Entrepreneurship

The third and final class I’m blogging about is my Entrepreneurial Thinking class. You know how when you’ve got a really good idea fresh in your head, you have that sort of “fire in your gut” as the professor calls it? That’s basically how I feel all class.

I think I’ve been thinking all wrong. Most of my business ideas begin with winning the lottery. One of my ideas is for a sort of two-way radio “superstore,” one right over the NH border and one in Boston. There are all sorts of neat details and I think it’d do really well. But I’d need a few million to start each of them, after you figure in real estate, labor, and carrying a big inventory. There’s a mall I’d like to buy out, revitalize, and make a killing on, but that’s a few more million. I have a lot of ideas that I think would work very well, but almost all of them involve having a ton of capital up front.

The way to do it, I’ve found, is to spend no money. We talked about SkyMall. Anyone who’s flown will recognize their catalogs, a sort of ‘metacatalog’ of all sorts of neat stores. They spent millions starting the business up, setting up a big IT infrastructure, setting up warehouses at airports for instant delivery, and so forth. The catalog companies were selling them the merchandise at substantial discounts, and he was paying just a little bit to the airlines to carry the equipment. The article we read about them ends with him on the verge of bankruptcy. Obviously the story doesn’t end there, because they now seem successful.

So then the professor told us we could start it in class, using only the money we had with us. We’d sort of already established that the real business was acquiring new customers for the catalogs: SkyMall isn’t a retailer, they hook you up with the appropriate retailers and take a cut of the sales as commission for connecting customer and retailer. So we got that step right in class. But then he asked what was next. No one said anything, so I volunteered that we’d then have to get the airlines to cooperate. He asked me how. “I think they did it right. You pay them a small cut of your profit for each sale, and promise to pay them a modest amount if no one buys anything, to cover their costs of carrying the catalogs,” I told him. “Really?,” he asked. There was sort of a silence, since no one else saw what I had missed.

He proposed that we not pay the airlines anything. I couldn’t understand why the airlines would agree, but he explained that in short answer: passengers are stuck in a big metal tube for hours, and it’s up to the airlines to entertain them. You don’t market it as, “Will you please carry my catalog, I’ll pay you?” You market it as, “Hey, why don’t you carry my free catalog to entertain your passengers! That way you don’t have to pay anything!” And there’s no risk to them–at worst they just throw your catalogs away.

So I was reminded of one of my favorite business models ever. For all I know it’s urban legend, but I don’t think so. Some guy found that some industrial process was being left with some sort of waste that was extremely expensive for them to dispose of. They were paying through the nose for someone to dispose of it. The same guy knew that somewhere else, people were looking for that waste, because it actually had uses. (Sort of like the fruits I mentioned from negotiations class, actually?) So he charged for the disposal, but undercut the competition. And, having just been paid to take it, he then sold it to someone else. Rather than the traditional, “Buy and sell at a markup” model, he was charging at both ends. As the Guinness guys would say, brilliant!

But I guess the other aspect is a sort of “organic” growth. Some things might work if you start with millions of capital upfront, but if you can start a business from the spare change on your desk, you really don’t have much to lose. Your real asset becomes your networking capabilities, not your fundraising capabilities.

Homelessness

My second class was my “Outsiders and the Law” class. We were supposed to be discussing homelessness that class. Our professor got a group of four people to come from a homeless shelter. The director was one of them, and explained that we’d do a sort of ‘exercise’ first–of the three others, two were staff members at the shelter, and one was a resident of the shelter. We’d get to ask questions, and, at the end, guess which was the homeless person. (Something about “Spot the homeless guy!” seems insensitive, but I digress.)

The first person was a guy, probably mid- to late-twenties. He had several facial piercings, was clean-shaven and well-groomed. #2 was fairly well-dressed. #3 looked a little unkempt. He definitely fit my mental perceptions of what a homeless person would look like. Before we even started, I was positive it was #3.

One of the first questions was something like, “What’s it like to be homeless?” #1 said it could be liberating at times, but that at other times, it was no fun. (I’m heavily condensing answers.) #2 basically said it was a drag. #3 began with, “There’s nothing liberating about it.” I told you it was #3!

