Dr. Jerry Pournelle

Email Me

Why not subscribe now?

Chaos Manor Subscribe Now

Useful Link(s)...


Hosting by

Powered by Apache

Computing At Chaos Manor:
October 19, 2010

The User's Column, October 2010
Column 362
Jerry Pournelle jerryp@jerrypournelle.com
Copyright 2010 Jerry E. Pournelle, Ph.D.

The other night I was the principal speaker at a dinner in the old Northrop Grumman building down by the Hawthorne airport. Hawthorne airport is directly south of LAX, one of the surviving general aviation airports in Southern California. I used to go to meetings down there when I was a studies manager for Aerospace Corporation in San Bernardino back when I was in the aerospace systems analysis business back in the 1960's.

I always thought of myself as doing Operations Research - we called ourselves OR men and there used to be advanced degree in OR at Case and other technical universities, and I was made a Fellow of the American Association for the Advancement of Science because I was a Fellow of the Operations Research Society of America (ORSA) - but someone decided that OR wasn't distinguished enough and all the OR troops became specialists in Systems Analysis, and our job titles were changed. Eventually ORSA merged with The Institute for Management Science to become something else, but by that time I was writing full time, and hadn't been to an ORSA conference in a decade, and for that matter hadn't paid ORSA dues in years. I still try to get to annual AAAS meetings sometimes.

Jerry with Iron Man
Jerry with Iron Man
Stephanie Bednarek
Stephanie Bednarek
Stephanie giving the tour
Stephanie giving the tour

The old Northrop Grumman building is now the headquarters and principal factory of Space-X, a private rocket company started by PayPal founder Elon Musk . Space-X is working full time three shifts a day to build new rockets to supply the International Space Station, and doing very well at it. It was also used as one of the sets for filming Iron Man, so I was able to get Stephanie Bednarek to take the first picture at right.

Stephanie is an aerospace engineer for Space-X, and the fact that they used her rather than a PR specialist to show the visitors around says a lot about Space-X. It seemed only fair to include her picture as well. I note she got me in better focus with my camera than I managed with her. In the final picture of the sequence, she is giving a tour to a number of guests including a congressman.

The dinner was held on banquet tables set up in the employees' cafeteria at Space-X, and Iron Man and his nemesis watched us at dinner. Food was excellent: the dinner was catered by Wolfgang Puck and his wife who were also guests. My speech went well, and I had a great time, but that's not really the reason for including the event in the column.

First, I hadn't been in the Northrop Grumman building since it was used as a production line for fighters, and I haven't been in a production rocket factory for nearly a decade. The changes are astonishing. Everything is more spread out. There's more room - and it's quiet, at least on swing shift. There was plenty of work going on, but most of it was controlled by computer programs. No riveting, which is why it was quiet. Some welding, but mostly the work consisted of installing electronic units and stringing cables. Not like the factory floor of, say, Boeing in the days of B-52 production, or for that matter Grumman when that Long Island company was the Navy's favorite factory as Boeing was the USAF Strategic Air Command favorite. In those days the first thing you saw on the factory floor was enormous crews, mostly riveters, and electronics got installed after the ship looked like an airplane or a missile, not as an integral part of the construction. All this is a consequence of the computer revolution which reaches into everything.

Okay, no surprises there, although it's one thing to think about when we contemplate the unemployment situation. Factories like Space-X turn out products. They produce highly valuable stuff; but they don't employ very many people. There were about as many aerospace construction workers and engineers employed in the Bomber Weapons Unit at Boeing in Seattle back in B-52 days as the entire payroll of Space-X in all its locations in California, Texas, and Florida, and Washington, DC. The computer has made everyone from the design engineer to the factory assembler far more productive while also making the final product more valuable. As I said, no real surprise, but it never hurts to be reminded of this.


My real surprise came before I ever got to the Space X factory. I live in Studio City, and getting to Hawthorne anywhere near rush hour is a big pain. It's one reason I seldom fly out of LAX, which means that when I go to Washington I generally stop at Dallas because there aren't direct flights from Burbank to DC. I'd rather spend two hours in the Dallas airport than on the road getting to LAX - Hawthorne is south of LAX and even harder to get to. Fortunately one of those attending the dinner was Edwin Sahakian, who flies a Cessna Citation Mustang, a small four-seater two-engine jet, out of Burbank airport, and of course he intended to fly, so it was arranged for me to meet him at Atlantic Airways on the Burbank airstrip for a ride.

