Bohemian Paris, Arkansas

Status update

2008-10-28T18:54:00.000-07:00

Sorry about going quiet for so long. My blood counts had been back in the normal range and my symptoms had gone away, so I didn't have much to write about. We thought that the CPAP had taken care of the problems.

Apparently, not so much. In mid-Sepember, the headaches started coming back, and were back full force by early October. I became cold-sensitive again, and started having all the other problems. So, I went in yesterday for something totally unrelated, and had blood drawn.

It's back: hematocrit of 50.3, hemaglobin of 16.7. The 16.7 is actually high-normal, but the hematocrit is abnormal again.

Oh, and my blood is black again.

Searching for stupid people with Google

2008-06-15T13:30:00.000-07:00

Nicholas Carr is right up there with Mark Andresson in my pantheon of people who I admire -- I hold him in almost equally high esteem. His provocative, insightful piece on the relevance of IT departments moved me beyond words, as I wondered how any editor could so cynically want advertising revenue as to publish such tripe.

So I was utterly amazed at the insight, the depth, and the reflective brilliance of his most recent piece on distraction in the modern world. You should read his heartbreaking piece on college students failing to appreciate a brilliant lecture on a massacre that eventually led to the end of British colonial rule in India -- although one might argue that the public committee of inquiry that followed had as much to do with it, and wonder if a similar committee of inquiry on, say, the abuses as Abu Ghraib wouldn't have similar effects -- and then reflect on the response of the students listening and not listening to the lecture.

It's hideous. It's awful.

And it is PAR FOR THE COURSE.

Students don't listen.

Students have never listened.

Lecturing has got to be among the world's most unrewarding tasks. You don't do it for the class; you do it for one or two absorbed listeners, and, if you're honest, for yourself, in order to see if you hear anything new in the material as you run through it. I always did; I never went through a class without learning something new in one of my own lectures.

[As an aside, that's saying more than you might think, since those classes included teaching the equivalent of high school algebra long after I had earned my own Ph.D. Yes, one can learn things from teaching very basic facts, although one does have to refrain from sharing them when they suddenly jump out. Just yesterday, I was talking through the fact that the derivative of the area of a circle was its circumference with my eldest son, who is taking calculus from me, and it struck me that not only is that true for the k-sphere for all dimensions, k, but that that fact is not only significant, but actually an important property of the topological sphere, arising from the unique relationship between the normal of the sphere and its surface. So even now, more than two decades after my degree was granted, I can learn something from going through a simple homework problem in Calculus I with a student.]

Anyway, back to Carr -- if he's never noticed that, then he's more of a fool than a thought...which is more of an achievement than you may have realized.

The funny thump as the car hit the book

2008-06-10T21:08:00.000-07:00

I don't remember the time of year, but it can't have been deep winter, nor could it have been summer. I doubt it was spring; my guess is late fall, and probably November. I don't remember how old I was, but I can't have been younger than seven, nor older than nine. I don't remember much at all.

I don't even remember the title of the book. I just remember that it was cloth-bound, and I remember the rough texture of the material that had been used.

Dad and I were on our way to the Howe Library, and it was dark and raining hard. The street was black with the rain, and suddenly his hand grabbed my shoulder, hard, and yanked backwards. It hurt. There was a loud "thunk", and the book jerked in my hand. The red, faded material on the cover was torn by the impact.

The car drove on, and the driver didn't even slow down.

I still wonder if he even noticed how close he came to the little boy he almost killed.

Silent far too long

2008-06-01T07:49:00.000-07:00

I've been quiet for far too long -- sorry.

News on the medical front has been boring. The various doctors are all in agreement that they don't know what's wrong -- something clearly isn't right, but whatever it is has decided to not be obliging about identifying itself. For now, the approach that they recommend, and that I'm following, is to wait and watch. I'm scheduled for an appointment with a sleep specialist after I get back from my next business trip mid-month to look into treating the sleep apnea. (I need to take care of that, if only so that FDDD can get a good night's sleep, I guess.) We'll track my hematocrit, so that if it spikes upwards, we can change approaches.

Other than that, we do...nothing.

well, well, well...

2008-05-02T09:55:00.000-07:00

So, I'd been planning my own funeral. But life has a way of getting in the way.

The JAK2 test was normal.

Whew.

Now, let's be real -- I still have the symptoms. I'm still BARELY normal in the most advantageous possible situation. There's still a 1 in twenty chance that it's PV...but there is at least one other possible cause, and that one would also explain the polycythemia.

So, now we're off to try and treat sleep apnea.

And I still pump demon's ichor through my veins. So there, D&D fans!

On being a real engineer

2008-04-24T21:04:00.000-07:00

In his blog, my brother calls me a "genuine engineer". [Quick aside. I don't name or provide pointers to any other member of my family, and I try very hard not to provide identifying details of any other person with whom I have direct contact. Earlier this evening, for instance, I edited out a person's name from an older post when I realized I'd unintentionally included it. As a result, I won't provide a link to the posting. Sorry. It's kind of fun. I also won't provide long block quotes, since, given the block quotes, you could search for the original source, and, from that, figure out who he is.] It's a high compliment, coming from him, and I think it's how most people would see me -- but, what's odd is that it's about the last term I'd use for myself.

I'd be willing to lay claim to the term "failed scholar", "failed actor", or "failed musician", although, to be fair to myself, none of those claims is true. I wasn't a failure as a scholar, merely not a success to the standards I demanded of myself. I published a number of interesting papers, and, even now, continue to do interesting and original work, albeit far-removed from my original field. I never tried to make it professionally as an actor or a musician -- I knew that I didn't have the fire in my belly to make it on Broadway, and my voice, while exotic, is really too heavy for me to have found many roles. After all, you can only sing Sixteen Tons so many times each year, there's a limited market for interpretations of Tom Lehrer songs as German Leider, and you can't count on a dinner invitation from Don Giovanni every weekend.

If there's one role other people see me in, it's "computer programmer". It's one I'm good at, and one that I enjoy, particularly in combination with the "algorithm and architecture designer" role -- which is why people call me an engineer -- but, oddly, it's not the label I apply to myself. I'm not an engineer, you see. I'm a craftsman.

Engineers worry about things like reproducibility, cost, and process constraints. They worry about large scale production constraints. I worry about very few of those things -- software is easy to copy, and, once one takes the inevitable hardware-induced failures into account, scaling and scalability aren't really interesting or difficult problems to solve. (No, really. They just aren't. Don't let the hypemeisters tell you otherwise; if someone is telling you haw hard it is to scale something out...they're lying to you.)