Another question basically asked their stories. #1 started using drugs, got hooked, and ended up being homeless. He told us, rightly somewhat proudly, that he’d been clean for 82 days. The inclusion of specific details like that made me reconsider a bit. But then #2 went and gave specific details galore. She was a victim of domestic violence. Her voice was really quiet and melancholy, and she looked at the floor the whole time she talked. She lost her job and ran out of friends to stay with, and her therapist referred her to the shelter. #3 was disabled, and one day his wife decided she didn’t want to be with him anymore. He left with not much but the clothes on his back, no source of income, and he couldn’t work. He receives SSDI, but not nearly enough to pay for an apartment.

All three basically said they were there until they got things together. None intended to stay there a moment longer than necessary, but with no job and no home, there wasn’t much else to do but stay in the shelter until they could land a job and get an apartment.

Someone else asked about the role of families, and why they weren’t with family. #1 praised his family, and said they’d really been there for him, but that they eventually stopped letting him into the house when he was on drugs (most of the time). And when he started stealing from them to buy more drugs, they kicked him out for good. #2 gave a somewhat vague answer, but I think her father was never there for her, and her mother supported her somewhat but usually wouldn’t let her stay over. Just the way she spoke screamed of utter despair. I started to think that it was #2, not #1. #3 was in his 50s, so moving in with his parents wasn’t exactly an option.

I asked about how they spent their days. #1 said that there were some “day programs” the shelter put on, but that there really wasn’t much to do. #2 has a part-time job at Dunkin’ Donuts, but talked about how hard it is to find jobs. You need to list a phone number on the application, which is the payphone at the shelter. It’s not answered too reliably. And everyone knows the address of the homeless shelter, so even though there’s no, “Are you homeless?” question on the application, most employers know it. Worst of all, she said, the shelter has a 4:30 curfew–if you’re not in by 4:30 p.m., you can’t come in. This doesn’t work too well with her job, and she’s also scorned by coworkers. (“She can’t work any later because she’s homeless!”) #3 talked about how he’ll spend some time in the library, and, with all the heat, he’d spend a fair amount of time at the beach, but he noted that it’s not nearly as glamorous as it sounds–with no money and nothing to do, spending time at the beach was really pretty miserable.

When we guessed, #1 received 1 vote. #2 received about 50%, including me. #3 received the other 50% or so.

It was #1. He’s been clean for 82 days, but he doesn’t have much to his name, and he describes himself as still healing. #2 and #3 both work at the shelter, but noted that their stories are a mix of their own past experiences and experiences of those they know.

It was really a good way to break some stereotypes, namely:

  • The homeless person “looked” the least homeless…
  • There’s an image of homeless people as burnouts who stay homeless forever. #1 was working hard to get his life back together, and seemingly making great strides.
  • The perception of the homeless as all being alcoholics isn’t that accurate. Apparently about 30% self-identify as alcoholics, but we saw a myriad of reasons for homelessness.
  • Sometimes the homeless hold jobs, but it’s hard for them to get those jobs, and having a job doesn’t automatically buy them a house.

Negotiations

This week I’ve had four classes so far. In three of them, I’ve learned something that’s kind of profound in a way. So consider this the first of three posts in a series.

One of my classes is Negotiations. It’s part of my management major, but just talking, a lot of people think it should be a required course, since it’s a skill that’s not at all limited to management majors. I’ve had the professor before. He’s one of those great professors who keeps you engaged and yet doesn’t give much work. And despite the fact that there’s not a lot of work, you really learn a lot. So it’s sort of a win-win class.

When we met Monday night, he talked a bit about negotiations, and then gave us an exercise. We split up into teams of two. The assignment was that we were both bidding on a very rare fruit. I was part of a UN delegation, and we were extracting part of the prune to help grow food in infertile regions. My work would save 20,000 lives, and I was given a budget of $2 million to get these prunes.

It seemed like I had a slam dunk case. My strategy was to not even mention my price, just that I was going to save so many lives. How could he let all those people die?

