facilitator of most of the spam out there. Microsoft is the known lowest common denominator for computer security problems. I would not allow my mother to use any Microsoft product for accessing the internet. So it would seem that the ISP policy is to aggravate spam and other security problems in the name of fixing them.
So I got another ISP. Netscape works fine.
While I was thrashing around, I sent an email message to my mother from an old CompuServe account I had kept around for just such emergencies. The local ISP called it spam and bounced it. It was icing on the cake, because I found out after I had already jumped ship.
I'm a computer expert, not a banker, but I know a little about accounting. One of the inviolate principles of sound fiscal policy is called an "audit trail," a record of every transaction so that if something goes wrong you can track the problem down and fix it. Most -- and until last fall, all -- banks give you with your checking account an audit trail on every check they pay from your account: who you wrote it to, who endorsed it, what bank accepted it, and when. That audit trail is mostly on the back of each check, which they keep a copy of, then return to you for your records. Of course handling all that paper is expensive, and the bankers would much prefer to pocket that money instead of spending it on verifying that they are acting in a trustworthy manner. So they lobbied Congress to get the law changed (your tax dollars at work, again), so they are not required to return the cancelled checks. I don't exactly know what changed, because they were never required to do that; they just did because smart customers insisted. I do know it was the bankers who wanted the change, because my local bank here tried to foist the idea off on me more than a year ago. I refused, and for a year I paid extra for the priviledge.
Three months ago they told me that the law gave the digital images the authority to serve as proof of payment. I'm sorry, folks, but a tiny pixelated image of the front only of a check I wrote has no audit trail, no proof of payment at all, and besides that, it is easily forged. I'm a computer expert, I know these things. I also refuse to sign the touch screen at Wal-Mart, because that digital image is so easily forged. I make them print out a piece of paper to sign. They also know that the digital image is worthless in court, and for big-ticket purchases, their register automatically spits out a piece of paper to sign. If the new Federal law gives legal standing to the digital images of the checks, the first time a high-powered handwriting expert goes up against them in court, the banks -- or more likely, you and I, their customers -- will lose. Anyway, for three months now the bank wasn't even giving me tiny pixelated digital images of the endorsement, despite my continuous and vehement protestations. Nothing. When and if this practice becomes widespread, you will hear about the next big bank scandal. All it takes is somebody with a little more greed than integrity (nobody like that anywhere in this virtuous country, is there? Want to buy a bridge?) to realize that you have no way to prove he got your money. The banks are not at risk, they sent you the monthly statement, and you didn't complain. Read the fine print on the account disclosure you signed. I sat there in the bank today for a half hour, reading the fine print, while the manager waited. How are you going to complain about diverted funds, if they don't provide the necessary audit trail? By the time you get past the hassles from the creditor you thought you were paying, the 60 days are past and you just lost it. If like me, you have been complaining all along about the lack of an adequate accounting, you might persuade a jury that the bank is at fault; otherwise, you lost it.
Except for this one local bank with whom I am in the process of severing
all ties, all the banks I deal with return the endorsed checks.
And when the Federal law takes that little piece of paper away, I will
insist they replace it with a comparable audit trail. I will do business
with the banks that serve my needs. They make their business decisions,
and I make mine.
It's a Lie, first told in the Garden.
"No man is an island," the poet tells us. He's right. We are not, nor can we be, truly autonomous -- especially in today's industrial world. We continually depend on products and services supplied by other people, and if we are not nice to them, they won't be there when we need them. That's true of God, too, but God is more gracious and forgiving than other people -- so far. The time will come when all the atheists' denials will make about as much sense as a modern Luddite (also known as "environmentalists") refusing to live in houses or wear clothes or eat food made with electricity. It can't be done, no matter how much we pretend.
But today I am thinking about another aspect of The Great American Faith, which impacts the market for computer products. Every couple years, the technology wizards and prognosticators announce the impending demise of the personal computer, which, they tell us, will be replaced by terminals connected to central computers. In the 1960s computers were too expensive for everybody (including almost every business) to have one, so the idea made sense. That was the last time it made sense.
By the 1970s minicomputers had come down in price to something affordable by small to mid-sized businesses. I know, because I wanted my own computer, and was watching the prices creep down to the magic $1000 mark. Then the microprocessor happened, and prices went through the floor. Everybody who wanted one could have their own computer. That is still true today, only more so.