Software "engineers" build one-off items, with simple tools. We smooth out rough edges, and our standards of productivity are those of a medieval guild master -- elegance, and fitness of decoration to form and function. In fact, our internal hierarchy is very similar to that of a medieval guild's, with apprentices, journeymen, and masters -- only we call them "contractors", "full-timers", and "distinguished engineers".

So, much as I am flattered by the term...I'm not an engineer. I'm a guild-member, and a journeyman, at that.

Demon ichor...

2008-04-18T17:34:00.000-07:00

One of my friends went through chemo for lymphoma earlier this decade. He described seeing PV patients being phlebotomized while he was being infused, saying "Their blood is just weird: it's black."

At the time, I thought he was exaggerating, but now I know I was wrong: whether or not I have PV, my venous blood is now black and opaque, like India ink, even in a 5 mL tube. Not red, not even purple, but black and kind of..syrupy.

It's cool. My blood now matches my clothes.

Hematologist update

2008-04-17T19:16:00.000-07:00

It turns out that one of the hematologists who literally wrote the book on PV is based just a short ways away from where I live -- and I was fortunate enough to be able to get an appointment to see her today.

So, I took the afternoon off from Gollum, drove to the hospital where her clinic is, and had a quick (approx 1 hr.) meeting with her.

(Quick aside: 90+% of all cases of PV are due to a small class of loss-of-down-regulation mutations in a single gene on the X chromosome, referred to as JAK2. The so-called "JAK2 screening assay" tests for the presence of erythrogenous stem cells with these mutations, although I don't know the mechanism for the screen. I'll need to follow up in a later post to explain that.)

In any case, the numbers are less dire than I had initially been led to believe. That said, I'm still a clear candidate for PV. Her argument was that although, yes, the headaches, the itching, and the confusion I've experienced were Protean symptoms (her word, not mine -- what a lovely term), they all are symptoms of PV, and, in conjunction with the clearly elevated hematocrit levels in my blood, and with the apparently sudden increase in the levels, the possibility of PV needs to be kept in mind.

I'm taking that as a pretty broad hint that she thinks it's PV, but, being a smart woman, knows better than to commit with the additional diagnosis of secondary polycythemia due to sleep apnea still live.

All that said, I don't have an enlarged spleen, I'm not very flushed, my liver is of normal size, I have a "thunderous" pulse in my ankle, and while, yes, I have a heart murmur, it's exceptionally subtle. (Dr. B. (sardonically): "That's quite a subtle murmur." Me: "You can actually hear it? Most physicians can't, you know. I've decided that I don't actually believe in it, Doppler echocardiogram or no." Dr. B.: "Well, yes, I think I can, but it's certainly tenuous.")

Essentially, despite an "exciting medical history", I'm basically a medical conundrum: a 47 year old man in exceptional health with an odd hematocrit score.

It could have been cancer

2008-04-03T11:38:00.000-07:00

About a decade ago, I started having night sweats. Bad ones -- you know, soak the sheets night sweats.

Say that to your doctor, particularly when you dress in black and have a voice minor, and you get knowing nods of the head and questions about HIV status. No amount of calm recitation of the risk factors, and of how they don't apply to you, will satisfy them -- if anything, knowing the risk factors makes things worse -- they want an HIV test, and so you say "sure" and of course it comes back negative. The feeling that they just called you a liar, and implied that you'd willingly threaten the lives of your wife and children...you're supposed to ignore that.

(Yes, of course there are other vectors. It's the knowing nod that is the insult.)

And, while you're at it, if you're me, you pick a new doctor, and don't say why. I never have until now. That's why I didn't like that particular physician.

But the night sweats never got better. I just stopped talking about them. And the fatigue kept getting worse. In the last couple of years, in fact, I'd gotten to the point that I was sleeping fifteen to eighteen hours every day when I didn't have to go to work. I was dizzy a lot. Then the headaches began.

and the joint aches.

and the dizziness.

Nothing crippling, but...a bunch of little things. Nothing incapacitating, but the collection was starting to wear on me. Was I working too hard? Well, I couldn't rule it out -- Gollum is an exciting place to work, and I do love my job. Was I just turning middle aged? Well, that too.

FDDD finally convinced me to go in for a check up; I'm a few months overdue.

And that's where I got the news. The doctor went through the usual routine -- prostate, reflexes, blood pressure, the lot. off the platform, over to a chair, to review the test results...

"Wow, your glucose results are really good! That's much better than I'd expected, given your history. Your lipids -- well, your hdls are low, your ldls, high -- you know about that, and I've already lectured you about that [quick smile] -- don't suppose it'll help to go through it again? no, probably not -- your blood counts -- yeah, norm...wait."

She stopped mid sentence, looking at the results, and it hit me that she (a) must not have looked at them before coming in, and (b) had just seen something that probably wasn't normal.

She resumed, carefully.

"Your red blood cell count is...quite high. It's 36 million cells/[gobbletygook], that's unusual."

And that was all. But that one sentence changed my life forever.

You see, it explained everything, and it opened the door into a funny grey area between "real" sickness and "normal" health.

If you go do a little looking around for "High red blood cell count" (and 36M/mcL is very high; normal for someone like me would be 4-6M/mcL), you quickly encounter the words "polycythemia vera" -- a clonal myeloproliferative disorder resulting in excessive production of red blood cells. It's a sort of poor cousin to the Big Bad Boy -- acute myelogenous leukemia, which is the same thing with white blood cells. The difference? Well, to put it simply, PV is uncomfortable and humiliating, causing night sweats, headaches, itching, fatigue, thromboses, and general irritation. AML causes one thing -- death.

Given the choice, I'd rather have PV, and, fortunately for me, I do.

Indeed, according to the doctor during a later visit, that's almost certainly what I have. There are a few things I need to do for now -- aggressive hydration, meds for the headaches and to help me sleep, CBC monitoring to see when I need to start being phlebotomized, maybe some changes to my work environment to compensate for the fatigue -- but...that's all.

It could have been cancer or AIDS. I can be very glad of that. On the other hand -- not only did I dodge the bullet, I heard the crack as it went by my ear.

