USS Clueless Archives

  USS Clueless

             Voyages of a restless mind

Main:
normal
long
no graphics

Contact
Log archives
Best log entries
Other articles

Site Search

Stardate 20010714.2349 (On Screen): As part of the continuing Republican initiative to abrogate the ABM treaty, there was a test today where a mock warhead in sub-orbit was intercepted and destroyed. They say.

I'm highly skeptical about this whole business, but that's because I know something of the history of military weapons testing. The problem is that the tests aren't really what we engineers would ordinarily consider reasonable, which is to say that they don't represent real-world or worst case circumstances. Rather, they are deliberately chosen to be as benign as possible, usually to the point of being completely unrealistic. The purpose isn't to prove that the system would really work, it's to get brownie points to take to Congress to get more money for the program.

So, for instance, there were tests done of a surface-to-air missile intended to hit an enemy jet. They were tested by firing at a drone, a pilotless plane flown by remote control. The drone flew relatively slowly, in a straight line at a constant speed, and these are things a real hostile jet would never do. Worse, the drone contained a transmitter and the missile which hit it was homing in on that transmitter -- and generally speaking, hostile jets don't tend to have such transmitters. A more realistic test would have had the drone maneuvering violently, not carrying a transmitter, but on the other hand actively using counter measures (such as dropping flares or ECM). That's what you would do if you were really trying to find out if the missile was any good. Problem is, it would have failed an embarrassing amount of the time and the program might have been cancelled. The military's traditional approach has been to get the program approved and get the weapons deployed, and then figure out how to make them work, and discard them if they don't. This is, as you might imagine, extremely expensive and the Air Force has been one of the worst offenders. The history of weapon acquisition is full of programs where there were a series of four or five tests all of which failed, after which someone said "That's good enough. Deploy that sucker."

Today's test is extremely suspect. The administration really needed a success, so I wonder just how reasonable a test this was. For instance, the interceptor system seems to have known exactly when the target would be launched and exactly on what trajectory it would be coming. The true target was not accompanied by any decoys and wasn't using any countermeasures. Odds are that it wasn't using "stealth" technology, though a real warhead definitely would do so. But they got their kill and that gives the President the political ammo to stand up and say "See? It works!". I don't buy it.

Only I'm going to have to buy it, because the President wants to use my money to pay for it. I don't think we should build a missile defense system because I don't think it can be made to work, and even if it did it wouldn't remove any serious threat. It couldn't be large enough or comprehensive enough to stop a major attack from an established nuclear power, and any rogue state or terrorist group who had a bomb and wanted to attack the US wouldn't use a ballistic missile to deliver it. This system can't stop cruise missiles, or smugglers. If you defend against one form of attack without addressing others which are simpler, easier and more likely to succeed, you're wasting your time (and my money). (discuss)

Update: Yup, it was unrealistic, alright.


Stardate 20010714.1613 (On Screen): One of the interesting ways to differentiate nations is by observing how they deal with cases where the government or its agents make mistakes. For example, what if you arrest and charge an innocent man? Some nations will hold a trial, deliver a verdict of "not guilty" and then release him (and occasionally suffer a civil suit for false arrest). Other nations will hold a trial, find him guilty anyway, but set a penalty which isn't really a penalty such as "time served" for a native or "deportation" for a foreign visitor. (In the worst cases, they find him guilty anyway and lock him away.)

Finding an innocent man guilty because you've arrested him is a form of lying. It says "We really were right to arrest him. We didn't make a mistake. See, he got convicted, didn't he?" It's a way of covering up. But I think it runs deeper. Nations whose governments are afraid to admit mistakes are nations where the governing elite is afraid that it is losing its grip on power. As with people, nations who admit their mistakes forthrightly are the ones who are most secure and confident in themselves. (discuss)


Stardate 20010714.0832 (On Screen): Sometimes I think that there's a deeper justice going on in the universe. For example, when someone falls for a blatantly obvious hoax and loses a bunch of money as a result, I find myself strangely unsympathetic. I know we have to pursue and prosecute the scam artist, but if someone buys oil drilling rights on the moon, then they deserve to get taken for a ride.

I'm not too keen on the idea of someone who attempted a murder being free on the streets, but I also have a sneaky suspicion that someone with the guts and brains shown here really should be free. Or conversely, if the County of Los Angeles is so incompetent at keeping prisoners, then it doesn't deserve to have any. (discuss)


Stardate 20010714.0809 (On Screen): If, like me, you're getting fed up with ICANN's slow, bureaucratic muddle, then you can prove once again that it is the users who ultimately control the Internet. You are not limited to the DNS's which toe ICANN's party line; you can use other domains even if ICANN doesn't like it. So tweak ICANN's nose by loading new.net's plugin; it works fine. It's an ActiveX control and if you're using Win2K you have to be logged in as Administrator to install it, but once you do you'll be able to get to all the new top level domains that really should have been approved by ICANN. It's also Mac-compatible, and one of the new sites is www.mac.tech. For the rest of us (ahem) this opens up such sites as www.fishing.sport, www.guitar.mp3, and www.maxim.free. If your ISP has decided to activate new.net anyway you might not need the plugin, but most have not. But who needs them? You can do it yourself. Strike a blow for freedom, brothers and sisters! (discuss)


Stardate 20010713.1832 (On Screen): By the end of World War II, the Japanese fleet was a no-show. Nearly every ship they had was sunk or damaged. About all that was left were a handful of submarines, but such few as there were of those were still quite deadly because they were armed with the war's best torpedo, the dreaded Long Lance. Still, the Pacific Ocean was huge and there weren't all that many submarines, and they hadn't been an effective threat for the last year of the war because they had spent most of their time delivering small quantities of supplies to Japanese garrisons which had been cut off by the American advance. So it's somewhat understandable that after delivering the components for the two nukes to Tinian, that on 30 July 1945 the heavy cruiser USS Indianapolis wasn't too worried about submarines. But one found her and hit her with two torpedoes, causing her to sink.

