Saturday, 25 May, 2002

Book Review: Bias

People have been pointing out a decided liberal bias in the media for years.  Especially television news.  And for years, news anchors, liberals, and others have dismissed those complaints as radical right-wing conservative rhetoric.  Admittedly, Rush Limbaugh screaming "liberal bias" has a certain hollow ring to it, but when a self-described "old-fashioned liberal" like Bernard Goldberg documents it, somebody's got to take notice.  Bernard Goldberg is a 30-year veteran of CBS News, an Emmy Award winner, and well respected in the television news business.  In his book, Bias, he gives very strong evidence showing how your evening television news is slanted towards the liberal viewpoint.

I picked up the book out of curiosity, and approached it with some skepticism.  Something in me automatically distrusts "whistle blowers," perhaps because so many of them are blowing the whistle just because they have some bone to pick with their employers.  The book is very readable, and Goldberg's points very well supported by the evidence.  The primary point is something that I hadn't considered.  Goldberg points out that the liberal bias in the media isn't some master plan conceived by television journalists or politicians who are hell-bent on furthering their agendas.  Rather, the bias is a natural product of the media types' personalities.  Unsurprisingly, the vast majority of journalists have liberal leanings.  For example in a 1996 survey of 139 Washington bureau chiefs, "89 percent said that they voted for Bill Clinton, compared with just 43 percent of the nonjournalist voters."  As Goldberg points out: "There's hardly a candidate in the entire United States of America who carries his or her district with 89 percent of the vote."  61 percent characterized themselves as "liberal" or "moderate to liberal."  9 percent said they were "conservative" or "moderate to conservative."  These are but two examples from the many that he presents in the book.

All that would be okay, provided that journalists recognized their own biases and made a concerted effort to present real balanced reports.  Sadly, they don't.  It's the dishonesty that's the real crime.  At least I know, when I listen to Rush Limbaugh and his ilk, that I'm going to get a one-sided view of the world.  For journalists (especially television journalists) to pass their lopsided views off as balanced reporting is just short of criminal.

Goldberg's book is an easy read, and quite worth the time, regardless of your political leanings.  Highly recommended.

I almost drove off the road this evening.  I was listening to the BBC program The World on our local public radio station (KUT, 90.5 FM).  The news announcer was reporting onthis story.  I burst out laughing when the announcer said "Researchers in Africa have discovered that preventive measures are less costly [in reducing the spread of HIV] than treating infected patients."  I actually had to pull over for a minute, I was laughing so hard.  They reported that as news?

Whoever summarized the report for the announcer got the sense of the story all wrong.  What the researchers did is quantify the cost.  It turns out that prevention is 28 times more cost effective than treatment.  As the article puts it:  "In other words, for every life-year purchased with treatment drugs, 28 life-years could have been purchased with prevention."

The more interesting information is in this paragraph:

...because of misconceptions about HAART's efficacy in preventing HIV transmission, both the US and Brazil have seen an upswing in sexual risk behaviors and an increase in HIV incidence after the introduction of HAART.

So, prevention is 28 times as cost-effective as treatment, and more effective in preventing transmission.  It seems to me that we should be concentrating most of our available resources on prevention.

I've been wondering recently about time.  Not hours and minutes, or millennia, but rather simple calendar time.  Today is May 22, 2002, by the Gregorian Calendar that's been adopted by most of the world—at least for purposes of navigation and scientific reference.  There are countless other calendars, a brief study of which is very revealing.  A good place to start is http://www.ernie.cummings.net/calendar.htm.  As good as the discussion is, though, it doesn't answer my question.

How do I know that today is May 22, 2002?  I think we have reasonably good records for any date after 1582 (when Pope Gregory changed the calendar), and perhaps for a century prior to that.  But prior to, say, the middle of the 15th century, written records are somewhat spotty.  With all of the calendar machinations between the Julian calendar's adoption in 46 B.C. and today, can anybody say with any certainly what today's date is?  Can we say with any authority exactly how many days have passed since that date?  I somehow doubt it.  I think we'd be hard pressed to say, with an accuracy of plus or minus 10 years, how many actual years have passed since the Julian calendar was adopted.  The AD/BC system wasn't adopted until 500 or so A.D., and at that time there was some question as to the exact year of Jesus's birth.