He represented a major pharmaceutical company. So my strategy shifted a little–he probably wanted to research what was in them. So, in return for him not bidding on them, I’d give him a few of the fruits to study for free. That still left us with most of them, and would satisfy both of our needs.

It turns out that he needed all of them, for a new medication that would save lives, too, by preventing heart attacks. There goes my humanitarian appeal. And his budget? $5 million.

I tried to talk him into splitting the lot. I’d buy 2/7 and he’d buy the remaining 5/7, with our $7 million combined. But he had no incentive to do it–his boss wanted as many as they could get. (He also wouldn’t take a bribe.)

We went back to class and reported on our findings. A few groups, like us, couldn’t find any compromise. A few found bizarre compromises. But only one group found the ‘secret’ of the exercise.

We didn’t need the same part of the fruit. What was a ‘waste’ product to my firm was exactly what the pharmaceutical company needed. We even talked a little bit about our goals, but never got into enough details on what part of the fruit we needed. We just assumed we needed the whole fruit. I won’t be lame and try to point out the morals, but suffice it to say there are several.

Rain

Do you think my professors will accept, “It’s pouring out and my umbrella is at home” as an excused absence?

Have humans always been reluctant to go outside in the rain, or are we just reaching new levels of absurdity?

P.S. – I’m not actually going to skip all my classes because it’s raining out. I’d like to, but I won’t.

The Magic SysRq Key

Today I came across (via Digg) some stuff on Linux’s Magic SysRq key. I’m yet to be able to try it, but it’s apparently a low-level way of controlling the system when it goes haywire. You do need to enable it in the kernel, but I’m under the impression it’s enabled by default*.

I’ve been running Compiz on XGL, both of which are experimental software. Certain obscure combinations of activities will freeze it, and the system appears locked up. I strongly suspect that it’s not ‘truly’ frozen, though, but since it’s my user interface, short of pulling up another computer and remotely logging in, there’s not much I can do. Traditionally I just turn the machine off. The “REISUB” method will hopefully let you ‘cleanly’ shut down the system, by getting the kernel to terminate all processes, unmount disks, and then reboot.

It looks like there’s more, though… You could regain control of the keyboard and then just kill everything on the current VT, in theory, at least, letting you just restart X.

* You can check if it’s enabled by checking for a /proc/sys/kernel/sysrq, which should contain “1” if enabled. This page also covers making the command ‘sticky’ so you’re not in the awkward situation of trying hold down Alt + SysRq + another key. (And in my case, SysRq is a ‘function’ key, so it’d be Alt+SysRq+Fn+r, for example.)

Intersections

I almost bought* an office park in Chicago (I think?), and noticed that it was on the intersection of High Street and Gay Street. It seemed ripe for jokes.

WPD is currently at what seems to be a major single-car accident at the intersection of High Street & Moody Street. For added amusement, they’re just referring to it as “High and Moody.”

* If I had won the lottery.

The Perfect Radio

So as anyone who’s seen me in person will surely know, I have a lot of radios. I’ve sold a few lately, but I’ve owned a wide variety. I have the VX-2R, one of the smallest radios ever produced. It’s got an incredible frequency range, too. I have the ASTRO Saber, one of the biggest radios ever made, capable of APCO25 digital voice, trunking, and MDC ID decoding. I’ve owned police scanners and mid-range radios.

Perhaps unsurprisingly, I’ve found that there are big differences between a tiny, $200 radio and a gigantic, $5,000-new ($250 on eBay a decade later) radio. But the huge expensive one doesn’t always win out. If I were designing a radio (which I’d like to!), here’s what it would be like….

