Turtles all the way down

When I read it a few years ago, William Dement’s The Promise of Sleep shaped my understanding of the nature of human sleep.

In a nutshell, the take-homes were: Most or all people need right around 8 hours of sleep a night, sleep disorders can lead to various physical and mental ailments, getting less than your needed amount of sleep puts you at a sleep deficit that will you’ll need to repay, and no one knows the long term impact of not getting enough sleep.

For the last few years I’ve been averaging about 6-7 hours per night. When uncaffienated, I’m tired, which is a sign that I don’t get enough sleep. My sleep deficit is somewhere in the thousands of hours. I’m not sure if this makes a difference or not in the long run. But I can’t seem to conform my sleep patterns to what society expects.

Naturally, I’m busy. Admittedly at my own choosing, but I’ve got my main job, a handful of small side-projects, and some hobbies. Of course, there’s the family, owning and maintaining a house, and all that goes with that. There aren’t enough hours in the day and every day I leave stuff undone.

After so many years of not being able to get enough sleep, now my body can’t seem to get enough sleep. I have trouble sleeping more than 5-6 hours at a clip and am often up in the middle of the night. On one hand, I use these extra hours to get things done. On the other I’m not sure what this behavior does to my productivity or my long term health.

Additionally, I have a strong diurnal cycle. I get sleepy around the same times every day…from 11pm to about 5am and again from 1pm to 3pm. Within those windows I can sleep without too much trouble. Outside of those windows I usually can’t sleep at all, even if I’m tired. If I wake up at 4:45am, I just get up because I know that the chance of falling back to sleep is pretty small. But I love mid-day naps and would take one every day if I could.

Contrast this to a friend of mine who can sleep 9-10 hours each night but can’t nap. She can be woken up 3-4 times in the middle of the night and fall right back to sleep. Wake me up in the middle of the night and I’ll be up for two hours. This made the first few years of being a parent rather challenging.

Travel gets interesting as well. In Europe or Asia, I get nailed with jet lag. I can sleep in chunks of 3-4 hours, maybe 5 at most, but I’m always up in the middle of the night. Even if I’m dog-tired, I just can’t sleep. It may be 3am in the UK but my body thinks its still 9pm and is not ready to wind down.

Another observation is that TV is bad for sleeping. If I put the TV on in the night, I’ll stay up an extra couple of hours watching. On the other hand, books are good for sleeping. I start reading and I’m usually out pretty quickly.

My ideal day? Up at 5am, in the gym by 6, out and showered by 7:30, off to work, lunch around noon, nap from 1-3pm, work until about 7, head home and get about 3 hours of family / free time.

Remember the 80’s? Me neither, but something interesting started happening back then. We had mainframes, mini’s, oodles of UNIX-variants on workstations, and these new-fangled PCs. Different platforms, each with different development environments. You know, different compilers, libraries, APIs, tools chains, and so on.

By the late 90’s we had WIN32 (that’s the Microsoft environment), Linux, and a bunch of also-rans. Today, in the server space, its pretty much the same, Microsoft and Linux. Now, I’ll be honest, there are different variations of each platform. MS Server 2003 is different from NT Server and there are several Linux distributions, but the differences between those variations are relatively minor compared to the differences between completely disparate platforms.

On the client side, however, its almost all Microsoft.

I’m not a Microsoft fan, either as a user or developer. I’ve always found WIN32 variants to be a clumsy environment compared to something like Debian Linux. However from a business point of view, having a common development platform is huge. If I want to write a client-side application, I write it on WIN32 and I’ll get the biggest possible total addressable market. On the server side, I’d prefer to make it a web application if I can and then integrate it with Apache or IIS. This has narrowed the choices of platforms that an application developer has, which is a good thing.

Switch over to the mobile client platform world. Today you have J2ME, BREW, Symbian, Microsoft, Linux, and an obscure Korean BREW knockoff called WIPI. Each has their variants. Symbian and Microsoft seem to have the fewest intra-platform deviations, BREW isn’t too bad, and WIPI is not widely spread enough to consider. However J2ME (that’s Java for mobile devices) varies greatly between manufacturers and sometimes even within manufacturers. Remember the jokes about Java circa 1998, “write once, debug everywhere”? That’s J2ME today. And Linux is the native OS behind a bunch of phones from a bunch of manufacturers, but each manufacturer has their own APIs and essentially their own Linux platform.