I'd like to know how far back we can reliably count days.  The only way I can think of is to find dated documents that were written on each and every day, going back as far as we can.  And the dates on those documents would have to be corroborated by references from other documents, verifiable content (i.e. today was the great fire), or perhaps the reputation of the document's author.  I would think we could get back 100 years in a single newspaper's archives.  Could we go 200 years?  Could we get back to 1582?  How far back would we have to go before we couldn't say for sure what day it was?  Why this is important, beyond rubbing a rad fundie's nose in it from time to time, I can't really say.  But it sure would be an interesting archaeological exercise.

Tuesday, 21 May, 2002

Stallman Unfairly Maligned

Richard Stallman, founder of the GNU Project and the Free Software Foundation, is in the news again.  But this time he's being unfairly maligned.  Apparently, RMS (as he likes to be called) was invited to speak at the University of Texas Linux user's group, whose name is SIGLINUX.  RMS declined, saying that Linux should be called GNU/Linux, and he would be happy to speak if the group changed its name to something that acknowledges the role of the GNU project or the Free Software Foundation in creating the operating system that is most commonly called Linux.  The details are in this article (linuxworld.com) by Joe Barr .  RMS recently wrote an article (linuxworld.com) responding to Barr's criticisms.

So perhaps you think that RMS deserves the criticism he's getting from the open source crowd in this situation?  If so, think again.  RMS was invited to speak at a user group—something he is happy to do in most cases.  However, he has no obligation to speak to any group, and he can decline for whatever reason, or no reason at all.  He is not holding the group "hostage" or somehow harming them by electing not to speak.

Personally, I think RMS is being unrealistic, just as I believe that his dream of eliminating proprietary software is unrealistic.  But RMS is an idealist, and he's perfectly free to live in his own delusional world.  He doesn't deserve to be attacked for adhering to his principles.

Monday, 20 May, 2002

Star Wars:  Attack of the Clones

It seems every year about this time, my employer takes us all to see a movie.  This year's choice was, of course, Attack of the Clones, Episode II of the Star Wars saga.  Technically, the movie was excellent, with beautifully rendered scenes and lots of eye candy.  I could do without the computer graphics characters, though.  At least they didn't use CG for a primary character like they did with that annoying Jar Jar in The Phantom Menace.  I thought that the acting was mostly uninspired, and I still haven't figured out why they cast Samuel L. Jackson in that role as a member of the Jedi Council.  That role is a poor use of Jackson's considerable acting talent.  The action scenes are reasonably well done, but the movie drags a bit (okay, a lot) between action sequences.  This is not a science fiction action/adventure movie like the first three (and to a lesser extent, the fourth).  It's an historical epic disguised as a science fiction movie.  That's fine, I guess, but it's not my favorite type of movie.

I saw the movie more because I want to see how Lucas presents the story.  We all know what the story is—Obi Wan told it to Luke at the beginning of the first Star Wars movie over 25 years ago.  I'm interested now in how the final part of the story will be told.  I don't expect any surprises, though.

Sunday, 19 May, 2002

Garden Fence

Work on the garden proceeds slowly.  Last week I had time enough to put in the T-posts for the fence, and yesterday I attached the wire on three sides.  Today I spent setting the two wooden gate posts, and constructing the gate.  An interesting little engineering problem, that gate.  I have to wait for the concrete to cure (24 hours) before I can attach the gate and string the rest of the wire.

My real question is whether a 6 foot fence will prevent the deer from hopping in and mowing down the garden.  I've seen deer jump a 6 foot fence, but I've been assured that they won't jump into the enclosed garden because the raised beds break up the open space too much.  If the 6 foot fence isn't enough, we'll string a wire across the tops of the posts (another 8 inches up), and hang streamers from it.  The combination of the wire and the streamers apparently confuse the deer enough that they won't attempt it.  We'll see.

Saturday, 18 May, 2002

Spider Man!

Debra and I went to see the new Spider Man movie.  I was pleasantly surprised.  I was never a huge fan of the comic book, but I read it enough to know the basic story.  The movie was true to the comic, but yet done in such a way as to stand on its own.  The action sequences were superb, and the pacing of the movie was right on.  The sound, especially, was excellent.  They made very good use of the theatre's surround sound.  I have a few quarrels with how some parts played out, but for the most part I thought the ending was perfect.  And, of course, they left plenty of room for the sequel, which is due out in two years.  Considering how well the movie did on opening weekend, I'm confident that the sequel will indeed appear.  I wouldn't call it "best picture" material, but it's a rockin' good action flick with a decent story line.