End of an era

2008-02-03T08:36:00.000-08:00

A new voice entered the tech scene about ten years ago. Brash, radical, and revolutionary, Slashdot helped to propel the tech bubble into the twenty-first century, and to push Linux and its associated technologies into the forefront of the Valley's reporters' collective minds. That's important: like the Village Collective in politics, the Valley Collective has a disproportionate effect on tech coverage. It also has an abysmal record of accuracy or foresight.

Yesterday, Microsoft bid for Yahoo, and, in the process, showed how how much the world had changed. Oh, yes, the bid for Yahoo was a big deal since the story of the failing tech behemoth being swallowed by the fast growing upstart is always a big deal. What I was interested in, though, was the reaction on Slashdot.

None. A bunch of stories, and no more than a couple of hundred responses. Lots of opportunities to bash Microsoft, and...nothing. I did an experiment, and posted a particularly obnoxious message attacking the herd mind's interpretation...and nothing. I haven't even been troll rated. (Update. Okay, something has happened now -- someone has posted a response to my comment supporting it.)

Here's the big news from yesterday's offer: Slashdot is dead. It's the end of an era.

Your numbers don't add up

2008-01-25T19:21:00.000-08:00

I've enjoyed Brad DeLong's fruitless search for a grown-up Republican, even though he's had about as much success in his search as Diogenes did in searching for an honest man. It may be that the Party which hates ... oh, look, a jackalope! simply doesn't contain any grown ups any more, but, as he and others have pointed out, there used to be grown ups in the Republican Party -- people who'd look at you, smile condescendingly, and say "It's a great idea, but it isn't going to work. How much will it cost? How will you get people to comply? If they don't do it voluntarily, then how much will enforcement cost? Your numbers don't add up -- and changes, however beneficial they appear to be, don't stick unless their numbers add up, too."

Damned if they didn't keep being right, too. Filthy grown-up Republicans. I miss them sometimes.

One of the reasons I miss them is that it leaves me in the uncomfortable position of needing to be a grown-up Democrat. I've got the condescension down, mind you, and the whole superior tone of voice -- it's the responsibility I hate. It would be so much easier to just ignore the whole thing and move on.

The idiocy of carry-on limitations

2007-12-29T09:10:00.000-08:00

Let's start by saying two things. First, lithium batteries are a fire hazard if the are punctured. Second, flammable or explosive liquids and gels on a plane are a threat to the passengers. There are easy ways to create a hazardous environment on an airplane given either one. For instance, alcohol is highly flammable, burning in 20% (well, actually, 21%) solutions in water in at surface oxygen tensions. At higher concentrations, such as 40% alcohol by volume, it burns quite readily. Clearly, we should be banning it for fear of terrorists using it as a weapon.

For some reason, however, airlines are not eager to forbid vodka, brandy, and whiskey in the passenger cabin.

Similarly, lithium batteries are a fire hazard. No doubt about it. There are a slough of reports of them catching fire while on somebody's lap, and the chemistry involved is certainly highly energetic. If you drive a nail into an battery pack with enough force, the pack can even explode, although that's never been recorded "in the wild" as far as I know. There have been no reports of spontaneous combustion of disconnected batteries at all. As a result, it's clear that lithium batteries should not be allowed on airplanes, at least not in the dangerous configuration in which they're plugged in to a computer.

For some reason, airlines are also not eager to forbid business travelers the use of their laptops during flight.

Safety regulations exist for a good reason. Feel-good Code Brown restrictions, though? Not so much.

Death march

2007-12-25T14:25:00.000-08:00

My product is (finally) wrapping up version one, running down to release, however you want to say it. At Gollum, this is a particularly big deal, particularly the company has a history of releasing products very fast. That works for some products, but not for others, and since mine is one of those others, this is a big deal.

Either way, it's meant that I haven't had much time to invest in the blog, because all I'm thinking about is the product. With luck, that will change next month.

H1B

2007-11-27T06:53:00.001-08:00

I've been thinking a lot about the H1B shortage recently. I think it's likely that the availability of foreign-born indentured servants through a training period probably does depress the starting wages in the market, simply by availability. (The availability of a pool of H1B candidate makes the prevailing local wage a cap instead of a median. Since this is iterated, it serves to artificially depress the local wage for all candidates over time.) Moreover, not only do artificially depressed starting wages directly depress wages for more senior employees, but the inability of those workers to seek employment elsewhere further depresses wages throughout the first few years of employment, the time when most salaried workers see their fastest wage growth. The net result is a broad negative impact on the incomes of technical workers.

Now, to be fair, this is a problem which could be easily fixed by changing the terms of hiring. If, instead of setting the price of an H1B hire to be the prevailing wage, employers were required to pay each H1B at or above the 90th percentile of comparable workers in the local economy, to publish their pay, and to maintain parity at that level throughout the training period, H1B holders would become net inflationary to all wages in the market. That would make them less desirable hires for "average" workers, and would actually reserve them for only the superstars who we, as a country, want to hire and for whom the H1B program is allegedly designed.

There are excuses, and then there are lazy excuses

2007-11-27T06:41:00.000-08:00

I was reading through one of the blogs on zdnet this morning (here, if you must know), I found a particularly choice piece of idiocy:

For those of us incapable of such engineering George’s efforts are interesting. For those of you inclined to take on these projects yourself go for it. I’m stuck in the “buy because there’s no way I can build without blowing something up” camp.

Building a computer is easy. Really. It's pretty tedious, but it's easy. Routing out all the corners of a wood box isn't quite as easy, but...look, even I can do it. No, the only reason to not build your own boxes is laziness.

Anonymity as a principal value.

2007-11-11T10:10:00.000-08:00

John Cole, who writes at Balloon Juice, makes an interesting claim in a recent posting:

Privacy no longer can mean anonymity, says Donald Kerr, a deputy director of national intelligence. Instead, it should mean that government and businesses properly safeguards people’s private communications and financial information.
... [T]his is antithetical to the concepts that used to be the backbone of conservatism

(The outer quotation is Cole, commenting on Kerr's remark.)

I don't agree with Cole, though. A right to anonymity was a core conservative tenet only when only the wealthy could actually afford it.