My chariot awaits
My chariot awaits
The glass cockpit
The glass cockpit

I got to the airport, and Edwin told me to climb aboard and take the right hand seat while he attended to some last minute detail with the ground attendant. You can see my chariot at right, the picture taken as Edwin does paper work. There's still a lot of paper work in modern general aviation. I got into the airplane, sat down, and looked around, and the more I looked the more astonished I became.

I used to be familiar with aircraft instrumentation, which for a multi engine airplane like a B-52 or KC-135 can get very complicated, what with tachometers, oil pressure and temperature gauges for each engine, and a myriad other dials which don't demand attention unless one of them gets out of normal range. Aircraft control panels were egregiously complicated, too much so, and layout designers spent a lot of time laying things out so that pilots would be able to pay attention to instruments they needed for ordinary flying - altimeter, rate of climb, attitude indicator - without being distracted by those that only become important when something is wrong. I had anticipated that a lot of the secondary instruments would be automated, some of them perhaps being replaced by warning lights. Having seen the navigation screens in the most modern cars, I was pretty sure that there would be even more fancy 3D navigation displays in new airplanes.

What I wasn't prepared for was no instruments at all on the main control panel. None. The panel display is a computer screen. Actually multiple computer screens, one for each seat with a central navigation and general purpose display between them; but unless the airplane's central computer's are working, there's not an instrument in your normal vision field. It's all computer display. At first look some of the front and center instruments look traditional, but they're all computer display simulations. Everything. Temperatures, pressures, fuel levels and other information you don't need continuously, but also the altitude and attitude indicators. My first thought was that display in the airplane was computer generated.

That turned out to be incorrect. There is a mechanical attitude ball up just below the autopilot controls; which is a relief, because without that, if you lose power you'll be lost for sure. There are also mechanical air speed and altitude instruments, a bit changed from the ones I was used to, but they were there. The mechanical backups are up where they won't get in the way and are seldom seen, because you don't use them. Actually, you don't use much of anything. Flying nowadays consists of punching in numbers to the flight computer and getting out of the way. But I am getting ahead of the story.

The old mechanical instruments are up out of the way rather than right in the center of the display because they aren't really needed. It's easier to see and use the computerized readouts for attitude and altitude. That's probably just as well. The old three-pointer barometric altimeter was notoriously easy to misread, and most aircraft designers provided alternate (and redundant) instruments just in case.

It was even more startling to realize I was looking at a computer display rather than the mechanical attitude indicator. That flight ball used to be the main thing pilots watched when visibility got bad, and one of the first things you learned in flight school was always to trust the ball, not the seat of your pants. When you can't see the horizon out the window it's important to know whether you are straight and level, or banked, or worse banked and headed down. Everything can feel right, but if the ball says you're not straight and level, it's important to get that way.

Failure to believe the instruments was probably what killed John John Kennedy; or perhaps he misread it. There used to be a long standing debate over whether the attitude indicator ought to give an outside in or inside out view. Outside in: a small airplane pictured with a fixed horizon. If the airplane picture looked tilted or to be diving, that is what your airplane is doing. That view seems intuitive - but it's not what pilots are taught to use. They use inside out: the wings of that airplane are stationary (painted on the case), and the horizon tilts so that if you could actually see the horizon you would see what the artificial horizon indicates. If that doesn't make sense, look here and you'll get the idea. It usually takes beginners a bit of time to get used to that view.

The old flight ball is mechanical, with a gyro spun by vacuum; it doesn't require electric power except for illumination and you can always shine a flashlight on the ball if you lose electric power. The old altimeters are barometers, again mechanical, and don't need electric power, again except for illumination. You can read them in the light of a flashlight. I am assured that today's electronic flight computers and displays are so robust that there's no real need of the mechanicals, but it's a relief to know they are there as backups.