Browsing in the book store last week, I picked up Donald Knuth's book Things a Computer Scientist Rarely Talks About.  The book contains the transcripts to six lectures he gave in 1999 for the "God and Computers" lecture series at MIT.  Knuth covers a lot of territory in his book, including randomization as a means of scientific inquiry, aesthetics and computers, language translation, and creativity.  He ties it all together with his work on his previous book 3:16 Bible Texts Illuminated.

The book is a pleasant read, humorous in places, and spattered with valuable insights into how a scientist thinks of God.  Perhaps the most powerful insight is this:  there is no practical difference between "infinite" and "really, really big."  Knuth "invented" a number he calls "Super K" (see pages 171-175 of the book), which is so huge as to be far beyond human comprehension.  The number is finite, but for all intents and purposes, infinite.  Is it possible that God could be limited by a number that large?  It is, after all, way more than the total number of subatomic particles in the universe.  If so, would that God appear any different to us than the traditional "infinite" God?  I'm going to be pondering this one for some time.

Thursday, 16 May, 2002

What's Wrong With Yes and No?

Have you noticed that more and more software is using "OK" and "Cancel" in dialog box prompts when "Yes" and "No" seem more appropriate?  The software installation I ran today presented me with a dialog box similar to this one when it was finished: 

The answer to that question should be "Yes" or "No".  I don't know who decided that "Ok" and "Cancel" would be easier for users to understand.  My understanding is that Apple strongly discourages Yes/No selections.  If somebody can give me a reasonable explanation as to why and how Ok/Cancel is somehow better, I'd be grateful.

Wednesday, 15 May, 2002

The Hardest Test of Maturity

From the Anonymous file comes this quote:  "The hardest test of maturity is knowing the difference between resisting temptation and missing a once-in-a-lifetime opportunity."  I can't decide if that's supposed to be humorous.  Certainly in my case, most things that look like once-in-a-lifetime opportunities really are temptations that are best resisted.  Or perhaps I haven't yet mastered that one.

Tuesday, 14 May, 2002

Walking Stick

I came home this evening to find awalking stick (click on the picture at left for a larger image) hanging out on the lattice work of the pool house.  The body on this guy (actually, probably a gal) is 4 to 5  inches long.  It looks longer because the two legs at the bottom left in this picture (just below the wide protrusion there) are stretched out rather than to the side like the other four.  I can't say for sure which end is the insect's head.  I haven't yet identified the particular species.  At first I thought it was an American Walking Stick, Anisomorpha bupestroides, but now I'm not so sure.  I haven't found a picture of anything like this on the 'net, yet.  Guess I'll have to look in my Texas Bug Book.

I've known about these things for years, of course, ever since I lived in South Texas, although I didn't realize they lived in the Austin area until we'd been here over a year.  I don't see them very often because they typically hang out in the trees (they like my oak and pecan trees).  I was particularly surprised to see this one on the lattice work.  Usually I'll find them stuck to a tree trunk, or walking on the ground among the real sticks.  What a fascinating insect.  Can you believe there are people who keep them as pets?  Very strange.

Monday, 13 May, 2002

Open Source Propaganda

A Slashdot posting yesterday discusses this article at Newsforge about a "panel of experts at the National Academy of Scientists" that wants to end the software industry's special exemption from product liability suits.  A "lemon law" for software.  The article tries to make the point that Open Source software developers would incur liability under such a law, in much the same way as would a commercial vendor.  Perhaps so, perhaps not, although considering that lemon law claims typically are limited to the price paid for the product, what's the big deal?  Open Source software is, most often, obtained at no cost.

My primary beef with the article, beyond its coming to no solid conclusion (or even attempting to), is the author's use of unsupportable blanket statements, most of which are mindless parroting of Open Source propaganda.  For example, the third paragraph contains this oft-repeated myth:

We all know that the open and distributed model for development described in Eric S. Raymond's book "The Cathedral and the Bazaar" is much better and creates more reliable products than any closed non-distributed development model.

I certainly don't know that!  There is some evidence, anecdotal at best, that the distributed "bazarr" method can create good software.  Understand though, as I've pointed out before, that most of the projects that the Open Source advocates hold up as examples (the Linux kernel, Apache, Samba) were developed using a traditional model of a small tightly controlled team.  It's only after the software was stable and substantially complete that it was released as Open Source.  There's nothing even close to proof that the "many eyes" method of debugging makes for superior software.  Conversely, there's a mountain of evidence to show that the "million monkeys" approach to development makes for clearly inferior software.  The one project they thought was going to prove their point conclusively,Mozilla, is a laughing stock even in the Open Source community.