To be anonymous in a world where the dissemination of information required a printing press meant not only owning your own printing press, it meant having a place to hide it, and a mechanism for disseminating your published works. The resulting anonymity was largely a matter of convenience -- there were individuals who could be worked backwards through, and who could be suborned or threatened if the need genuinely arose. In order to be truly anonymous, you needed to be able to do the equivalent of paying someone who wouldn't remember you to carry the papers around. That takes time, and that, in turn, takes money. (Similarly sending something by post with no return address, etc. How are you going to publicize your work? Public meetings? Flat out.) With enough money, a close approximation of anonymity could be built up, but it was only available to those who could afford it.

Today, almost any sufficiently paranoid person can be genuinely anonymous if he or she wishes to be. It isn't easy, perhaps, but it is possible. The intermediary people, the weak points in the previous model, can be replaced by computers with long enough keys. (For those of you who want to talk about NSA spooks reading things using atomic force microscopy or the like, there are ways around that, too. It has to do with recovery rate and key length.) Those computers can't be bribed or threatened. Distribution can be performed by sending email to a throw-away account.

Now, if I were a cynical cretin, I'd suggest that this was due to the "equal rights for rich folk" behavior of American conservatives. For once, though, I think that this is something else. You see, even the "anonymous rich" would not, in fact, have been truly anonymous. Encountering a genuinely anonymous actor, I'd know immediately that there were only a few people who could actually pull it off, as only a few people would have both the money and the contacts. Given that and enough resources, I could still track packages back in time, and people forward in time, to find a nexus. It would take time and money -- but those are things governments have. The myth of anonymity was really a big lie, one which we all told ourselves because it was convenient.

That has changed. I'm not surprised to see the authoritarians around Bush running away from their alleged principles when they become inconvenient, of course -- but I don't expect anyone on the political spectrum to actually hold onto true anonymity now that it's possible.

Of course, that may simply prove that I'm truly a cynical cretin. I certainly haven't yet found an honest man that I would fully trust, so I probably have the key characteristics.

Guys, he's not running.

2007-11-08T20:30:00.001-08:00

OK? Al Gore is not running.

Here's a bit of evidence:

Everywhere he goes, the winner of the 2000 popular vote is asked: "Would you consider running for President in 2008?" His answer is that he has no intention or plans to seek office.

Where is that drawn from? The lower left hand corner of http://www.algore.org -- one of the many Draft Al Gore sites.

Now listen very carefully. That's called a denial. He isn't running. He's said that repeatedly. It isn't news. Yes, you and your co-confabulators may have made yourself believe that he doesn't really mean it, but that's kind of like the guy who keeps sending the girl flowers even after she says she never wants to see him again. And after she puts the bouquets on the top of her trash pile. And after she eventually hires an anonymous assistant to make sure she is never even seen in company with one of the bouquets.

Stop asking for an answer from him -- you already have the answer, and you don't like it. You're busy blaming Gore for saying no repeatedly because you don't want to hear that answer. You were owed an answer; you got it. You are not owed a thank you, though -- you went from a suitor to a stalker when you kept sending the bouquets. All you're owed is the political equivalent of a restraining order.

Now, please...go away. Don't go away sad. Don't go away mad. Just. Go. Away.

Looking for Godot

2007-11-04T10:49:00.000-08:00

In his classic play, _Waiting for Godot_, Samuel Beckett sets two men to waiting -- interminably, and probably hopelessly -- for "Godot", who will come and rescue them. Godot, obligingly, never arrives.

Now, in the modern theater of the absurd which our Presidential elections have become, our press has gone the next step, looking, desperately looking, for Godot to come and rescue them from Hillary Clinton. I can't blame them: where Reagan was frequently on-message, he was on-message because he was reading somebody else's script, and his gaffes made good television. Bill Clinton was relentlessly off-message, however brilliant he was, and so made good theater. The Bushes, father and son, are both war Presidents, and, again, war makes great ratings, making up for their respective blemishes, the father's patrician colorlessness, and the son's smirking folly.

I suspect, that the press fears Hillary Clinton. Someone who is as smart as Bill, but won't stumble, who is relentlessly on-message, but can ad lib while staying on message -- such a person will make their job much harder. They want either Bill's lack of discipline or Ronnie's unpredictability. Hillary, cheerless, hard, and driven, offers them none of that. Worse, they fear -- correctly, I suspect -- that she'll hold her political enemies responsible for their behavior, thus reducing the value of Republican Congressional sources of vitriol. She looks like a school-marm, and she's not going to be "fun". It's going to be a long dry season if she's elected.

No matter, though; this posting isn't about the press and its fears, but rather about its follies. For better or for worse, Clinton is building a strong plurality among Democrats, and, given the nation's current bent, she'll win going away in 2008 if she gets the nod. That's going to be terrible for the bottom line of the networks; they make (quite literally) tens of billions of dollars off the advertisements the election spawns, and if Clinton is a walk-away, they'll be forced down-ticket, reducing profits. Worse, news stories and other "public service events" about a contested election draw in consumers, who then stick around to watch more advertisements on subsequent shows. Since politically-aware Americans tend to be people with more disposable income (after all, they can waste time watching the political chat-up shows), their eyeballs are quite valuable.

And it is beginning to look a lot like Clinton with relatively little internecine conflict. No conflict == no primary viewers -- the only thing worse would be no conflict in the general, leading to fewer eyebal...err, I mean, engaged voters to show commercials to. So, if you're a news hound, confronted with this dreadful prospect, what do you do? Why, it's obvious; in a tradition going back to _The Maine_, you *create* a controversy.

No, I'm quite serious. Look at what's happened in the last week. First, Chris Matthews throws negative question after negative question at Clinton during the debate, hoping to catch her out. He fails outright at least 21 times out of 22; on try number 22, he tries to rope her in with a question about Elliot Spitzer's proposal to give driver's licenses to all residents of New York, ignoring their legality of residence. She actually gives a good answer, talking about the complicated reasons that it might not be a bad idea, but then refuses to either endorse or condemn the plan. Immediately, MSNBC starts trumpeting that as the "first crack" in her candidacy, and asking, in purple prose, if there's finally a horse-race. The rest of the press jumps in.

We're ginning up a controversy, yes we are, laydees and gennulmin! Cept for a small problem -- the polls come back, and the American people have responded by shrugging. No, it isn't a definitive answer, but, gosh, you know, it's not a problem with a definitive solution. It's certainly not a stupid answer.

So, what's on today's front pages? A declaration that if Michael Bloomberg jumps in, he'll create a big third party threat. Why, it's Ross Perot all over again!