Most of the crew died. 316 were rescued. The first friendly to find the crew as it floated in the water was an American PBY Catalina, a two-engine amphibious plane heavily used for reconnaissance and for air-sea rescue work. Hundreds of American fliers owed their lives to the PBY's (affectionately known as "Dumbo") and the men who flew them. After this one landed, it used its engines to taxi around as its crew pulled sailors out of the water. It never even tried to take off again; instead it loaded up with as many sailors as it could possibly carry, even to the point of having them laying on its wings, while its radio operator frantically radioed for help. Eventually surface ships showed up and rescued the remaining men in the water.

883 men died. Inevitably someone had to be blamed and the captain of Indianapolis, who was one of the survivors, was courtmartialed for it. In 1968 he committed suicide (and became one of the last casualties of the war). Now he's been exonerated, as he properly should be, though it makes no difference to him.

There was a crime for which there should have been a court martial, but it wasn't his. War is a dangerous business; sometimes ships sink. The real crime was that Indianapolis wasn't missed, and no search was ever ordered. It didn't get a chance to radio an SOS. The PBY which found its crew was on routine patrol, it wasn't looking for her. Indianapolis was supposed to report in by radio periodically, and when she stopped reporting no-one noticed. A ship that size doesn't sink rapidly, and there was time to abandon ship. She took fifteen minutes to go down, and about 800 of the crew of 1199 got off her before she sank. And they floated in the water, waiting for rescue which never came. Over the course of four days, the captain encouraged his men, trying to keep their spirits up as they became thirsty and hungry and cold and tired and gradually lost hope. And men lost strength or gave up, and one by one nearly 500 men slipped beneath the waves, never to be seen again. Maybe he wanted to die, too, but he owed it to his men to stay alive to help them stay alive, as they waited for someone to notice that their ship was gone and to come look for them.

If anyone was courtmartialed for not noticing that Indianapolis stopped reporting, I've never heard of it. But that's who really should have stood before a court. (discuss)


Stardate 20010713.1447 (On Screen): Iowa Attorney General Tom Miller says that after New Mexico's settlement that the remaining states who are plaintiffs in the Microsoft antitrust trial are still united and resolved to proceed. Why is it that this strikes me as whistling in the graveyard? "Of course we're united, aren't we fellows? Aren't we? Aren't we? (I can't hear you!)" (discuss)


Stardate 20010713.1253 (On Screen): Santiago Calatrava is the most revolutionary architect since Frank Lloyd Wright. His designs for bridges are beautiful while being eminently practical. But it is the Seville bridge which takes my breath away. We are used to vertical lines in large structures, something Calatrava avoids. The Seville bridge is not only beautiful and audacious but it is actually a better design than a more traditional approach. Usually a suspension bridge uses a tower in the middle, with stresses balanced on both sides. But that means that the tower has to have its base in the water, which is far more difficult to do and thus more expensive. Calatrava's Seville bridge places its tower on dry land. And a more traditional suspension bridge like the Golden Gate requires immense anchorages at the ends to carry the force of the main cable. Calatrava's bridge only uses direct cable links from its one tower to the roadbed itself. If you want to know what an engineer means by an "elegant design", this is an elegant design. When you see one, you have two simultaneous reactions: "What was he thinking??" and "Why didn't I think of that?" (discuss)

Update 20010714: Lia has actually seen the Seville bridge. (I'm jealous.)


Stardate 20010713.1158 (On Screen): I try to keep my coverage of Rambus to a low level, but two stories about them have hit in the last day both of which are major bad news. (Heh-heh-heh...) Yesterday Rambus announced its quarterly results and they were terrible, and the long term outlook is not good. Revenues are down, expenses (especially legal) are up, and the long term trend is for revenues to continue to decline and for legal expenses to rise.

And now it's been revealed that Rambus has renegotiated its contracts with some specific unrevealed RAM maker to reduce its royalty payments. This can't be good for the company, since royalties are its sole source of income. This would be in response to the legal decision in the Rambus versus Infineon suit that Infineon SDRAM did not infringe any of Rambus's patents, and also that Rambus had engaged in fraud to get those patents. As a result, Infineon will not have to pay Rambus royalties for SDRAM sales in the US. In that trial it came out that Samsung had a clause in its royalty agreement permitting its payments to be reduced to match those of any other competitor, and it's suspected that this agreement was with Samsung, but regardless it's going to ripple through to all the other licensees, and Rambus's royalties are going to take a dive. As I write this, RMBS is taking a beating on the NASDAQ. Of course, it's been taking a beating for a long time and its price curve looks like a Linux company, not like a real company which actually has a product to sell. (discuss)


Stardate 20010713.1125 (On Screen): Well, if nothing else was going to kill Napster as we knew it off, this will do it. Judge Patel has ordered Napster to stay offline until it can guarantee that no unauthorized copyrighted material will ever appear on its system. (This really shouldn't surprise anyone.) The article asks rhetorically if this is even possible.