I thought about all that as Edwin took about ten minutes getting our flight plan into the Garmin computer system.

After which we got flight clearance, and the airplane pretty well flew itself. It rolled down the taxiway, turning at the right spots, went to the end of the runway, and took off. It flew in a big arc that took it to Cucamonga, turned, and next thing I knew we were descending. I knew that because I had kept my eyes on the electronic equivalent of the old flight ball when I wasn't looking out the window. "Why are we descending?"

Edwin grinned. "We're there." I had sort of forgotten that this was a jet; nearly all of my flying in small airplanes was in little single engine high wing airplanes, and it takes them a lot more than seven minutes flying time to get from Burbank to Pomona and thence to Hawthorne. Edwin got the tower, told the airplane what runway to use (actually there is only one at Hawthorne) after which the plane landed itself. He did steer it to where we'd park it during dinner.

Getting back was a bit more complicated: we were taking off into the LAX landing pattern. We'd come in from the inland side flying into the breeze coming off the sea. To take off headed inland would require getting aloft with a tail wind, and there was enough sea breeze to make that less than fun; we'd have to go out over the water, and we were not far south of LAX with all its traffic. It took about twenty minutes to get air traffic control to find us a hole in the LAX traffic and we only had a five minute window to get to the end of the runway and get aloft. That turned out to be no problem. There were clouds over the sea at 1000 feet, and the world disappeared as it does when you fly through clouds. The airplane always knew where it was - clouds don't stop GPS signals - and although I couldn't see a thing out the windows as we flew through one of the busiest traffic patterns in the world, there were no problems at all, and presently I could see Santa Monica off the starboard bow. Six minutes later we were on the ground in Burbank. On the way in I was able to realize just how complicated flying in the Los Angeles Basin has become compared to the 1960's when I'd sometimes get an Air Force Johnny to fly me in to Santa Monica airport for a RAND consultation, or out to Edwards in the Mojave Desert. In those days it was easy to tell where you were, even at night. No more. I was lost. Fortunately the airplane knew exactly where we were, and had no problem finding the proper runway - there is more than one at Burbank, of course - and getting us on the proper descent path.

For some of you this is no news at all, of course. I gather there is a whole generation of pilots who never learned to "fly by the seat of your pants", and who never suspected that you could navigate by following highways and railroad tracks in airplanes without much in the way of instrumentation, like the ancient Piper Cub I got my flying lessons on in Memphis, Tennessee in 1949. There have been more changes in general aviation since 1990 than there were in the fifty years before that. GPS isn't just something you add to your navigation skills: it pretty well takes their place, and allows a precision of air traffic control that was never possible in those days.


It's dangerous out there. A reader sent me a copy of a phishing scheme he'd received, purportedly from me, discussing his recent email to me. Since he hadn't given his email address in that letter, he thought it might really have been from me, except that the grammar was atrocious so he thought he'd ask.

I did a Google site search of my web site on his name, and discovered that he had some years ago sent me email with his name and address on it hoping that someone would answer a question for him, and I'd published that. The email address appeared with his name only once, and that years ago. He hadn't done that again, but once was enough: someone was able to harvest his name and address, and the fact that a person with the same name had recently had mail published on my site, and that was enough to generate this email message. It didn't want him to do anything but answer some harmless query about his message; but that would have been enough to verify that this was a valid name and address, and I can only guess what the next step would be. Note that someone went to a fair amount of trouble to do this - and also that whoever it was hadn't put similar effort into learning English grammar.

I do not publish people's email addresses without permission, because we know that spammers harvest email addresses from lively web sites. This was worse, since it purported to be an inquiry from me and faked my return address.

I was pretty sure I knew how this had been done, but to be safe I did what I always do when there's any possibility my system has been compromised: I Googled ESET, clicked on the online scanner, and let it do its thing. For good measure I did that with all my other DOS systems as well. Interestingly, ESET found something on Emily, my superfast Intel system, but it found it in an archived file that hasn't been open for years. The system has been scanned many times before, so this must be due to a new revision of the scanning software. I let ESET quarantine the file anyway.