  • Medium size. The tiny radio is handy, but it ‘feels’ crappy just because it’s small. The Saber and ASTRO Saber feel like some of the most solid radios ever built, but they’re almost comically large. I want something in the middle: solid, with controls big enough to use, but something that I can put in my pocket.
  • A good speaker. I can turn my Saber / ASTRO Saber up halfway and hear it more or less throughout the house. By comparison, if it’s noisy, I can’t hear my VX-2 unless it’s pressed against my ear. And turned up all the way, it’s heavily distorted. This is probably true of the Saber-based radios, but you’d probably blow out your eardrums before you noticed the distortion.
    • The sound quality is just as important as volume. The Motorola radios have a nice ‘deep’ sound, whereas most other radios sound somewhat tinny.
    • It probably costs $5 more to include the best speaker ever put in a radio in. I’d gladly pay $50 to upgrade to a radio with substantially better audio quality. Why don’t more people include good speakers?!
  • Notch filter or similar. There’s lots of extraneous noise on most signals. It’s actually pretty easy to filter it out, and ‘base’ HF ham rigs have been very good at it for a long time. Something as simple as a notch filter would eliminate a lot of the nuisance noises and make listening much more pleasant. (You could do a lot with DSP and make audio sound much better, but someone should at least do the minimum…) This is also the place to mention that I’d really like it if your radio would do some volume normalization.
  • A good microphone. For normal ham radio stuff, this doesn’t really matter and any 19-cent microphone can be soldered in and work just fine. But ‘real-world’ stuff doesn’t work that way.
    • At work, I can almost never hear the mechanics when they’re out back trying to talk to me. They could be telling me that they’d like me to ask the snack bar to cook them a hamburger, or they could be telling me to shut down the machine because they have their arm stuck in the gears. I think a good directional microphone would be a big help here, in only picking up what’s directly in front of the radio and not the (very loud!) ambient noise. (I remain convinced that another microphone on the back of the radio, ‘subtracted’ from what’s coming into the front micrphone, could produce amazing results.)
    • Campus Police responds to a lot of fire alarm activations. The fire alarms are extremely loud, to the point that it’s literally painful if you’re there in person. You can only hear what they’re saying between the buzzing of the alarm, and even then it’s hard because it echoes. I don’t know that this can be solved easily, but I’m sure a good design could at least help.
  • A nice big screen. In ham radio, giving me eight characters is considered amazing. Many commercial radios just give you a numbered readout of what channel you’re on. The ASTRO Saber has an incredible 14 (I think…) characters. But even then, fitting “Boston PD Channel 4 – Area ‘E’ – West Roxbury, Jamaica Plain, and Hyde Park” is a chore. You end up with something like”BPD4EWRX-JP-HP,” which is not that helpful until you get very familiar with it. At which point you’ll probably know what Channel 4 means without the label anyway.
  • A sane way of organizing channels. Motorola gives you 16 ‘zones’ of 16 channels each for 256(ish) channels. (This is technically not true but it’s a practical limit.) Many ham radios just give you 200 channels. Some of the better ones (and some scanners) let you use ‘banks,’ which are sort of like ‘folders’ of channels. But there are almost always limits: a bank can almost always store 40-50 channels max, and you usually can’t assign a channel to more than one bank. In my mind, it’s idiotic to still have these limits. What if I want 60 channels in one bank? What if I want to have 72 banks? What if I want Campus Police in a “School” bank and a “Waltham” bank and a “Waltham – Active Emergency” bank?
  • Nuisance Delete! Motorola got this a long time ago, but until I bought the ASTRO Saber, I’d never heard of it since it seems like nothing else supports it. When you’re scanning a range of channels (which is, you know, what scanners are for), there’ll sometimes be one that you don’t want to scan. Maybe you have the police, fire, and the local ham repeater, and there’s a big fire going on. The police are scrambling to get someone out of the building while the fire department is attacking the flames. And then the scanner stops as someone on the local ham repeater discusses how he doesn’t really care for Taco Bell that much these days. Most scanners have a “Lockout” which will ‘permanently’ delete the channel from the list of channels it scans. Nuisance Delete is temporary and gets wiped out once you stop the scanning ‘session.’