It's possible but not practical. The only way it could be done is by changing from a disapproval system to an approval system. Instead of permitting songs through unless they're recognized and blocked, the system would deny all songs unless they're explicitly recognized and permitted. If it was a file exchange system, that would mean that a trusted (hah!) Napster staffer would have to listen to the song and confirm that it wasn't anything public and then enable it, which is not economically viable.

Only it actually is, and that's what we're about to see. The next incarnation of Napster (if it survives) will not be as a music exchange system. It's going to be a simple music distribution system, where music moves from Napster to its users but not from users to each other. All the music Napster vends will be licensed, and it will cost users a small amount of money for each and every song they get. Napster will become an electronic publishing business for music, rather than a file exchange system. What I'm expecting is something akin to an Internet radio, where a user sets up a special client program which receives and plays the music without storing it. The user selects the songs to hear. The billing might be a one-time fee per song for unlimited plays, but I'm expecting a per-minute usage fee instead. (discuss)


Stardate 20010713.1103 (On Screen): I am now and have been for a long time a rabid opponent of mixing church and state. That's because I fear greatly the experience of living in a theocracy. I think some people wonder why I feel so strongly about this. If so, all they have to do is to study what's happening in Afghanistan right now, where the people are living in conditions of thought-control which were beyond even that of the Third Reich. The Gestapo has nothing on the Taliban, folks. They're even trying to make the Hindus wear yellow insignias "for their own protection".

But ideas are pernicious and people want to think freely. Given the Taliban's balls-to-the-wall attempt to enforce mental conformity, they've decided that exposure to foreign ideas is simply too dangerous, and have decided to cut off the Internet from the country, thus joining those sterling examples of free thinking and liberty, Myanmar and North Korea. Access to the internet is in itself not particularly special, but this is the latest in a long series of moves by the Taliban which are becoming very alarming. I think three things are coming in Afghanistan: economic collapse, mass starvation, and a terrorist campaign against the Taliban. (discuss)


Stardate 20010713.1048 (On Screen): So there I was, in a store full of pots and pans, and sitting on the shelf was a brand new coffee maker which contained its own grinder. How could any techno-toy freak resist? There's a hopper in the top and you shovel whole beans into it, load it with water and turn it on. The grinder crunches the coffee which falls into the filter, and the hot water sluices through the grinder to remove any excess grounds or powder before going to the filter. Since the coffee is ground about ten seconds before the first hot water hits it, it's just about as fresh as it can possibly be.

I also then purchased about four different bags of decaf (sigh) beans of various kinds, and one of them was Don Francisco's Decaf Regular 100% Columbian Whole Bean Coffee. I am now sitting here having an amazing experience: I am drinking a cup of coffee which tastes the way that ground coffee smells. (It is very good.) (discuss)


Stardate 20010713.1038 (On Screen): It's another installment of "shooting a flea with a cannon": a study finds an average of 14 distinct alleles for 313 human genes they studied in an ethnically varied group of 82 subjects. How, exactly could that happen when we're all descended from just ten chromosomes of each type? That would be two each from Noah, his wife, and his three daughters-in-law. (I got an email from a creationist yesterday telling me that I must be wrong about it because I don't have a college degree, and since I am "uneducated and inexperienced" [his words] my opinion on the subject is worthless.) (discuss)


Stardate 20010713.0837 (On Screen): Online gambling is one of those social problems which won't go away, so they've been considering it in Congress. They've talked about the idea of making it illegal to use checks, credit cards and debit cards for online gambling (which would pretty much kill it) but have decided that it is unenforceable.

Actually, there's a really easy way to make it stop. Pass a law saying that online gambling debts can't be enforced and that anyone who loses can sue in civil court to regain their money. If that happens, and if there are a number of cases like that where people lose thousands of dollars and then go to court about it (which has happened already) then the online casinos themselves will "voluntarily" stop using those payment methods. (discuss)


Stardate 20010712.2006 (On Screen): NIST has developed a new form of atomic clock which is accurate to one second in fifteen billion years. This is a preposterous degree of accuracy, and in fact they've reached the point where they can measure the relativistic time compression effects which would occur simply by moving this monster across the lab and back. (Assuming they had two of them and the other sat still.) For that matter, if there are multiple instances of these beasts, they'll measure time differently as a function of the latitude and altitude that they're at, due to relativistic effects of the rotation of the Earth. Sheesh.

In fact, this clock is not as accurate as they say. What's happening here is a confusion of repeatability and accuracy. If a measurement device is very "repeatable" then if it is used to measure the same phenomenon multiple times it will provide you with a very consistent result. If it is very "accurate" then multiple instances of the device measuring the same phenomenon will give you very consistent results. Usually a measurement device will have a better level of repeatability than accuracy. Accuracy is a hell of a lot harder, and I don't think it's possible to measure time accurately to the level of error described here. In fact, I know it can't be done because it would require an absolute frame of reference, and Special Relativity says there is no such thing. (discuss)


Stardate 20010712.1943 (On Screen): I always thought it was a damned good thing for the Daleks that everywhere they wanted to go, even in limestone caves, had perfectly flat polished floors. As powerful as they were, something like stairs would have defeated them. (discuss)


Stardate 20010712.1131 (On Screen): I have no kids but I don't like getting smut spam any more than anyone else does. But this article describing a proposed law to make it illegal to send such spam to children without a tag which could be used by parents (and me) to filter. It also would impose immense penalties for doing so. It is really quite pointless. The last line of the article reads "It is unclear at this time how e-mails originating overseas would work with such a tag."