Norton also offers an on-line scan service, and I sometimes use that just for variety. Periodic on-line scanning is a precaution, a checkup to be sure my main line of defense, Microsoft Security Essentials including automatic updates, hasn't failed.

I recently heard a caller to Leo Laporte's radio show ask how she could turn off automatic update, because it took so long to do and slowed her system's Internet access down. Turns out she was on dial-up. Windows updates can be many tens of megabytes - hundreds, even - and take a long time in dialup. Leo's suggestion was that she periodically take her laptop to a Starbuck's and make use of the wireless connection to update the system, and be sure to do that on the second Tuesday of each month, which is Microsoft Windows Update Day. My wife commented that you don't need to pay Starbuck's prices; many McDonalds now have wireless connections, and their coffee is just as good.

Note also that Microsoft sometimes makes unannounced security updates available whether it's second Tuesday or not.

It's vitally important to keep your system up to date, and do that as fast as you can. When Microsoft releases a security update, they acknowledge that there was a system vulnerability to be fixed. That's a signal to every highly skilled hacker to get to work, because a lot of people will take a while before they update their systems, and thus leave a window of opportunity to the bad guys. Nowadays the hackers don't work for glory, or reputation among other hackers, and they avoid fame: what they want are systems they can command, and the usual worm does no more than insert a tiny vulnerability, protect your system from further infections (which the worm sees as rivals), then go to sleep for a while. Eventually it calls home for more instructions. By that time you may have updated your system and be no longer vulnerable to that worm, but it's too late. That's why I periodically scan my systems with an outside well maintained scanner like ESET or Norton.

Virus Protection

I am often asked if regular users need ESET NOD32 or Norton Anti-Virus; and of course you hear them advertised often on the radio, and see their advertisements in the few computer magazines that still survive.

My answer is that many think it's better to be safe than sorry, and either of those programs will work. I used to swear by anti-virus programs, beginning with Dr. Solomon's. I then switched to Norton Anti-Virus and used that for many years, until, sometime after the turn of the Millennium, Norton System Security got so loaded with "features" and just general gunk that it was more trouble than it was worth and I dumped it. I replaced it with NOD32 and that nagged me so much that I dumped it in favor of the Microsoft replacement. Microsoft has converted to Microsoft Security Essentials which you can get free, and I have that on all my Windows systems, and so far that has been good enough.

I don't want to be in a position of advising people not to use an anti-virus program, because if I do, sure as anything, someone will get a virus and blame me. I don't use them myself, but I am the first to admit that may be a mistake. Our hardware is far more powerful than it was in the days when the cost of an anti-virus program was a noticeable slowdown in system efficiency. I will say that if I had my choice of an anti-virus program vs. keeping my operating system up to date, I'd choose the updates.

But when you come down to it, your best protection is still common sense. I don't know of any software that can protect you from cooperating with the enemy. If you're determined to fall for a phishing scheme, you may be lucky enough to be saved from answering the wrong questions, but you're sure taking a chance. Be careful about opening attachments, even from people you trust. Preview your mail in plain text. And be aware that the electronic "Hallmark" card someone has sent you is as likely to lead you to the Ukraine as to Kansas City. It's a pity because genuine Hallmark eCards can be beautiful, but alas, most of the "Hallmark" eCards I get are phishing schemes. If you really know what you're doing it's possible to weed the genuine from the phishing eCards, but there are a lot of smart guys out there coming up with new ways to fool you; and once you open up something as complex as an animated greeting card with music, common sense says that a lot of other stuff can slip in along with that. If you're in the habit of reading eCards, by all means get a good anti-virus program. And still be careful.

The Largest Non-Nuclear Explosion Seen from Space And Other Stories

Viruses, worms, and Trojan horse software packages are nothing new, and some have had rather important effects. For example we have stories of the infection of Iran's computers which control their nuclear program being widely infected by Stuxnet, with what long term effects we don't know. Since most Iranian software is pirated and thus not maintained and updated, it's possible the effects will be quite profound, not only on the nuclear program, but on Iranian commerce in general. I expect to find out more about this in the near future.