I wish the Open Source crowd would come up with some objective proof of these claims.  Instead, whenever somebody calls them on it, the reply is "You always say 'unsubstantiated' and 'anecdotal'.  You're just spreading FUD (fear, uncertainty, and doubt)."  Sorry, guys.  You're the ones running around with chips on your shoulders, making wild claims that you expect us to believe despite all evidence to the contrary.  If you want to change my mind, you had damn well better show me conclusively that your way is better.

Sunday, 12 May, 2002

A Gentle Ribbing

If you're going to write a daily journal, be sure to write in it every day.  If you think of an idea for the journal, jot it down and write it up as soon as possible.  Otherwise you're sure to have forgotten it by the time you sit down to update your journal.

Not that I'm speaking from experience or anything.

Tuesday, 07 May, 2002

Stop the Registry Madness!

I thought that everybody, Microsoft included, admitted that the Windows Registry was a really bad idea that looked good on paper.  The number of things that can go wrong with a single file containing all system and program settings is astronomical, and 7 years or more of tweaking and patching hasn't made it any better.  The IIS team apparently agreed, because they decided to move most of their configuration to a separate file called the MetaBase.  Not that the MetaBase is such a good idea, either, being a weird indexed proprietary format, but at least it reduces the chance of losing your IIS settings when the Registry gets corrupted.

But, no, programs are still storing all manner of data in the Registry.  Today's evildoer?  Outlook Express.  Would you believe that Outlook Express stores its message rules (blocked senders and other filtering rules) in the Registry?  No wonder I couldn't find them by searching the hard drive.  You'd think that, since Outlook Express doesn't provide a way to export or import message rules, it'd at least put them into a file so that you could back them up.  But, no.  In order to back up your message rules you need to start the Registry Editor,  navigate to the proper key, and then select "Export Registry File" from the File menu.  The proper key, by the way, is:

HKEY_CURRENT_USER\Identities\{GUID}\Software\Microsoft\Outlook Express\5.0

The {GUID} part is replaced by the globally unique identifier of your identity.  Likely there will be only one to select from.

In there, you'll find keys for Blocked Senders, Rules, and Signatures.  Export what you want to a .reg file (which is text format), and you can edit to your heart's content using Notepad or your favorite text editor.  If you want to import the rules to another computer (or for another user), then simply log in to the machine where you want to import the rules, change the {GUID} portion in the .reg file to match the GUID for the identity on that machine (get it from the Registry Editor), and then select "Import Registry File" from the File menu.  Note that this may very well overwrite any existing rules, so be careful.  It's probably a good idea to back up your existing rules before you do the import.

My friend Jeff Duntemann has pointed me to PocoMail, which I need to try.  Not only are the message rules stored in plain ASCII text files, but it also looks like the program's message storage is much more reasonable.

I ran across Eric S. Raymond's How To Ask Questions The Smart Way today.  It's a nice piece of fluff, I guess, and even good advice for programmers or other technically-minded people who want to ask questions on open technical forums.  It's hardly designed to encourage the uninitiated to ask questions, though.  The document says a lot more about  the elitist, condescending attitude of most hackers than anything else.  It's a great advertisement for not hiring anybody who calls himself a hacker.  With attitudes like that, what employer in his right mind would hire them?  Perhaps an example will serve to illustrate my point:

We've found by experience that people who are careless and sloppy writers are usually also careless and sloppy at thinking and coding (often enough to bet on, anyway).  Answering questions for careless and sloppy thinkers is not rewarding; we'd rather spend out time elsewhere.

Elsewhere being, I suppose, hacking away at some code.  You know as well as I do that most hackers are careless and sloppy writers.  Does it follow, then, that they also are careless and sloppy at thinking and coding?  I believe that to be the case, and the state of most open source software I've seen certainly supports that assumption.

Sunday, 05 May, 2002

Building the Garden

Seven cubic yards of gardening soil makes a pile about 8 feet long, 5 feet wide, and 4 feet tall when it's delivered by a dump truck.  It takes about 1.2 cubic yards of soil to fill an 8' x 4' x 1' garden bed.  One bed takes 9 wheelbarrows full of soil, at about 25 shovels full of dirt for each wheelbarrow.  So in filling those 5 beds today, I bent and lifted 1,125 times.  No wonder my legs, back, neck, and shoulders hurt.  Will somebody please remind me why I do this?  The other yard of soil is for some containers.  This week I'll be putting up a fence around the new garden area, and perhaps Debra can plant next weekend.  We won't have to worry about the deer getting to the plants this year.