Actually, it's completely clear: it would not be enforceable. If this law passes, some spammers will start using the tag (and will see their effectiveness drop through the floor) and others will go offshore to do their spamming, or do their spamming in a fashion which is untraceable. If you receive a letter advertising Gloria's Lovely Lesbian Farm Girls (yum!) or some such thing, how do you prove that it was really sent by Gloria? Remember, there's presumption of innocence in the US legal system, and if Gloria covers her tracks well, there's nothing that can be done.

This is the latest example of a long line of cases which points out the fundamental truth of the Internet: it's beyond the control of any single country's laws. The global village has arrived, and no-one governs it because it's bigger than any nation. (discuss)

Update: "A European Parliament committee has voiced objection to a proposal supported by most European Union member states that would virtually ban unsolicited commercial e-mail." Virtually-ban is right, because it sure wouldn't really ban it.


Stardate 20010712.1116 (On Screen): "White House budget chief Mitchell Daniels blamed the economy Thursday for shrinking federal budget surplus projections." Of course, a massive tax cut couldn't possibly have anything to do with it, could it? Hmmph. (discuss)


Stardate 20010712.1035 (On Screen): The "greens" are on the wrong side of this issue. I have been running into a lot of people who are afraid of genetically modified foods, even to the point of making up funny nicknames for it such as "frankenfoods". Such people show both ignorance of history and an apalling lack of understanding of the world situation. First, what most of them don't realize is just how heavily genetically modified everything we eat now already is. Humans have been crossing plants with each other (including doing cross-species pollination) for thousands of years. (The Maya had a saying that their maize crop would be improved by having a local grass called "teosinte" growing nearby. It's now known that maize is in fact a mutated form of teosinte and modern plant geneticists have been using wild teosinte as a source of genetic variability for quite a while to improve maize in a number of important ways.) Starting big-time about sixty years ago, genetic modification of food ceased to be something done at random by individuals and became a form of organized engineering. In the 1960's there was a real fear that by the year 2000 we'd be facing mass starvation (i.e. cumulative starvation deaths of more than a billion people). You may have noticed that it didn't happen. That's because the plant geneticists produced miracle crops which were then distributed widely, to the point where for a while India actually became a net exporter of food (instead of the place where chronic famine was expected). But the problem is not "solved", though much good has already been done. And while the overall situation is much better, you can still have small, or even microscopic, disasters.

The Green objection to GM food seems to be predicated on the following logic:

  • There is plenty of food. (Just look at how high the shelves in the grocery store are stacked!)
  • Genetically modified foods are qualitatively different than anything which has come before
  • There's a chance that one of these experiments might go wrong and ruin some food crop.
  • Since there's a risk and no potential for gain, we shouldn't do it.
If you go talk to a poor farmer in Kenya or India, though, you won't get the same story. For them, loss of their crop on their particular five-acre farm is disaster. There may be food but they won't have money to buy it, so if their crop fails, their children will starve. There's no single solution to this (and pouring huge amounts of money into the Third World is not the answer) but genetically modified foods can make a really big difference. We know that because they already have.

One of the things which has to be realized is that there is no fundamental difference between the results gotten with the new genetic techniques and the results which were possible before using straight cross-breeding and radiation and chemical mutagens. However, the new approaches do have the ability to focus better: it's the difference between a searchlight and a laser beam. The changes which can result from the new techniques are no more radical than before, it's just that the old techniques involved making a lot of variation and then preserving whatever came out of that which was best (whatever it happened to be), while the new techniques actually permit design. There are a number of ways in which this may help.

The single biggest cause of deforestation in much of the world is for firewood. This is a particular problem in Africa. In areas where there are few (and soon no) trees, the biggest source of burnable materials to prepare food and keep warm at night turns out to be dried manure (which burns quite well). But this is a disaster for farmers because they also need to use that manure to fertilize their fields. Since you can't eat most foods without cooking them that has to come first, and that has resulted in a decline in yields in many places as the land becomes exhausted from lack of fixed nitrogen. Some food crops, however, can fix their own nitrogen. This includes legumes, clover, alfalfa and certain other plants. None of the grains can do this, though, and grain is the primary source of food in the world. What if we could modify wheat or maize or rice to do this? We might be able to now.

One of the big disasters which can happen is attack by insects. If a 5-acre subsistence farmer has his crop wiped out, it doesn't matter if there is plenty of food around because he won't have the money to buy it, and his children will starve. (Which is why there have been suicides when this happens.) There has already been much progress on making plants which will be able to protect themselves against insects. Given that overall close to a quarter of the food we grow is destroyed by vermin, the potential gain from this should be obvious.

But another threat to world food production is limitations on the supply of fresh water. There are many places where there simply isn't enough, and this has lead to wars. There are also places where there is some, but the use of it causes salt to accumulate in the soil over time. The only solution is to thoroughly flood the land each year to wash the salt away. This means that vastly more water is required than would be needed simply to keep the plants alive, because the plants we need are extremely sensitive to salt and won't grow if it's present. But it doesn't need to be that way. There are many plants who can grow quite nicely with their roots immersed in ocean water, and some of them are grasses (the family which includes all grains). And there are a lot of areas where salt water is plentiful, and a lot of areas where salt accumulation in the soil prevents agriculture because there's enough water to irrigate but not enough to wash the accumulated salt away. Suppose we could change grains to be less sensitive to salt, or even to be able to grow when irrigated with plentiful seawater instead of precious fresh water?