Another story of deliberate software infection with large effects has been circulating for years. William Safire told it here. I knew Safire, but not well: we were both paid speakers at several conferences, one as I recall at Sandia. He had a reputation for reliability - and I have other sources for this story. The USSR was notorious for its industrial espionage. This was in the early 1980's, when software was more mysterious than now, and the Soviets were trying to restrict access to computers, so that their smartest programmers often didn't have frequent access to the machines they programmed. This generated a desperate need for software for industrial processes, including control of natural gas pipelines. The Soviets couldn't buy the software they needed because the US blocked the sales, so they stole it; only the US was waiting for that and poisoned the well. The result was that they lost control of the pipeline and the result was the single largest non-nuclear explosion in history. It was so bright that USAF sources at the time thought this might be an attempt to blind our observation satellites.

Most viruses and Trojans won't have that dramatic an effect...

Book Review Section

Lego Technic Idea Book(s) &mdash No Starch Press has a series of idea books on building things from Lego. The series has the title The Lego Technic Idea Book, and there are three of them, all by Yoshihito Isogawa: Simple Machines, Wheeled Wonders and Fantastic Contraptions.

The titles pretty well explain what the books are about. There's no text in these books: they consist of hundreds of pictures of Lego devices, some fairly conplex (even in Simple Machines) and many rather elegant. There are a few wheeled wonders in the Simple Machines volume and a couple of the Wheeled Wonders could easily have been in the third book. Some of the Fantastic Contraptions truly live up to that name.

I enjoyed browsing through them, and if I were a Lego addict I'd appreciate getting them as a gift.

Best iPhone Apps: The Guide for Discriminating Downloaders by J. D. Biersdorfer is titled by O'Reilly as a second edition, but there's little overlap between this one and the first edition by Josh Clark. Biersdorfer is the author of the iPad Missing Manual, and according to her official biographical sketch had to stand in line for six hours last June in order to get her iPhone 4, which says something about Apple PR if nothing else. There are over a quarter of a million iPhone Apps and the number grows every month. There were "only" 50,000 when the first edition came out. The second edition has mostly new applications to recommend in a variety of fields, from games to getting work done to lists to notepads and editors. I don't completely agree with all her selections, and her taste in games is not mine, but then I don't play many iPhone games. Even so, I'd be astonished if you read this book and didn't learn about an App you need that you never heard of.

Building Android Apps with HTML, CSS, and JavaScript by Jonathan Stark is O'Reilly's introductory book for JavaScript programmers to get into the Android Apps game. Once you have written and proved out an App for Android you can publish it yourself - unlike Apple there's no need to get approval or go through the company store. If you know HTML, CSS, and JavaScript you can write Android apps; there's no need to learn Java. I had no problem following the logic of the examples given, and there are a number of examples. I don't have an Android so I didn't write a simple one for testing, but I don't doubt I could have done that.

Leo Laporte and others have chosen the Android as the pocket computer/phone of the future. I'm staying with the iPhone (and iPad), but I haven't got a strong rational argument for doing that; those who have gone over to Android seem happy with their decisions. If you're somewhat familiar with CSS and JavaScript this book will guide you to writing your first App and even tell you how to publish it. After that it's up to you.

Adaptive Project Framework: Managing Complexity in the Face of Uncertainty (Kindle edition) by Robert K. Wysocki is another book in the series of books on Effective Project Management by that author. I've never met anyone who read one of them, but they are apparently popular, and the Agile management system has built itself a sizable following. One of the problems with being an old Operations Research man is that OR consists largely of studying ways to take tasks and operations apart into smaller pieces that can be modeled. You get in the habit of thinking that way, so when you encounter a management techniques textbook a good bit of what's in it seems obvious. That doesn't mean that OR people all make good managers. For that matter, most of those who graduate with MBA's don't turn out to be as good at management as they'd hoped to be. Art is art and skills are skills.

Like many management books, Adaptive Project Framework has many broad statements that turn out to be hard to pin down. For example:

We have seen how APF not only anticipates adaptations but also expects them. We have already discussed that APF is not a recipe to be blindly followed. Rather, APF offers a structured framework - a strategy - for thinking about how best to manage a project. However, APF is far more adaptable than even the situations in the preceding chapters have indicated...