Except that there's no reason to believe it's true. Mostly, it's a puff piece, again, trying to create advertising impressions, in a desperate hope that the American people will welcome them with candy and flowers. It isn't true this time, either, guys.

Once more with feeling

2007-10-28T09:06:00.000-07:00

In a piece republished from the New York Times today, I see an echo of the defeat of the press. The piece (here) says:

However conceived, though, the result is a new interest in public policies that address problems of peace, health and poverty — problems, unlike abortion and same-sex marriage, where left and right compete to present the best answers.

Help me here, please. How is it that the left and right don't actually compete to provide the best answers for abortion and same-sex marriage? Is it because, like on issues of race and gender equality, the left was right all along, and it took a long time for the right to divest itself of the bigots that "led" on those issues? Or is it, because his editor's ears had gone deaf, the reporter got to slide the newest lie about the left into the text? I suspect the latter, if only because the reporter is so utterly stunned that the Democratic front runners are more moral in person than the Republicans are this year.

It's no accident that the Democratic front runners lead more moral lives than their right wing rivals. The Democrats, liberals all, are comfortable with the core doctrine of liberalism, that people really are qualified to make their own decisions, whether individually or collectively. That doesn't mean that government has no role to play outside a very narrow (and objectively anti-poor, anti-middle class) set of issues, but rather that people are qualified to make decisions about their own lives, and that the power of enforcement should be used as little as possible. True liberals trust other people to make reasonable decisions about their own lives, although they understand that people may fail to make such decisions.

Beliefs have consequences, though. If I repeatedly say to you "I trust you to make this decision or that decision", then I must be willing to believe that people typically make good decisions, and, at the end of the day, that depends on me, myself, having the experience of making difficult rational decisions. As a result, someone who's willing to extend freedom to others over a long period of time will usually be doing so based on having made good decisions themselves, despite temptation or abuse. That is, liberalism and libertarianism are political philosophies which select for self-controlled people who can be trusted with power over themselves.

By corollary, authoritarianism and radicalism would tend to be most attractive to those who have "uncontrollable urges" -- a pattern we've seen a lot of in recent years in the leaders of the "religious right". Tom Delay, Grover Norquist, and Jack Abramoff were all terribly susceptible to sin of greed, and were unable to restrain themselves from taking everything they could. Ted Haggard, Larry Craig and Mark Foley were all horrified by their homoerotic impulses, and by their inability to channel them into normal behavior, and so they were and are unable to imagine that normal people feel sexual urges all the time, while still managing to keep from tearing each other's clothes off in the office. Rudy Giuliani is a serial betrayer of wife and family, and so believes that nobody is trustworthy. Mitt Romney will do anything for power, and so imagines that all others are corrupted by the same vice.

There's nothing special about conservativism itself here, by the way. Stalin in Russia, Mao in China, Kim Il Sun and Kim Jung Il in North Korea, Hussein in Iraq, Assad in Syria...there are plenty of left wing authoritarians with similar patterns of deceptive left-wing thugs. Closer to home, one can raise the examples of the Black Panthers, Elijah Mohammed and the Weathermen as well. It's only that the right in America right now is the home of our home-grown authoritarian movement.

Maybe the author of the New York Times piece was saying that the left was unquestionably correct about about gay rights -- which include, but are certainly not limited to the right marry -- and reproductive rights -- which include, but are certainly not limited to the right to a therapeutic abortion. I kind of doubt it, though.

Hating on Apple (III) -- one for the (e)rode

2007-10-27T10:19:00.001-07:00

I thought about upgrading my MacBook Pro to Leopard, figuring that I couldn't possibly have a less positive experience with it than I do with the current version of the OS. I never got around to it, though, because, last night at 2:47am, the computer died. Dead. As in, "I won't boot, demi. Deal with it." So, come Monday, I get to return my consumer device and get a new one.

I think I'll go with Gollubuntu on my next laptop. It's wretched, but it's less wretched than MacOS. (Gollum maintains a custom version of Ubuntu for laptops. I figure it's my next stop on the path through the OS's available locally.)

Before I leave the tale of the fruit, though, I want to mention the particularly hideous experience of multiple monitors. Now, this surprises a lot of people, because Macs are famous for their superior multimon support. I'm told that the experience for horizontally arrayed monitors is truly exquisite, and I'm willing to believe that. However, the experience for vertically arrayed monitors is dreadful.

Contextual menus regularly show up on the lower monitor, rather than upon the top one. Windows appear and disappear randomly. There are multiple mouse cursors visible when the computer goes to sleep.

The Mac -- it just works you up if you just want to work.

Wanted: a +5 circular saw of eternal sharpness

2007-10-22T06:31:00.000-07:00

I don't quite know what to say about this, but I'm having a grand time thinking about the classifieds it will spin up.

Friedman misses the point

2007-10-21T11:02:00.000-07:00

In his 10 October 2007 editorial, Thomas Friedman opines about the failure of modern college students to protest in the streets, unlike his generation. He marvels at it, wondering how such idealistic students aren't following their ideas like his generation did. (Friedman was born in 53, for what it's worth.)

Of course, the truth is that they are. Friedman, like many of the people seven to ten years older than me, has a "French Resistance" lie about his time in college: everybody was out there fighting the bad folks, and everybody was doing it for a good reason. In fact, like the French, most students of Friedman's generation didn't participate, and, also like the French, not everyone did it for a good reason.

Unlike the French, though, almost nobody who was active in the "anti-war" movement in 69-72, particularly, did it for a good reason. The truth of the matter, Thomas, is that your generation evaporated from the streets the instant the draft was repealed -- you guys posed about opposing the war, but really opposed the draft. That was a fine thing to oppose, perhaps, but it's not what you and your ilk say.

And, as someone who is the right age to have watched the older, cooler kids betray me by lying about their ideals, I am not about to forget that.

Hating on Apple (II) -- more on X11

2007-09-23T11:28:00.000-07:00

Let's drill down on that putatively improved implementation of X11 while we're at it.

In a Windows environment with a rootless X server, the individual applications running through the X server can easily be made first-class objects, which can be tabbed through just like any other application window. That's important, since an X server can easily host many applications -- it is, after all, a virtual display which services one or more computers. On a Linux box with an X server running, the same situation obtains.

On OS X? X is a single application. That means finding a window served by the X server requires bringing up all X windows, then carefully walking through the windows without any sense of which one is which. There's no way to build a window group corresponding to "emacs windows targeted at possessing-size.inner.gollum.com", even though every other X server provides exactly that. (And expose doesn't solve that problem. Emacs windows look very much like other emacs windows, so it's still all guesswork.)