Some food crops are nutritionally unbalanced. When they are nearly all of the local diet, nutritional diseases can become common because of this. Corn is missing a couple of amino acids, for instance, and rice is a poor source of thiamine, whose lack causes beriberi. No grain is a good source of ascorbic acid, whose lack causes scurvy. But we might be able to modify such crops as rice and corn to be nutritionally complete (the way potatoes are now). There's no important reason why corn couldn't produce vitamine C; it's just that it doesn't actually do so.

"But we don't need these things!" That's right. We Europeans and Japanese and Americans don't need these things. But Kenyans and Somalians and Nigerians and Malaysians and Indians and Indonesians damned well will need them soon, so we'd better get cracking on developing them now. If we don't, then ten years from now we're going to have millions of people starving, and the same people who are marching in the streets now to try to prevent development of GM foods will be blaming everyone except themselves because of it. (discuss)


Stardate 20010712.0945 (On Screen): How many companies are capable of writing off $3.9 billion in one quarter and still ending up with a profit? Not to mention showing a $700 million rise in gross sales in the midst of a meltdown of its market segment. (Betcha can guess which company this is without following the link.) (discuss)


Stardate 20010712.0728 (On Screen): Oh, goodie! This is just what Representative Condit's spin-doctors wanted to have come out right now. Couldn't be better. In addition to the now disappeared intern, and a stewardess, it comes out that a few years back he was also humping the 18-year-old daughter of a pentecostal minister. Just peachy. I think our hero's political career is over now; he has a snowball's chance in hell of being reelected. (discuss)

Update: Oh, yeah, it's just a marvelous day for his staff.

Update 20010731: Turns out that the claim by that minister was not true.


Stardate 20010712.0701 (On Screen): What with the WinXP's hardware fingerprint algorithm having been reverse engineered, someone is now speculating about the possibility of actually reverse engineering Microsoft's key-generation code so as to make it unnecessary to call Microsoft. The goal would be to make it so that you could say "no" to the request to activate, get told you had thirty days to run, and then to run a handy-dandy utility you downloaded which would then spiff up the registry with an activation key.

If Microsoft did this correctly, it's not going to happen. What happens is that XP examines your hardware and from that generates a fingerprint (as described here) and that gets sent to Microsoft either by direct internet connection or by you reading a long string of digits to someone over the phone. At Microsoft that number is logged, and it also gets enciphered. The output of the cipher is then given back to you and gets stored in your registry. Each time you boot your computer, XP regenerates the fingerprint, and also deciphers the key in the registry, and compares the two. If they match "close enough" then it will let you run. Otherwise it forces you to register again.

But they're using an asymmetric cipher, also known as a "public key cipher". Presumably they're using RSA; the number of bits involved is small enough so that RSA's inefficiency is unimportant. The whole point of an asymmetric cipher is that knowing the public key doesn't help you to find the private key. This article proposes creating a whole series of fingerprint/validation-key pairs and from that figuring out how they're created. That won't help you if indeed they're using RSA. Anyway, he's working too hard; if he really wanted to find out what was going on then the way to do it would be to find the code which interpreted the activation key from the registry and reverse engineer that. Then you could get the public key, and see how many bits it is. If it is a thousand bits (and it's likely to be more) then you can forget about cracking it.

So what other alternatives are there? One would be to actually replace the module in XP which does the validation, to make it simply say "We're validated" unconditionally. The problem with that is that Win2K (and presumably XP) have integrity checks built in now. DLLs and other system code files have checksums and other validation mechanisms and they're checked before being used. This was put in place to protect against trojans and also to prevent so-called "DLL hell", and what happens is that if a bad (i.e. altered) file is found then you're prompted for your install disk or other access to a valid file) and the bad one gets replaced with a good one. So secondarily it would be necessary to figure out how those integrity checks work so that you could make your bogus validation module look correct. At that point you're well beyond the scope of effort that I think a typical cracker is willing to expend. (discuss)


Stardate 20010711.1932 (On Screen): Now this joke was old when I first learned to use UNIX in 1979. You'd find someone's terminal which was logged in, and type cd to get to their home directory, and then type touch \* which would create a file named "*". Someone who was clueless (like, say, an EE) who saw that would reflexively type rm *. I never did that to anyone but I heard of it being done. But it was destructive so it wasn't really a good practical joke. With the influx of new Linux users, though, I wonder if people are still falling for this 25 year old gag. (discuss)


Stardate 20010711.1750 (Crew, this is the Captain): I love baseball. I hate beer ads. (I don't drink.) They used to be dorky-guy-drinks-beer-and-gets-the-girl, which was stupid. But recently they seem to have changed the plotline, and I can't figure out why.

The new story is that some dorky guy meets one or more drop-dead-gorgeous right-out-of-the-pages-of-Maxim there-ain't-no-such women in a bar, tries to pick them up and gets snubbed and slinks away with his tail between his legs. The message seems to be "Drink our beer and you won't get lucky." Why are they telling me that? I can think of only two explanations. Either the real point is to make men depressed so they'll drink more to drown their sorrow, or the ads are actually targeted at women. (discuss)


Stardate 20010711.1547 (On Screen): I get some odd email sometimes, and today I got a letter from a woman saying she liked my site (blush) and telling me about hers. I actually suspect she didn't spend much time looking at my site; I think she was googling for "Friesland" and "Frisia" and "Frisian" and mailing to every website owner where she found the word. Since it appears in my biography page, that's how she located me. Still, the site she sent me was interesting.