The previous chapters showed a series of flow charts and check lists to use so that you might anticipate adaptations; I presume that is what is meant by "APF expects" these changes. One might also call that common sense.

This book expects you to already be familiar with Agile management techniques, although it never explicitly says that it does. If I were just getting started in learning formal management, I'd start with an earlier book in the series. Agile is as much a way of life as a technique.

The Process of Software Architecture by Peter Eeles and Peter Cripps, Foreword by Grady Booch, is a general purpose introduction to program design. It doesn't look at any single procedure or computer language, but rather, as per the title, the process of architectural design. It's in the tradition of Wirth's Algorithms + Data Structures = Programs, and thus very worthwhile for budding computer scientists. It describes procedures, but more importantly, explains why the procedure is needed and what it means to accomplish. Every would be software architect should read several books of this type, since they raise the kinds of questions that ought to be addressed in the design of major software, but which, alas, are often overlooked. This one is readable, and has examples and case studies.

Head First Data Analysis by Michael Milton is an interesting book. It's an introduction to a lot of techniques that are traditionally part of the tool kit for Operations Research, and indeed one of the tips given in the book is that if you need to learn more about quantitative analysis do an Internet search on operation research. Like most books that contain an introduction to statistical analysis, it gives too little attention to the philosophical stuffing under statistical inference, but otherwise it does well, and there's a better coverage of experimental design than most such books have.

The coverage is quite good, and there are lots of examples; this is one of the better introductions to data analysis and Operations Research that I have seen. Of course it presents it in the "Head First" format, which is breezy and informal, but that's what the Head First series does; you either like that or you don't. I didn't, at first, but I have come to appreciate it. It starts simply and goes in easy steps to quite sophisticated OR techniques.

Most people don't think in OR terms, and that's a pity. I wish public school teachers and administrators did. If this book were required reading for teachers I suspect that public education would greatly benefit from it. Recommended.

Winding Down

The game of the month was We Rule, which is an app for the iPad. I don't quite understand why the silly game is addicting, because nothing really happens in it. By rights I'd simply give it up as a waste of time, but I find myself checking my kingdom several times a day just to be sure no one has placed an order, or delivered something I have ordered. I tend to time my visits for just after my magic cauliflower has ripened, and yes, I know how silly that sounds. Oh. Well.

I continue to rely on the iPad as the primary device for reading books, usually with the Kindle App. I'm told that some general aviation pilots have bought iPads just to keep all the airport communications channel information in. Easier to find, easier to update, and a lot less stuff to carry.

The computer book of the month is Facebook The Missing Manual by E. A Vander Veer is the book of the month. I have never put up a Facebook page, although one of my fans seems to have created one with my name on it; I've never looked at it. But I have become convinced that I need to get into the Facebook game, and this book is quite clear on how to do it.

There has been a storm of controversy over Facebook and its privacy invasions, web tracking, release of information put on the Facebook site, and such like; there was another blob of news on this today as I began to write this. My view of Facebook is pretty well the same as my view of all of the Cloud - if you put the information out there, it will be found, and it will be public. Storing secrets in the Cloud is in my judgment a contradiction in terms. I never expected anything put on Facebook not to be public no matter how restricted I might want it to be - one reason I have up to now stayed off of Facebook.

I'm about to construct my own Facebook page. It won't have a lot on it, but it will be a convenient place to announce new books and call attention to new projects. I expect to use this book for that construction: I often rely on The Missing Manual series, and from my reading of this volume I see no reason to change my mind. Emily Vander Veer has written a dozen readable books, and this one is no exception. Recommended.

The movie we saw this month was Eat, Pray, Love with Julia Roberts. She's still charming and interesting, and it makes for a romantic chic flick, but the story doesn't stand up to much analysis, or at least it didn't for me. I don't really accept the premise that you have to go to foreign lands to find religion, and in her case she went to a foreign place - Rome - which already has a religion along with running water and good plumbing, as opposed to the Far Eastern places she went to get spiritual at the instructions of gurus and shamans. I wasn't sure at the end of the story that she was in a much different situation than when she started her journey. Perhaps I missed something.