Oh, yeah, and then there's the behavior of caps lock. On many X applications, caps lock is intercepted by the server about half the time. Thus, I hit caps lock, and it's both handled locally and handled remotely -- CL turns on in the X window. Then, I hit it again -- and it's only handled locally! Now, CL is off locally in all windows outside X, but on inside X. There's no way for me to figure that out -- which means that typing in passwords becomes an exercise in frustration. Tap it again? On both places. Tap it a fourth time in quick succession, and now, lo and behold, it's off both places.

This is a clean UI? Please.

The Macintosh -- it just works. As long as you don't actually use it as a computer.

Hating on Apple (I): picking the portable up

2007-09-22T13:55:00.000-07:00

I guess that if I propose that Apple MacBooks suck -- as, of course, they do -- I might be generous enough to provide evidence for that fact. I mean, even such notorious haters as Andrew Orlowski claim that the machines have a better and more secure operating system than Windows, yet I have the temerity to believe otherwise.

Here's a chronic nit. I have two workspaces in my home, one upstairs in my bedroom, and the other downstairs in my office. I carry the machine from one to the other when I change places, and, whenever I do so, it fails to waken from slumber. You'd think that Steve Jobs and the job shop schedulers could manage to make a portable that could be moved, but evidently not.

Maybe it would be better if I ran any non-Apple software on the machine? Perhaps I should connect wireless while I'm carrying the box around? I don't -- the only way that computer ever gets used around the house is either untethered and completely disconnected, or when plugged in to a wired network. (Yes, I'm a geek -- I have RJ45 jacks in both my office and in my bedroom. I'm only redeemed by the fact that the RJ45 jack in my bedroom is almost thirty feet from the bed.)

I run the X window system in order to connect to my workstation back at the office. Almost nobody does that, but it's one of those things which makes a Mac a Unix box. Did I mention that it's actually (allegedly) Apple improved, and that it ships as a part of the operating system? Yes, it's cool that way. In fact, it's been improved more than I could imagine -- I did not think that anyone could mess up the clipboard any more than it is on a "normal" Linux box, and the Mac shows me that I was wrong.

I mean, on a Linux box, even when all other forms of the clipboard have failed, you can always use highlight-and-middle click. On a Mac? Nope, not even if you've got a three button mouse.

I thought for a while that it was just a completely separate clipboard, but then, the other say, I discovered that I can highlight text in some X applications, and paste it into TextEdit, and from there paste it into almost any other Mac application. Of course, I can't copy the other way.

The MacBookPro: it just works. As a doorstop.

Hating on Apple...still

2007-09-16T08:10:00.000-07:00

One of my alteregos was recently told to get past his ph33r of teh Jobs, and take a walk on the Mac side. "The water's warm," my colloquitor said. I found this unutterably amusing, as I read it on a MacBook Pro.

I've been told for years that "In order to understand the Macintosh, you need to really use one for a month, six months, a year. You need to get your mind into it. To work in the framework." OK, so I was offered a chance to do just that: my current employer allows us to specify the kind of laptop we get, and so I chose a Mac. Gollum takes good care of its engineers, and I'm no exception; they provided me with a top-of-the-line MacBook Pro, 2GB of RAM, 80GByte hard drive, the whole package. The company provides great tech support in all cases, and, in my office, for Macintoshes, it provides unparalleled support. My Mac has had the best possible chance to serve me.

That was a year ago. I've given Tiger the year it deserves. I've worked on this computer, both locally and remotely, experimented with it, used it for multimedia work...in short, I've given it a good shot.

It sucks. It is awful. It is an unutterably painful experience.

Don't get me wrong. The hardware is lovely to look at. The screen is quite nice: high contrast, bright (when plugged in). The default screen background is a little too aggressive for my tastes, but there's a great selection of pretty backgrounds provided. The computer itself is pretty, in a cold and metallic kind of way, and it is thin and light. When compared to the Windows/Linux laptops of the same era, it is just prettier. In fact, when compared to the Windows/Linux laptops of any era, it is just prettier.

If it were a coffee table decoration, then, my Mac would win hands down. Coupled with Intel's (probably illegal) price subsidies for the Core and Core Duo, it's even a not-entirely laughable coffee table decoration. I mean, when I bought my eldest son a computer to take off to college earlier this year, I got a Windows laptop running Vista, with a larger screen but a smaller amount of memory and a slower processor -- and no brushed metal finish -- for about a whole third of the price I'd have payed for a Mac. Including the memory upgrade and a retail upgrade to Vista Ultimate, I'd have payed almost half what a Pro would have cost!

For a thousand bucks, I could have had a better coffee table decoration. Unfortunately, I don't need a coffee table decoration; I need a computer, and it's there that the Mac falls down.

Mark Twain was wrong

2007-08-18T18:55:00.001-07:00

One of the most moving passages in Life on the Mississippi, at least to me, is in Chapter Nine, "Continuing Perplexities", in which, Twain writes a marvelous elegy on the cost of expertise:

Now when I had mastered the language of this water and had come to know every trifling feature that bordered the great river as familiarly as I knew the letters of the alphabet, I had made a valuable acquisition. But I had lost something, too. I had lost something which could never be restored to me while I lived. All the grace, the beauty, the poetry had gone out of the majestic river!

It's a wonderful piece, expressing the loss of wonder, and the cost of knowledge. To Twain, then a man of roughly my own age, the acquisition of skill had erased his ability to see the river as a thing of beauty, replacing it with a book of poorly known but deeply studied hazards. He's talking about one profession, but clearly means to apply the knowledge to all, repeating one of those things which makes sense, and which is clear and sure: knowledge conquers innocence, and, in so doing, destroys simple joy.

Problem is, it's not always true, at least for me.

I'm a computer programmer, and, as a result, when I look at a program running, I can often see the subtle pitfalls and design mistakes that the author made. A hesitation in the UI, and I can feel the on-coming crash -- like Twain's master, intuitively telling it from something scary but benign. For my own code, when I'm working on it, that feeling dominates my experience of the program; each click and whir of the disk drive, each extra, unexpected bit of white space on a display, can be the electronic equivalent of the sharp pain in the left arm that signals cardiac arrest.