The mobile application development world today is where the wired application development world was in the 80’s. It would radically simplify mobile application engineering if there were only one or two ubiquitous platforms. To the end user this would mean better applications, that run on more devices hitting the market faster. Arguably the most successful mobile application developer is RIM and they not only control the platform (which is their proprietary version of J2ME) but are also a manufacturer.

The downside of course is whoever controls the platform has a near-monopoly ala Microsoft with PCs. The PC application market and the web application market boomed because the platforms were ubiquitous and anybody could develop on them. This needs to happen in the mobile space. There’s room for two, maybe three, platforms. That’s it.

If it takes 20 years to get there, we’re all in for a lot of pain.

I’ve had pretty bad luck in this area.

My first external was an 80 GB Fantom. Worked really well for a while, but died in a little over a year. My next was a Buffalo 120 GB which lasted several years and died recently. However I had already moved on to a pair of Buffalo 250 GB drives. One croaked in less than 30 days, so I took it back and replaced it with a Maxtor 300 GB. The other Buffalo soon crapped out, so I replaced it with a Maxtor 320 GB. When the Maxtor 300 GB died, I replaced it with a Western Digital 500 GB. I recently ran out of space on the Maxtor 320 GB, so today I bought another Western Digital 500 GB.

I’ll estimate that I get about 9-12 months out of an external drive. On the other hand, my internal hard drives have lasted through years and years of heavy use (knocking on wood…). I’ve read up a little on this and have some theories:

• Overheating. Many external drive casings either don’t have fans (hi there, Buffalo) or trap too much heat. Heat is a major drive killer.
• Movement. Since they’re external and in a box, I tend to move them around and bump them more than I should.
• Humidity. My externals are in my basement office. Despite the dehumidifer, it is the dampest place in the house.
• Flaky controllers. In several cases, the drive itself didn’t die but the drive controller did. This is good to know if you don’t have a backup. Just take out the drive and pop it into a new enclosure.
• Always on. Leaving drives on all the time reduces their lifespan. For the last 6 months or so I’ve been turning them off except to do backups and so far so good.
• Bad manufacturers. Western digital and Seagate seem to have the best reputations. All other manufacturers are spotty.
• Bad juju. Maybe I just bought a series of bad drives. I could be cursed.

Perhaps it is some combination of the above. However, time and technology are in my favor. Drivers are getting bigger and cheaper. With my pair of 500 GB drives, I won’t run out of space quickly. I use them mostly to store home videos and MP3s, and my storage needs increase at a rate around 70 GB per year. I’ve got two years left in these drives, and given the pace of the storage market, I’ll be able to buy a pair of 1 TB drives for around $200 each by this time next year.

So I’m reluctantly accepting that external hard drives are disposable because they are relatively cheap and quickly obsoleted.

I’ve noticed this already. Of my two lecturing profs, one is very interested in the exact definitions of terms. For example, assault is, “the overt and intentional causing of someone to be reasonably aware that they will suffer immediate harm, where the harm can be carried out and there is no consent or privilege.” Or something like that.

The other doesn’t spend much time on defining anything, but instead is heavily focused on procedure. “The Trial Court ruled this, but the Appellate Court rule that. Then the Supreme Court held that both lower courts were dumber than a pantry full of twinkie wrappers and ruled something completely different.”

Details vs. process. Structure vs. onion peeling. Short vs. tall. One points out errors in the textbook, the other wrote the textbook. We’ll see where this all leads.

So now I know why its more fun to argue with liberals…

Exploring the neurobiology of politics, scientists have found that liberals tolerate ambiguity and conflict better than conservatives because of how their brains work.

Previous psychological studies have found that conservatives tend to be more structured and persistent in their judgments whereas liberals are more open to new experiences. The latest study found those traits are not confined to political situations but also influence everyday decisions.

Analyzing the data, Sulloway said liberals were 4.9 times as likely as conservatives to show activity in the brain circuits that deal with conflicts, and 2.2 times as likely to score in the top half of the distribution for accuracy.

Based on the results, he said, liberals could be expected to more readily accept new social, scientific or religious ideas.