The other night we went to the Los Angeles Opera. The presentation was Il Postino, a relatively new opera by Daniel Catán, an American educated composer born in Mexico and now the music administrator of Mexico City's Palace of Fine Arts. The libretto is essentially the script of the 1994 movie, set to lyric music quite unlike the stuff that usually passes for modern opera.

I can recommend Il Postino as an artistic success, but you do need to remember that it takes place in 1948 when Stalin was firmly in control of the Soviet Union. The opera is supposedly about a period of exile of the Chilean communist poet Pablo Neruda (Neftalí Ricardo Reyes Basoalto) who fled a Chilean arrest warrant in 1948 and spent some time in "exile" in Italy (although not generally in seclusion and certainly not on an island without running water, although he did visit Capri). If you want to know more about Italy in those times, a good source is a series of short vignettes collected into several books by Giovanni Gaureschi that begin with "The Little World of Don Camillo"; alas the books are long out of print. Those were the days when Stalin took over Czechoslovakia (Spring 1948) by the murder of Jan Masaryk and the Iron Curtain was descending across Europe (Churchill's Iron Curtain speech was made in 1946).

Il Postino assumes that the Communists were the good guys, and that the Christian Democrats who won the election in Italy were the oppressors. This is not the standard view of history other than in the minds of a number of Western intellectuals. Or perhaps it assumes that Neruda was not really aware of the Gulag and other Stalinist horrors (much like Peppone in Guareschi's Little World, but Neruda wasn't a small town mayor who had never been out of the Po Valley except for a stint as an anti-German partisan in 1944; he was a well educated and well travelled public intellectual).

Having said that, the "Pablo Neruda" of Il Postino is a thoroughly likable, indeed charming, person, and all the reds in the opera are clearly on the right side against oppression; in other words, the plot is fairly standard Popular Front viewpoint. There is no enemy to the Left. Such propaganda is so deep seated among Western intellectuals that many believe it, and others will dismiss any criticism of the view with a shrug and a remark about artistic license. I can sympathize with that. I enjoyed the movie, and I thoroughly enjoyed the opera, and I was very pleasantly surprised at the non-modern lyrical music, which Roberta terms "rich".

The part of Pablo Neruda is sung by LA Opera General Director Placido Domingo who is still in great voice and has no problem playing a 50 year old reasonably active poet. He's marvelous. Mario the Postman is played by Charles Castronovo, who's popular in Los Angeles: he started in the LA company and learned the craft in minor parts here before breaking out to be one of the leading lyric tenors of modern times. He demonstrated why he is in demand as a lyric tenor; he also did a great job of playing a country bumpkin who has never left his tiny island. Amanda Squitieri is in great voice and was utterly believable as a small town Catholic girl of 1948 just coming of age, and indeed all of the cast were believable. It's a great story, and well worth seeing.

The opera is worth mentioning in a computer column because there was a great use of high tech in the production. Back in the Dark Ages when I was Managing Director of the Seattle Civic Playhouse, the lighting control board had enormus high-resistor dimmers and a myriad of switches, and all the lighting moves had to be done by hand from a very precise lighting script. An opera in a foreign language — as for Americans most are — couldn't really be presented with simultaneous translations without distracting everyone watching. Set changes were done mostly by grips under the watchful eye of a stage manager. There certainly were no back screen projections on scrims, and there were no movie screens simulating the TV in a bar or showing back story details. All this and more was a routine part of Il Postino, and it was all done smoothly with a minimum of distraction. One problem with proscenium theater is keeping up the illusion of reality, and whether or not this is desirable; Thornton Wilder wrote extensively about this, and how he didn't believe for a moment when watching most plays. Bertolt Brecht deliberately put stage hands and spotlight operators on stage in full view of the audience to make sure there was no possible illusion of reality. Movies don't have this problem: they can be real, or they can deliberately expose the machinery behind the works, and the director gets to choose which approach to take. Opera and stage play directors often don't have the choice — but as our computer technology gets better, and more importantly cheaper, live theater including opera is more and more getting that choice.

A couple of years ago I wrote that we were rapidly approaching a time of computing plenty, when everyone would have on his desk more computing power than could be used. That time is coming on fast. It won't be long before we have 64 processors on a single chip.