So, when I choose to experience the world in that way, I lose the ability to play a simple computer game, or to use an editor. I become a slave to my own expertise. But what Twain does not say is that it is a choice. Yes, when I'm working, it's a rare moment that I see the experience of a program the same way that one of my users will, although, even then, I get the occasional flash of "gosh, that's just cool" which I want to convey to my users. But I can turn that off, too, and simply experience the program as a gestalt, working or not, exactly like anyone else.

An open letter to the House Democrats

2007-08-05T11:51:00.000-07:00

So, you caved once again to your right wing. Fair enough -- you understand the language of power, and your attitude towards your left wing is "where are you going to go if we screw you over?"

Ask Joe Lieberman.

You assume that I won't support a primary opponent against you at your peril. I will -- even if you're in a risky district. You're wrong. I will. More than that, since I'd rather stick to the devil I know -- I'll give money to your Republican opponent in hopes that he unseats you in November.

I can stand two more years of a do nothing House; on the issues I care about, I've gotten it anyway.

Swiss "efficiency"...I think

2007-07-28T10:11:00.000-07:00

So, this week, I'm in the City of the Gnomes, in Gnome-man land, in Zürich, visiting one of the offices here, where a project closely related to, and dependent upon, mine is based.

It's been an experience. The city is beautiful, and old, and pleasant. There are still fish in the river in the middle of town, for instance, and the water is the wonderful milky green of glacial runoff. For the most part, I've been able to get along in English -- the company office here works entirely in English, which is the only common language shared by all the staff -- and I can read enough German to order a meal, find my hotel room, or navigate the city. I've had to fall back on French once. I'd be ashamed to actually try to answer a question in my fractured German.

The place is almost eerily well-ordered, though. It's not just that the city feels more like Kirkland than Seattle, and certainly nothing like Boston, Miami, or New York, but that it's so meticulously regulated and controlled. In order to ride a bicycle in town, for instance, you need to buy comprehensive insurance from Swissposte, and put a sticker on your bike, or risk a very large fine if you're involved in an accident. Well, that's not the end of the world, I suppose, but...still...it bugs me.

However, the Swiss reputation for efficiency appears to be somewhat exaggerated. There have been several odd issues -- people calling me at midnight, asking for Herr Engler, for instance -- but I thought they were silly. Until this afternoon, when, as I sat innocently at my computer, reading, somebody opened the door behind me, looked in, and ran away, only to be followed by someone else, who asked me, basically, what I was doing in the room.

It turns out that the room is double booked, and that I was given the key to the wrong room. All I can say is...wow. How can a hotel screw things up so badly?

Stupid IM lines of the day

2007-07-20T08:26:00.000-07:00

On my status bar today, I see one of my colleagues' status line set to "have you seen all the people sunbathing by their idling cars outside?"

My response is, "well, are they supposed to sunbathe in their cars? Or inside?"

Mommy, mommy, I don't want to build manhole covers!

2007-07-04T12:01:00.000-07:00

OK, so I seemed a bit disingenuous in my recent post about hiring Ph.D.'s. I criticized Marc Andreesen for his criticism of hiring PhD's without taking into account that he didn't criticize hiring them, per se, but rather the fact that they're the hardest people to motivate to ship commercially viable products. (His words, not mine) It isn't the same thing -- after all, maybe Ph.D.'s can't really be motivated to produce high-quality products except their dissertation research.

Of course, once you put it that way, you realize that the statement is obvious, and stupid. Of course some Ph.D.'s are unable to produce anything commercial. So are a lot of B.A. candidates. Do you refuse to hire any B.A. candidates because some of them might fail? No, of course not -- that'd be stupid. You meet with them to determine if you can motivate them to work on things unlike the kind of project with which they have familiarity. You check to see it they can actually write code. You investigate their interests.

You interview them, and you look at things like how they came to the particular project that they wrote up in their dissertation. If you're hiring them to fill a job that requires writing code, you ask them to write code, and measure their performance at that task. And you ask them questions which measure their flexibility and willingness to take intellectual challenges that are unlike the ones they've taken in the past.

I've heard of a great question that they used to ask at Microsoft along those lines: "Why are manhole covers round?"

Other than that, Mrs. Lincoln, how was the theatre?

2007-07-01T11:02:00.000-07:00

From the front page of ZDNet today:

Despite missing features, a slow data network, and call quality that doesn't always deliver, Apple's iPhone sets a new benchmark for an integrated cell phone and MP3 player.

In other words, in spite of being a poor phone on all features that matter -- and, because of the slow data network, a poor integrated phone/MP3 player -- this sets a new benchmark for integrated phone/MP3 players.

But how are we going to make manhole covers?

2007-06-29T17:22:00.000-07:00

In a recent post, I talked about Marc Andressen's interesting piece on hiring, focusing on his misconceptions about Microsoft's famous pattern of asking puzzle questions. (By the way: the number of puzzle questions in an interview was greatly over-estimated. Most loops contained no more than one, and more than two was a sign of a failed loop. Candidates remembered them, so they became Microsoft's hallmark. The reality was rather different: most questions have always been focused on the candidate's direct professional abilities.) In this post, I'm going to talk about another misconception Andreesen perpetuates about hiring -- and, yes, relate that back to Netscape's fall.

Speaking of Google's hiring practices, Andreesen says:

Have a PhD? Front of the line. Masters? Next. Bachelor's? Go to the end.
In apparent direct contraction [sic] to decades of experience in the computer industry that PhD's are the hardest people to motivate to ship commercially viable products -- with rare exception. (Hi, Tim! Hi, Diego!)

First, and most importantly, the "apparent contradiction" reflects Andressen's conception of the market, not any real measurement of the industry. Both Google and Microsoft did a lot of measurement, and discovered that the "decades of experience" in the computer industry was imaginary. In fact, developers with doctorates and motivation to be developers are usually very productive.

That should be obvious. The key predictor of future behavior in an employee is past behavior. As a rule, barring injury or illness, employees repeat their past patterns of achievement, both good and bad. That's why behavioral questions (the ones which start "tell me about a time when...") are so effective if you know the kinds of problems the candidate will need to solve in the position.

What has a man or woman with a Ph.D. done, then? Seriously, getting a Ph.D. isn't all that hard. A new Ph.D. did well on some exams, took some extra courses, and independently completed a body of novel thesis research, then wrote it up.