  • Recording! Really, I can’t believe there’s almost nothing on the market that does this. I want to leave the radio ‘off’ (let’s call it “Standby Mode’) on the charger all day. And when I see five police cars go speeding by, I want to jump out of my seat, pick up the radio, and hear what the police dispatcher said two minute ago. I’m not aware of any radio that will do this. The thing is, voice takes up very little space. 64kbps MP3 would be plenty. Probably overkill. And a 1GB flash storage card is about $20. And I bet you could get it for a quarter of that if you were buying them OEM to embed as opposed to a consumer buying SD card. You could store weeks of audio. And how many times are you listening but you miss a key detail. (When I’m listening to try to figure out why the emergency vehicles went by, I’ll almost always hear, “We’ll be on scene with a –” “Hey Matt, do you know what’s up with that fire truck?”) It’d be great to just replay it.
  • Digital mode support. This is kind of vague, and could involve a lot of licensing / royalties. But public safety (law enforcement in particular) is very quickly moving to the APCO Project 25 Common Air Interface (generally “IMBE,” “P25,” or Motorola ASTRO). There are three scanners, out of probably 50, that do this mode, and their audio quality doesn’t compare to the genuine radios. There’s also trunking which is very common in cities. The ability to monitor paging networks (POCSAG/FLEX) is handy, but raises a lot of legal issues. (Intercepting other peoples’ pages is explicitly illegal and it’d be hard to design a radio that could decode the protocols without allowing people to see other peoples’ pages.) There are other experimental digital modes, too.
    • The best solution, IMHO, is to make the device run Linux (or any other common embedded OS) and release an SDK so people can write their own digital modes.
  • Muting of various junk, such as digital modes. These days I’m using the VX-2, and I really miss the ASTRO Saber’s “DOS [Data-Operated Squelch] Muting,” which would detect MDC1200 data traffic and mute the speaker. All too often what comes out of the speaker isn’t voice, but just various noise that gets transmitted over the air. It’s really not that hard to detect it.
  • A good battery. I should be able to use it all day, including periodic transmission, without recharging. (Motorola famously offers a 4,000 mAh battery for their flagship line of radios.)
  • A very readable display. Not just big as I said earlier. One of my radios is hard to see if I look down at the LCD. Another is hard to see if I look up at the display. Another is kind of washed out if the backlight is on. Anything with a graphic LCD (very few radios) is almost impossible to see in direct sunlight. (Frankly, I’m very impressed with e-ink displays like the Sony Reader, and would be obliged to buy a radio, no matter the price, that had one as an LCD.)
  • A frequency counter, to detect what frequency something is on. Generally you have to buy an external device to do this. The VX-2 is (as far as I know) the only radio ever made with a neat feature which is basically a ‘ghetto’ frequency counter: it’ll kick in a 100 dB attenuator and see what frequencies it can find, good for locating very close transmitters. But it’s not a real frequency counter. It seems pretty obvious to me that frequency counters and radios are generally used together, so it’s really kind of surprising that so few people have thought to blend the two.
  • Durability. I drop things. Public safety radios get it ten times worse. You’ll read stories about people dropping their high-end Motorola radios into the ocean, or someone dropping it and then backing over it in the fire truck. And they pick it up, dust it off, and realize it sustained almost no damage. There seems to be more of a focus on making things cheap than on making them durable these days, though. But we want durability! At the very least, I should be able to stand on top of the machines at work and drop your radio eight feet (or so) onto the concrete floor and just have to put the battery back on.
  • Scratch-resistance. Especially the LCD. For some reasons radios don’t seem to scratch as bad as cell phones and iPods, maybe because not many people put their radio in a pocket with their keys. But a lot of watches have faces made out of things like sapphire or crystal that are basically impossible to scratch. I’m sure this adds to the cost (watches like this aren’t exactly cheap), but I also bet much of that stuff could be grown in a lab. People do care when it’s for a wedding ring, but I’m pretty certain no one cares when it’s for a radio display covering.
  • Intuitive controls. I shouldn’t have to press ‘shift’ and then something else to perform basic functions. When it’s -10 out and my hands are almost numb, I should still be able to operate the radio. (And when I wise up and put gloves on, I should still be able to use it.)
  • Wide frequency range. I’d be interested in working the 6-meter ham band (50-54 MHz), the 2-meter ham band (144-148 MHz), VHF ‘commercial’ (136-174), the 220 MHz ham band, the 70 centimeter ham band (420-450 MHz), the enormous UHF commercial split (403-520 MHz), the 700-800 MHz public safety band, and the 900 MHz ham band. If I were to cover all of these, I’d probably need eleven radios. My little VX-2, smaller than a deck of cards, will happily receive all of them. There’s been one commercial radio that would let you transmit on more than one ham band, and that was discontinued a long time ago. (Many ham radios support several bands.)