I don't think of myself as being Frisian (or even Dutch); I think of myself as being American. Still, the majority of my ethnic heritage is from that small part of the Netherlands, including my surname (which I'm told was chosen by my patrilineal ancestor in 1811). The page has a nice history of the Friesian peoples up until they sort of vanish by absorption in the middle ages, and it mentions the main historical cities. One it mentions is Dokkum. I've been there, on my only trip to Europe.

I am, as regular readers will know, a student of military history. When we were planning our vacation, I demanded the right to choose a few of our destinations. (Otherwise my girl friend would have planned every minute, that being the kind of person I loved her for being.) I had mentioned that I wanted to get into Friesland at least a little while, and while we were looking over the mountain of tourist booklets she'd ordered through the mail, there was one about Dokkum which described it as a "walled city". Well! That was enough for me; I was ready to go. Images of curtain walls with crenelations, and knights on horseback filled my mind. I was psyched.

The reality was not at all what I expected. It was indeed a walled city (the old city, now long since surrounded by a huge suburb) but it was 16th century technology. No curtain walls or crenelations to be seen. It was, rather, a star-fortress. That was perhaps even more fascinating. We spent an afternoon walking the entire circumference of the old city as I described in detail the reasons for everything to my girl friend (who probably got tired of it). So now I'm going to tell all of you more than you want to know about walled cities.

The curtain wall was the end result of a long sequence of development of fixed fortifications whose purpose was to defend against pre-gunpowder armies. A well designed walled city and castle could only be taken by siege. One castle in modern Lebanon was built by the Crusaders and was held for years by a force of about fifty men against a siege by the Infidel (ahem). The weapon of choice for the defenders was the bow or crossbow, and the walls were nearly vertical and very smooth for maximum strength and minimum climbability. Also, a vertical wall made it easy to drop things on the attackers (like molten lead or boiling oil or stones). The 3-8 meter thick walls were two stone faces with the space between filled with rubble and soil, which made them invulnerable to battering rams. Crenelation describes the tooth shape of the fence along the top of the wall; its purpose was to give the defenders places to hide (the high parts) and to shoot through (the low parts) without having to duck, so as to minimize the effectiveness of attacker's bowmen. Castle technology developed over the course of nearly fifteen hundred years, beginning with the primitive "motte and bailey" (little more than a wooden palisade built next to a hill) and ending up with the castles of Wales, which were the ultimate form of the technology.

And then gunpowder appeared in Europe and a primitive form of cannon called a "bombard" was invented, and the game was over for the curtain wall. It turned out to be the ideal target for cannons, and where a pre-cannon siege might take years as you waited for the survivors to run out of food, bombards could knock a hole in a castle wall in days or weeks at most. It was evident that a new form of defense was needed. The result was the star fortress.

The main defensive structure of a star fortress was the moat. While many castles had moats, they were mainly intended to make it difficult for attackers to get to the primary defense of the curtain wall. But in a star fortress, everything is there to defend the moat. The star fortress gets its name from its ramparts, which are like the points of a star. You could use any number but six was common. (That's how many Dokkum has.) Obviously it would be dangerous if the moat could be drained, so you make it so that the moat is part of a river (which is the case at Dokkum) which makes it impossible to drain. The ramparts are low (barely fifteen feet high, if that) with gradual sloping walls made of brick backed by soft soil. You could quite easily walk up one, except that you weren't going to ever get that close if you were an attacker. The primary defensive weapon was the cannon (backed by muskets). Cannons were placed inside the rampart, and they had the ability to hit the ground outside the moat in about 220 degree field of fire, plus the ability to shoot along the wall towards each neighboring rampart. Every piece of land on the far side of the moat was within cannon fire of at least two ramparts (and sometimes three). Anyone trying to cross the moat on boats would get the crap blasted out of them, and anyone trying to attack the rampart with their own cannon were at a distinct disadvantage because the defenders already had their weapons registered, and the attackers had no physical defenses for their guns. The long sloping walls had the effect of deflecting near misses away from the guns they defended, so it took a lucky shot to take out even one gun, let alone the dozen or so which might be there. In the middle of three of Dokkum's six star points there were windmills for purposes of grinding grain into flour, though if they were all destroyed it would still have been possible to grind grain by hand. It was said in the Middle Ages that a castle was only as secure as its water supply, and with a river running through the middle of it Dokkum was not going to go thirsty. And as a final bit of defense, the city itself was actually bisected by the river, which was crossed by bridges. If one part of the fortress really did fall to troops in boats, the defenders could retreat across the river and burn the bridges, and defend what was left, which meant that anyone trying to take Dokkum would have to take it twice. So it was no surprise to learn that Dokkum never fell to siege until weapon technology moved to the point where its defenses didn't make sense. (For instance, it was no defense against Napoleon, who had cannon good enough to make short work of the defenses had they been silly enough to try to resist.)

Dokkum is a beautiful city, and the night we spent there they closed off the center of the town and held a street party, with different musical groups in different places. There was a string quartet, and a gypsy band, but there was another which was even more popular and they were great. We spent most of the time listening to them. So there I was, in the middle of Friesland, drinking Dutch beer, listening to a Dixieland band playing "When the Saints come Marching In" and marvelling about how universal American culture had become. (discuss)


Stardate 20010711.1432 (On Screen): Art is art, but business is business, and sometimes bad art is good business. Everyone's favorite A-lister Jeffrey Zeldman is out there pounding the table for standards compliance again (with his shoe), and he's trying to tell large companies that they should turn away visitors to their sites who are using old non-compliant browsers, for the greater good.