Yes, we're getting a late start to the garden, but that's okay.   We now have 7 beds in which to plant stuff.  The fall garden should be terrific, and next spring everything will be ready for an early planting.  This year, Debra planted potatoes and onions in late March, and today we picked our first onion.  I hardly can wait to see how it tastes.

Saturday, 04 May, 2002

Spam == Bad Data

In any software system, the only effective way to deal with bad data is to prevent it from entering the system.  User interface programs regularly do this by, for example, refusing to let you enter your name when you're prompted for your date of birth.  If the data is guaranteed to be valid when it enters the system, then the rest of the program can operate without continually checking.  Can you imagine how much more complex and fragile the program would be if it allowed you to enter "George" for your date of birth?  What would happen if the program tried to compare "George" to the current date?

Email spam is bad data.  And once it enters the Internet mail system, it's impossible to eradicate.  It gets passed from server to server, and then to the clients, wasting huge amounts of bandwidth in transmission and causing untold frustration to millions of users.  Why do we let it into the system?  The only effective way to cure the spam problem is to prevent spam from entering the system in the first place.  We have the technology to do this, but for some reason lack the will.  I simply cannot understand why ISPs and large companies have not come up with a system of trusted servers that guarantee no spam.  If somebody can shed some light on this, I'd be happy to listen.

Thursday, 02 May, 2002

Why Keyword Filtering Won't Stop Spam

It occurred to me this week, while I was evaluating anti-spam tools for Debra, and setting up the email filtering rules for her system, that any type of spam filtering based on sender's email address or message subject line is doomed to fail.  Message content filtering based on key words or phrases is also doomed to fail.

Filtering based on sender's email address is just silly.  It's trivial for me to generate a message that appears to come from any arbitrary email address.  Want a message that appears to come from Bill Gates?  No problem.  I can make the "from" line in an email address look any way I choose.  I could write a program that generates random email addresses—a different "from" address for each message.  As long as the domain name (the part after the @ sign) is valid, the message will get through.  So go ahead and block that random address.  What do I care?

There are two things to consider about key word and phrase filtering.  The spammers send mail that contains certain words and phrases that we add to our filters.  Be assured that the spammers subscribe to the anti-spam sites, and they know which words and phrases are being filtered.  So they rewrite their emails with new words or unique constructions of the old words and phrases.  We build a better mouse trap, they build better mice.  It's an escalation game in which the spammers have the upper hand.

The bigger problem with word and phrase filtering, though, is the increasing possibility of rejecting a legitimate message as spam.  As you add more words and phrases to your rejection list, you increase the probability that a good message will contain a "bad" word, and get shuttled off to the trash bin.  If you automatically delete suspected spam, then you've just lost a potentially important message.  And if you put suspected spam into a folder for later review, you've done nothing to reduce the workload.  Why spend time setting up message filters if you have to look through the messages by hand anyway?  The problem of legitimate information being lost during noise filtering is nothing new.  Audio engineers have dealt with it for decades.  That the Internet community seems to think that they can ignore the problem leaves me to wonder whether they've actually considered it thoroughly.

Wednesday, 01 May, 2002

A New Machine

I just can't believe how inexpensive computers are these days.  Today I went to the local Fry's for a motherboard to replace the one in my old Dell Dimension.  Silly me.  Apparently Dell doesn't use ATX form factor motherboards.  So I had to get a new case as well.  I picked up a case, power supply, motherboard, processor (666 Celeron, $20), RAM (256 MB, $60), and 80 GB hard drive ($90)  for a whopping $315.00.  That includes tax.  I brought it home and scavenged the drives (DVD-ROM, CD burner, 3.5" diskette) from the old Dell, and I was up and running.  Well, it wasn't quite that simple.  First I had to put everything together, which is something that I'd sworn off.  Tinkering with the hardware stopped being interesting to me a very long time ago.

People who have had bad experiences with Linux GUIs (slow, clunky, etc.) probably have experienced them on old hardware.  KDE 1.2 was tolerable, just barely, on my old Pentium 200.  KDE 3.0 was painful.  But on my new 666 machine with 256 MB of RAM, the thing is quite nice.  I'm looking forward to getting more familiar with it.  And installing from the DVD?  What a joy, compared to swapping 7 CDs in and out.