“There is ample data from the history of science showing that social and political liberals indeed do tend to support major revolutions in science,” said Sulloway, who has written about the history of science and has studied behavioral differences between conservatives and liberals.

The first thing I’m doing for law school that has any impact on my grades at all is to write a memo. It is an analysis of a potential client’s situation and whether or not she has a case for intentional infliction of emotional distress (she does).

I did my first draft last weekend. Wasn’t too hard. Except that there is a canonical format such an argument should be in, and mine wasn’t even close. And my citations were all screwed up.

I’ve spent a good chunk of time over the last few days cleaning it up. As of yesterday morning it was in good shape. But, the discussion section, which is the meat of the argument, can be no longer than seven pages. Mine was ten. I hacked and slashed down to eight, tossing a couple of paragraphs that I really liked but were probably overkill. Then I resorted to go through the memo, sentence by sentence, line by line, imposing economical language on it.

As of a couple of hours ago, it’s seven pages. A couple of proofs later today and tomorrow morning, and it should be done. I hope.

Of course, no one in their right mind would require such a strict page limit on a real-world memo. The point is to force the analysis of one’s language and content via the page limit. I think it worked, though I also believe that my memo would be better with another page or two.

Don’t know what an Erdos-Bacon Number is? Look it up. I’ll wait.

Ok, now you know as much as I do. Almost. Today I discovered my Erdos-Bacon Number: 8.

Alright, I’m technically cheating. My Erdos Number of 4 is legit: Me -> Srini Tridandapani -> Arun Somani -> David Avis -> Paul Erdos.

However you can only have a Bacon Number if you’ve appeared in a film. If I ignore that requirement for a moment and consider that a friend of mine is an acquaintance of Moon Unit Zappa, my Bacon Number is 4. So my Virtual Erdos-Bacon Number is 8. If I get cast in a Kevin Bacon movie, I will be tied for the second-lowest known Erdos-Bacon Number of 5.

So I am hereby offering to appear for free in any Kevin Bacon movie. Did you hear that, Hollywood?

…seem to go together like peanut butter and jelly. Google the phrase and you’ll find hundreds of lawyers who would be happy to try to get you out of a DUI charge, or sue the drunk S.O.B. who hit you.

During orientation week we were given a serious lecture about the dangers of becoming a lawyer. It had nothing to do with lawyer jokes or friends calling for free advice. We were given statistics on the abnormally high rates of alcoholism amongst lawyers. Apparently the problem is so bad, a non-profit organization was started to provide anonymous help to those in need, and it has chapters in most major cities.

This sort of revelation makes me think twice about my new career.

The next night the school throws a three-hour reception for new students with…you guessed it…free drinks. The next week the Dean invites us to a local bar for…wait for it…free drinks.

It’s all starting to make sense now.

When teaching at Northwestern, I used the Socratic Method without knowing I did so. Even in science and engineering lectures, where the facts are the facts, there can be more than one way to approach and solve a problem. Furthermore, not all issues are cut and dry, so a degree of discussion is welcome. If nothing else, talking for three hours straight gets boring for me as lecturer and I can only imagine its affect on my audience. So breaking it up a bit while getting students to think on their feet can’t be a bad thing.

In my first Criminal Law class, I read the assigned chapter and cases twice. I highlighted the key passages in the book and had a good grasp of the main issues at hand…you know the ones with Constitutional implications.

I was probably the least prepared student in class.

The professor dove into case history, procedure and precedents…who did what how and why. The individuals sitting on either side of me were prepared with detailed brief of each case. I had a few sentences highlighted in my casebook. When called upon, my fellow students recited facts of the cases from memory (either their own or their PC’s as displayed on their screens). I madly flipped pages to find the right paragraph.

So I’ve got to start briefing all the cases before class. Briefs are elaborate cheat sheets that allow summarize the key point of a case in an organized format. I’ve also got to un-teach myself 15 years of purposely not memorizing anything.

This post is not about underwear.

My first brief-writing experience was painful but I’ve survived it. During orientation week we were assigned to write six briefs. Never having done this before, I had no idea where to start. My first brief was seven pages and growing when I finally realized that it wasn’t very brief. The second two were shorter, and I finally got the last three down to an almost-acceptable two to three pages.

Ironically, we were informed that there is no one right way to do a brief, but we’d be graded on some very specific aspects of our briefs. Hmmm.