This proposal has been discussed before in numerous forums, but I don't think with regard to commercial sites. There a few good points made here: compact HTML is unquestionably a good thing for a big commercial site, for instance. But the idea of actually turning away potential customers because of their browser choice is ludicrous, and demonstrates a complete disconnect with reality. Standards are valuable when and only when they serve us. A standard is not an end in itself and will not be complied with if the expense is too high. Turning away potential customers is too high a cost.

Successful companies do not create business plans based on ideology. A good business plan is practical and resilient, such as to succeed even in less than ideal conditions. No company survives by telling its customers what to do, or by trying to sell them what they need. All that does is to drive potential customers to your competitor, never to return. Companies which work for the greater good tend to go broke.

Success is gotten by letting them do what they want to do and by selling them what they want to buy. Zeldman needs to study that lesson; he's trying to sell standards compliance, and he's not doing a very good job of it. (discuss)

Update 20010715: Oh, now this is vile. And it appears as a popup.


Stardate 20010711.1406 (On Screen): I haven't been able to decipher just exactly what the police theory is about the Levy/Condit case. It's been more than two months since the last time anyone saw that young woman and apparently they have no idea whatever as to what might have become of her. So now, some ten weeks later, they've searched Representative Condit's home (with his permission). He has, of course, had plenty of time to dispose of anything which might be incriminating (if there were anything like that) so it's not clear just what they thought they'd find doing such a search.

Equally, they're trying to arrange a polygraph test. I am worried about this because the polygraph is much overrated. (I talked about that issue here.)

I think I do know what is going on. Levy's family has orchestrated a truly superb publicity campaign about the story, trickling out news at a regular rate in order to keep it on the nation's front pages for an astounding two months, and have forced the police to investigate Representative Condit quite thoroughly whether there's actually any reason to suspect him or not. I personally haven't got the slightest idea whether he's involved in her disappearance, but I'm pretty sure that what's going on now isn't going to answer that question. (discuss)

Update: Great minds think alike. (Coicidence, so far as I know, but interesting.)


Stardate 20010711.1340 (On Screen): You know, when everyone disagrees with you, even your closest friends, it's time to consider the possibility that you might actually be wrong. (discuss)


Stardate 20010711.1248 (On Screen): Apple has finally given up on the ill-fated Cube, and having announced the termination of the product line a week ago, bargain hunters are snapping up the remaining units (at cut-rate prices.) One customer is quoted as saying, "I'm hoping that enough potential Cube purchasers like me will emerge from the woodwork to convince Apple to reconsider and resume producing Cubes."

I don't think so. These are not the customers Apple wants. Customers that wait until a product is terminated and then snap them up at cut-rate prices are not customers which will keep any company in business. They reduce losses but they do not result in a profit. The customers a company wants are the ones who buy at full price when the product is new, and continue to buy at a steady rate without price cuts. Bottom feeders are not an asset. (discuss)


Stardate 20010711.1235 (On Screen): Now let me get this straight: Microsoft has decided to let its OEM licensees remove IE from Windows. Not just the upcoming XP, mind, but also ME and Win 98 and Win2K.

Didn't someone from Microsoft get onto the witness stand during the antitrust trial and say under oath that it was impossible to remove IE from those operating systems? I wonder if there might be a perjury investigation coming.

(Of course, at that time Dr. Felten [the self-same professor who is now a thorn in RIAA's side] demonstrated that it could be done quite easily.) (discuss)


Stardate 20010711.1220 (On Screen): This gave me a good chuckle. It's not an exploit so much as a form of harassment; its purpose is not to break through a firewall as much as to shut the firewall down. It's a form of DOS, in fact. Still, it's quite clever.

As part of trojan checking, some sophisticated firewalls now check all attachments on incoming email, which is reasonable. Of course, the trojan could be packed into some sort of archive, so necessarily this also has to unpack the archive before it can be checked. Someone figured out that you could take a huge file consisting entirely of the same value and compress it into an archive, and it would compress really well and result in a quite small archive. But when expanded again by the firewall it would explode and bring the firewall to its knees. A 42K archive would explode into up to 16 gigabytes of trash and all that would have to be checked for viruses and trojans. Ship ten or twenty of those at some system and the firewall will be down for about half an hour.

It will, of course, be trivially easy for the firewall company to fix it. All they have to do is make the unpacker smart, so that when it reaches some arbitrary upper limit (i.e. 100 megabytes) on unpacking, then it will give up and refuse to forward the archive. It should take someone about two hours to make the fix, and possibly a day to test. (discuss)


Stardate 20010711.0619 (On Screen): Blackhat and DefCon are held in Las Vegas each year at the same time, and it's obvious that this was deliberate. Blackhat is a convention for system managers to learn how to keep their computer systems safe from intrusions. DefCon (nice name, that) is a convention for hackers to get together to talk about how to break into systems. (It's rumored that the FBI watches this one closely.) I think it is obvious that one of the started happening, and then the other was scheduled at the same time and place to deliberately coincide with it, but which came first? I honestly don't know. (I'd love to know who has the sense of humor.) (discuss)


Stardate 20010711.0615 (On Screen): It is the nature of the political process that it seeks compromise. When facing two groups with radically opposed ideas, the political process tries to find a middle ground where each side makes some concessions and gets part, but not all, of what it wants. (It's been said that a good compromise is one that each side hates equally.) For many kinds of issues this is possible. But there are issues which do not permit compromise, or rather where a compromise solution would be vastly worse than either extreme point of view. This makes politicians extremely uncomfortable because it forces them to choose sides.