But look at that last phrase again: "independently completed a body of novel thesis research, then wrote it up". Translate that phrase from academic jargon to industrial jargon: envisioned, launched, executed, and delivered on a difficult, novel project. If you hire a star college grad, you'll spend five or six years getting him or her sufficiently well trained and professional to pull that off. You'll beg and plead to hire any self-starter with that much experience from a competitor. That is exactly what each Ph.D. is -- only he or she isn't employed by a competitor yet.

How can any sane manager not want to hire as many of those people as possible? Wouldn't you put them in the front of the line to interview? Particularly if you found an undervalued population of them, who will make far less in a far more cutthroat environment if they follow the natural path of their career? In the "intellectual property" businesses, your raw materials are your people -- and, in ignoring those "useless" Ph.D.'s, the high-tech business has chosen to avoid outsourcing to a new market.

So, what does this have to do with Netscape's failure?

Well, Netscape eventually figured out that it was in the manhole cover business, at which point it, as a company, needed to figure out what to do to use its current facilities to compete in that business. What kind of people could it have had who had a demonstrated ability to investigate problems they didn't understand except peripherally, and make progress in solving them with limited time and resources? Obviously, the best possible source of such people would be a pool of people who'd done it before -- exactly as all those hard to manage Ph.D.'s had done.

The rot starts at the top.

We're not in the manhole business...yet

2007-06-22T06:21:00.001-07:00

A couple of weeks ago, Marc Andreesen put out a piece on hiring. I've never had a whole lot of respect for him, and the post didn't change that. (Don't get me wrong -- there is a lot of good, solid common sense in Andreesen's post, and I'll recommend it to people who are learning to hire. In fact, you should go read it. He says a lot of useful things in among the stuff I object to.)

Go on -- only my words are here, you know, so I won't even know that you're gone.

Back now? Excellent.

(While you were there, I hope you chased his link to Joel Spolsky's blog. Joel's one of my all-time favorite writers about technology, and watching him mature from Microsoft fanboy to seasoned manager over the last few years has been a real pleasure. Yes, I could have linked there myself. No, I didn't: I want you to read Andreesen's piece. )

Andreesen ticks me off by repeating one of those old mantras about Microsoft and Google and their hiring practices. He says "Most of the lore in our industry about the role of intelligence in company success comes from two stratospherically successful companies -- Microsoft, and now Google -- that are famous for hiring for intelligence." He then goes on to repeat a whole lot of conventional wisdom, some parts of which (Ph.D.'s are "the hardest people to motivate to ship commercially viable products") are merely counterfactual, and others of which (Claiming that that right answer to "Why is a manhole cover round?" is, "Who cares? Are we in the manhole business?") are worse than false.

I think those answers show why Andreesen is a serial failure in business. Contrary to a dearly-held belief in Silicon Valley, Netscape is relatively unusual in actually being pushed under by Microsoft, illegal, unethical, and slimy tactics or no. Yes, Netscape lost the browser war, but neither Google nor Adobe has gone under. Sun, despite repeated attempts at seppuku, still lives. And Apple, Steve Job's Apple, is still thriving, having been a thorn in Microsoft's side since before Microsoft even existed. The list of companies which MS has focused upon and then failed to beat is far longer than the list of companies which Bill's Behemoth has bested.

(Don't get me wrong: Microsoft's tactics were illegal, and Judge Jackson's Findings of Fact were entirely sound. They may well have helped beat Netscape, although the evidence presented in court tended to refute that claim. Mostly, IE 4.0 was a far, far better product than NS 4.0. The fact that it came with every copy of Windows 98 seemed to make no difference -- and that's also something Judge Jackson pointed out.)

In the end, though, Netscape failed because they were in the manhole cover business, and they didn't know it. Despite their pretensions, Netscape was not in the cross-platform browser business -- for the simple reason that there isn't one. Firefox is succeeding because the folks at Mozilla were "reality-based" -- they looked at the market as it was, IE quirks and all, and built a browser which competed in the real market. Opera, Firefox on Mac, Safari on Windows -- these browsers fail, and will continue to do so, until and unless they learn that lesson.

OK, so what?

The reason Microsoft interviewers asked "puzzle questions" that appeared to have no bearing on the industry was to find people who wondered about how to solve ridiculous problems that didn't appear to have anything to do with what they were supposed to work on. The best answer to the manhole cover problem was not important -- what was important was what the candidate did with the challenge of being thrown into the manhole cover business. How did he or she attempt to gauge the problem's technical requirements? Given an observed fact, how did he or she propose an explanation, and what did he or she propose as mechanisms for testing that explanation? What could he or she propose as better alternatives? What options were available to shape the market?

Netscape failed precisely because it couldn't answer that class of questions about its market. Andreesen's blindness shows that there, just as at Microsoft, the rot starts from the top.

A life full of doing evil

2007-06-16T15:32:00.000-07:00

I was joking the other day with a friend that every company I ever joined immediately turned to the dark side. This afternoon, as I set and thought, I realized how true that is.

Back in 1979, I went to work at Argonne National Laboratory -- just in time for Three Mile Island to blow up. As a result of my joining the industry, it went from the hope for America's future to the Great Satan of the environmental movement -- instant evil. In 1995, I went to work for Microsoft, and, over the next few years, watched as every bit of public credibility the company had about its practices was thrown away. Last year, after eleven years immersed in evil, I left Microsoft to go work for a company which promised to avoid evil altogether. Guess what's happening now? You got it: the mob is calling my current employer evil, too.

Is it me? Am I some kind of dark stain on the world's employee pool? I don't suppose it could be that the mob is not completely right about the risks of nuclear power, the poor quality of Microsoft software, or the privacy implications of using a certain large search engine? Naah -- it's got to be me.

I understand David Broder's call for a third party now

2007-06-10T10:45:00.000-07:00

McCain/Lieberman would be a dream ticket -- without Lieberman to egg him on, McCain might not actually bomb-bomb-bomb, bomb-bomb Iran...

There really is a Paris, Arkansas

2007-06-09T07:49:00.001-07:00

No, seriously, there is. It's about 28 miles from Havana. And this blog has about as much to do with it as it has to do with Paris Hilton -- or the Paris Hilton.

But, for all that, Paris isn't bohemian -- the nearest approximation to bohemianism is in Little Rock, or Hot Springs. And this blog is all about Bohemian Paris, Arkansas -- that is, about things which exist only in my imagination, or, perhaps, about nothing at all.

Welcome.