So what makes the prognosticators think otherwise? Data. Every business needs central control over their data. Right. Every business has their own central computers to hold and process their data. The control freaks in upper management would like that to be the only computers in their business, but it's not going to happen. Why not? Autonomy. Every low-level manager wants to control his own data. As I recently discovered to my economic chagrin, nobody in any business of any kind really supports the stated corporate agenda. Their personal agenda is always uppermost in their priorities, and they support the corporate policy only enough to keep from getting fired. And they make sure that anybody who conscientiously supports the corporate agenda significantly more than that does get fired; such people make everybody else look bad.
From a practical perspective, centralized data management makes good sense, and a few more enlightened businesses will choose to go that way. The rest of the country -- and probably the rest of the world -- will go with autonomy and local control.
Why is this interesting? I read InfoWorld, a computer industry news weekly, and they are constantly promoting "Service Oriented Architecture" (another buzz-word meaning centralized data processing). They do this because all their big advertizers want to sell big-ticket products (the median price on the computer products and services they review seems to be about $10,000). Centralized corporate Information Technology (IT) departments have big budgets for high-priced goods and services. Selling a hundred $10,000 server systems generates the same revenue as selling a thousand $1000 computers, but the cost of marketing them is only one tenth, so the profit margins are much greater. There is this constant market pressure between the vendors and their bean counters, who want to sell fewer high-priced items, and the public, who want to buy a lot of low-priced items. Frankly, I wouldn't want to be in either market. The high end vendors keep getting their lunch eaten by the low-end products, and the low end vendors' profit margins are so thin they cannot afford to stay in business. IBM is getting out of the desktop computer business.
The same is true in software development. The marketplace exploded when programmers could afford their own computer, and could therefore write software that ran on personal computers. I have a friend who sells (I think it is) software development tools and methodologies for IBM. He is always telling me how the only game in town is client-server applications. He may be right -- that's certainly the song they are singing at InfoWorld -- but it doesn't fit with human nature nor The Great American Faith. I see instead a zillion tiny (one-man) programming shops writing in VisualBasic or whatever else they can make sense of, doing their own autonomous thing. And for a while all the growth was overseas, but there is this (ahem) autonomy thing in American businesses: they want control over their programmers, which you can't do when they are in India or Russia. The back pages of InfoWorld are once again filled with employment ads.
Me, I work on small, stand-alone software for individual computers.
I may be out of the industry limelight, but I don't see my future as darkly
as the industry "experts" do.
He's right of course about the Bible and pride, but my professional specialty is semantics, the meaning of words and sentences, and I got to thinking: Being proud of a person for whose actions you cannot really claim responsibility is not quite the same as being proud of your own work. Thus it is entirely appropriate for him to tell his wife that he is "well pleased" with her, while it would still be sinful pride to be "well-pleased" in his own preaching (which he did not try to do from the pulpit), no matter what euphemisms he tried to cover it up with.
Today this is relevant to my own situation. With characteristic programmer optimism, I guessed it would take maybe two weeks to take the virtual machine I'm building my new operating system on, which has been developed on the Macintosh, and convert it to Windows (or unix or whatever). Most programmers -- myself included -- have such a high opinion of their own skills that you need to take their "conservative" estimates and double it before you see anything that runs longer than a couple seconds without crashing -- and then double it again before you can use it. Well, I started on my PC version of the virtual machine January 5; today, less than a week later, I'm already past the two-week stage. There are still a few bugs, maybe it will yet take three weeks to dredge them out. God only knows.
Let me further clarify one important fact: I believe in Entropy. Nothing comes out of my fingertips and into the computer that I didn't get from somewhere -- probably from God -- so there's not much to be proud over. I'm just doing what God gave me to do, and He also gave me the ability to do it. He could take it away again, just like that. But I am (ahem) well-pleased with the progress so far. It's better than I have any right to hope for.
Later in the same sermon the preacher made a complimentary personal
reference to me by name, so I thought I would pull his chain a little afterwards
by accusing him of feeding a sense of "well-pleasing" in me, but he didn't
get the joke. Probably too obscure. Pride is pride, no matter what words
you use.
Complete Blog Index
Last year's blog
Itty Bitty Computers home page