  • Simple programming. I like FPP (front-panel programmable) radios, but there are legal issues on commercial radios that usually prevent this. It’s also handy to be able to program radios on a computer, especially if you have a lot of channels or want to program a lot of radios the same way. (Aside: why can’t radios share data over the air? There could be an ‘over-the-air cloning’ mode that uses something like spread spectrum to avoid interference, which could make programming a fleet of radios much easier.) And it’s also neat to share frequencies with other radio users. As far as the computer programming, though…
    • The interface has to be intuitive. Motorola’s RSS, somewhat of a standard (until their Windows-based CPS replaced it), is probably the least intuitive piece of software I’ve ever used. Programming the ASTRO Saber, I had a huge sheet of paper. On one screen, I’d input a “personality,” which was the frequency for a given channel. Then I’d have to map that personality into a zone, hence the big sheet of paper keeping track of what went there.
    • The software should be free, or at least included with the radio. (It should really be open-source, in my opinion, so people can enhance it.)
    • Programming should be quick. USB 2 allows 480 Mbps. Why are people still designing connectors that use serial ports? The last few computers I’ve owned haven’t even had a serial port. Not only is it obscure, but it’s so slow!
    • As long as the device is Linux-based, why not just have an /etc/channels.xml file or something? Using an open standard like XML, and making it just a file that anything can read/write over USB, would make programming ten times easier.
  • GPS. Ham radio has APRS. And frankly, I’m very surprised that there’s no public safety equivalent, especially as they all go digital. (Random aside: LTR trunking uses “subaudible data” to pass the relevant control information: they pass the data in the ‘audio’ range outside of human hearing, so it’s there as part of a signal but not reproduce as audio, so it’s basically ‘hidden’ in the analog voice signal. This is ingenious. In the 20 years or so since that began, I’m surprised that no one has ever thought to embed PTT-ID/ANI (e.g., a way of identifying which radio is which) data that way. And now you could embed GPS coordinates that way… But I’ve still never heard of it even being attempted.)
  • Tones and a vibrate function. Cheap “bubble-pack” radios have a “call” function that transmits a ringing sound. It’s actually very useful at work for getting one’s attention. High-end radios have things like “Private Call” that send a digital signal to a particular radio telling them to sound a bell to get the user’s attention. This is also handy. A vibrate function on phones is common for quiet areas but many will agree that the vibrate feature is actually most useful in very loud areas where you’ll rarely even hear your phone ring. I want the same on a radio.
  • Remote control. This isn’t useful for individuals, but is actually fairly common already in fleets of radios. Some existing uses:
    • Motorola’s OTACS: Over-The-Air Channel Steering. You could want all the firefighters at a given call to switch over to a fireground channel. Why not let the firefighters tend to the fire instead of their radios, and just send their radios a command to switch over?
    • OTAR: You can push out a new encryption key to all the radios in a fleet. (As inherently insecure as this sounded at first, it’s actually a complex process that’s incredibly secure.) This is important since, in a really secure setup, the key should be rotated every few days. It’s not practical to pull in all the radios for reprogramming every few days.
    • Remote inhibit. When a radio goes missing, dispatch can send an inhibit/stun command, which basically renders the radio a brick: a brick that silently keeps listening for an uninhibit/restore command. This means that if someone steals a radio, and it’s detected, they won’t be able to monitor you, much less interfere.
    • Remote transmit. Usually in response to an officer transmitting an emergency call or just not responding. Dispatch can send a command to the radio instructing it to begin transmitting audio, so they can hear what’s going on.
    • General reprogramming: I’ve never heard of this! It could be extraordinarily useful, though!
    • This whole thing needs to be encrypted/authenticated. The current implementations are not, which means that anyone with access to a dispatch console (rare, but they show up on eBay periodically) could, say, start sending inhibit commands to radios in the field. This could be really, really bad if it ever fell into nefarious hands.
  • A clock. Just a simple clock shown on the radio. (For bonus points, the radio should be capable of receiving the time over the air anyway from various atomic clocks, so you could have a very accurate clock if you were willing to take the time to program it.)
  • RSSI: I want to be able to see how strong the received signal is. Almost all ham radios do this. Very few commercial radios do. (RSSI stands for “Received Signal Strength Indicator.” Hams generally call it an S-Meter.)
  • Text messaging. A lot of departments will give their officers pagers, too, to send supplemental information. Why not just do it over the air in text form? Some of the newest Motorola radios support this, actually.