These polarizing issues can take decades or even centuries to work out. In some cases they may never work out. One example is racial or ethnic conflicts, where two groups both want the same land that both are living on. Maybe a compromise can be found but it doesn't look hopeful; usually the result is a fullscale war leading to genocide or eviction (so-called "ethnic cleansing"), or a deliberate partition.

President Bush is facing one of those no-middle-ground questions now. As mentioned here a couple of days ago, he's in the process of trying to make a decision about federal funding of research into fetal stem cells, which have the potential to cure several wide-spread and horrible diseases, but which are a considered an ethical disaster to certain groups. There is no middle ground here. And now a clinic has raised the stakes. Until now, most of this research was done using embryos which were left over from in-vitro-fertilization clinics. The procedure for IVF is to treat a woman with certain drugs which cause her to have a stronger ovulation cycle than usual. (It turns out to be an unusual use of birth control pills, oddly enough.) Just as she is nearing ovulation, she is operated on and eggs are removed from her ovaries. Then they are are fertilized outside ("in a petrie dish") and some of them will become fertilized. They are permitted to divide a few times to confirm that they are viable, and then all will be frozen. Typically there are anything from three to ten of them. After the woman recovers from the procedure, one will be thawed and implanted. If it doesn't take, then another will be thawed and used a few months later. Eventually one will take, and then the couple get the baby they want. At which point the excess embryos will be discarded, or used for research. There are biological reasons why it has to be this way: the harvesting operation can only be done a couple times on any given woman, and the embryo has to be well along in its development process before it can be implanted in the woman's uterus to result in a pregnancy. So it is inevitable that if this process is successful that it will leave unneeded embryos behind. That's where the stem cells they've been using have been coming from; they've been donated by the couples using the fertility clinics.

Except that now a clinic has deliberately created embryos for purposes of research. There was never any attempt to turn any of them into real babies; all of the fertilized eggs were converted for research purposes and all participants knew that going in. I don't see that this is ethically any different, frankly, but I can see how others might be affected by it. I have to wonder to some extent whether the clinic in question went public about this now precisely to polarise the issue further.

On this issue there can be no "compassionate conservativism". Either President Bush is conservative (and will come down clearly against this kind of research) or compassionate (and will support it fully, which I think is the right answer). There is and can be no middle ground. It's time to make a decision and to accept the fact that someone is going to be angry. (discuss)


Stardate 20010711.0536 (On Screen): Polaroid is in deep trouble and is considering bankrupcy, including convincing its creditors to accept a partial writedown of its debt. As the old phrase goes, if you owe a thousand dollars to a bank and can't pay, you have a problem. But if you owe ten million dollars to a bank and can't pay, the bank has a problem. With Polaroid owing its creditors a third of a billion dollars, it's not surprising that they are "optimistic" of getting an agreement.

It won't make any difference in the long run. Polaroid is as obsolete as kerosene lamps. They've been blindsided by the invention of the CCD-based camera and there is nothing important that Polaroid cameras can do which digital cameras won't ultimately be able to do better. The battle isn't over yet but Polaroid is in full-scale retreat. For sixty years it's built its business around a technological edge, and now that's gone like evaporated dry ice. They didn't see the change coming and didn't begin to invest in digital technology soon enough, and in any case in the digital camera business they have no advantage. I think the company is doomed. Those who live by the niche will die with the niche. (discuss)


Stardate 20010710.1257 (On Screen): In the 1970's and 1980's there was a period when there was a flurry of really unpopular court decisions and a lot of people responded by saying "Well, then, let's pass a constitutional amendment to overrule that." Two which spring to mind were a suggested constitutional amendment to ban school busing to assist integration, and a constitutional amendment to ban burning of the American Flag. I had hoped we were past that kind of foolishness, but apparently we're not. Fortunately, the Founders made the amendment process a difficult gauntlet to pass, which is as it should be.

The US Constitution has been amended 27 times, so the process works. In some cases amendments have sailed through the system: the 26th amendment (18 year old vote) was approved by Congress in January of 1971 and was ratified in June of that same year. But the process requires approval by both houses of Congress and by the legislatures of three quarters of the states (meaning 38 of 50) and this means that the need and worth of the amendment has to be very broadly acknowledged to have a chance. The nearest an amendment has come without passing in recent past was the ERA which got 36 legislatures and then stalled and eventually timed out.

The flag-burning amendment was a case of political grandstanding by the right, a sort of holier-than-thou attempt to embarrass liberals by daring them to vote against the flag. It was blatantly transparent and a gross misuse of the amendment process, and it did as intended put the more rational members of Congress in a bind. Then Bob Kerry of Nebraska stood in front of the Senate on his artificial leg, which replaced the real one he lost while winning the Medal of Honor in VietNam, and denounced the amendment for what it was. His patriotism was beyond reproach (they do not give Medals of Honor out easily) and that then gave political coverage to the majority in the Senate who knew they should vote against this. I was thoroughly disgusted by the whole episode. (And the ironic thing about the whole issue is that according to US law, old flags are supposed to be disposed of by burning. You're not supposed to just

Captured by MemoWeb from http://denbeste.nu/entries/archive-07082001-07142001.shtml on 9/16/2004