Police: A Lifelong Career

I may have posted this on the old site, but I don’t remember…  Taking my Forensic Science class here, taught by a retired detective, reminded me of the thought.

I don’t think I’ve ever heard of a police officer who has changed jobs. I’ve heard of military people ending up in office jobs and cooks becoming teachers and janitors becoming construction workers, but I’ve never, ever heard of a police officer who become something else.

I’m sure it happens sometimes, but it seems to me that it happens much less frequently. And I think that’s a good sign of when you know a job is something you might like: when no one ever stops doing it until they’re forced to. Of course I haven’t interviewed thousands of people to develop this theory, so it could very easily be inaccurate.

But it’s also supported by the fact that policework is one of those things people always seem to love talking about. I’m not even sure what one of my uncles does. He works in the financial industry and doesn’t seem to like his job. He’s never said, “I have to tell you what happened last week!” But run into a retired cop and they could talk all day about their time on the job. Again, I’m drawing on limited experience talking to retired cops, so I could be wrong. But I’m not so sure I am.

I wonder if there’s a place that compiles statistics on this. I’d be interested to see what other jobs have low career-change numbers. Which are the highest? Waiters? Does that count, though? Not many people expect that to become a career. Is it linked to training? Being a police officer takes a lot of specialized training that won’t help you at all if you decide to become, say, an accountant. But being an accountant takes lots of specialized training, too. Do accountants have mid-life crises and become construction workers often?

Mailserver – Funambol?

I still have no mailserver running here. (I’m now forwarding all my other accounts through GMail.) I’ve wanted for a while to set one up, but it has to be good.

I just came across Funambol, which is apparently an enterprise-grade server package for syncing mail, PIM data, etc. with mobile clients after installing a client-side package on your mobile device (Palm, Blackberry, and Windows Mobile devices, as well as generic Java ones, are all supported.)

And then, of course, you’d want Roundcube for webmail. It looks good, lets you search your messages, and doesn’t log you out after you finish composing a lengthy e-mail. (Are you listening, OWA?)

Oh, and DSpam for Bayesian spam filtering with a good web interface.

What caused this whole thought process was someone asking if there existed a service that basically offered what I’ve just described. He stated he wouldn’t want to pay more than $10 per mailbox. (Hosting six mailboxes at that rate, I’d be making a profit!)

My Room

Continuing the tradition of illustrated posts, here’s a few photos of where I’m living this year. I’m in Falcone, which is located at the foot of all the academic buildings. Freshman year I walked 96 steps to get to class. (I counted, trust me.) And that was just if it were on the ground floor. I’d have to go up as many as four more stories. This year I have to climb six steps and I’m there.

Here’s the view out my window:

View out my window

The brick building on the left is Jennison, one of the busier academic buildings. Also visible is the library’s clocktower.

Here’s a quick interior shot, of mediocre quality:
My room

And, facing the other way:
My room

I also took a few shots the other day, on a very clear day, using my relatively-new polarizing filter on my camera. The effect is a little too much, I think…

Library

I think this one might become my new wallpaper:
Bentley's Clocktower

And illustrating a polarizer taken to the extreme… Straight out of the camera, rotating being the only post-processing done:

Library Clocktower