It's 7 am. Apparently I've stayed up all night. I can see the sky just starting to lighten. I just stayed up all night for no reason. I have a final on Friday around noon, so if I go to be at noon today, I'll wake up at 8pm, study most of the night, take my final, and collapse around 5pm. This is shaping up nicely, actually. I also have the false alertness of insomnia.

Except that I'm going to a birthday party Friday night. Hmmmm.

Mood: sleepy

Before I go to bed, entirely too late (2:30am!), I must share Telescopic Text.

None of the “Mood” choices match my actual mood. But then I can't really describe it anyway. Sort of ignored/abandoned because everyone's asleep. Tired because, well, it's late. I've got plans which have been partly thrown awry by staying up so late. Hopeful about certain things, pensive about others. I suppose I'll put down “composite”.

(Has anyone ever noticed that Myspace has the exact same list of moods as LiveJournal originally came up with?)

As I sit here in the library, listening to the subtle music of the floor being abraded, I write. I have nothing better to do while waiting for 300MiB of data to copy to my bulla (rfriel knows what I'm talking about, so ask him).

Some people, most notably tsukiyomijapan, are always whining about how the Metro system should have stored-value cards so that users don't have to carry about change. I agree, but there are several technical problems with that sort of thing.

In order for a stored-value card to work, the value must be stored. (duh) Early stored-value systems (ATMs in the 1960s) kept the balance on the card and phoned in updates all at once, at night, because phone charges were expensive. Of course, this allows a user to just use a magcard thingy to change the number encoded on his card and take all the fare he wants.

Modern subway systems work well with stored-value systems because each ticket booth can be in constant (or at least frequent) contact with the central database, stopping this sort of fraud.

It is possible to have a copy of the master DB on each bus, in a computer under the driver's seat, but that would be a nightmare. Somebody would have to maintain around six thousand databases. Frightening indeed.

So somehow the buses must be in constant contact with the central database, and the cards can only know what their account numbers are. But how are they going to stay in contact? Packet radio sounds like the only option. You have six thousand buses, all competing for airspace. Sounds like more fun, but it would be doable and all that. (And what about reception problems? Metro goes all over the county, so they would need a very powerful transmitter on each bus, or else a massive repeater network.)

It would be very expensive, though. All the buses would have to be upgraded and a new system would have to be put in place. Worst of all, it would be very confusing for some buses to have the SV system and others not to, so it would have to be done all at once.

I'm going to post the contents of my poetry journal in the future.

Here's what I wrote yesterday:

NO. 5 XBR SYS TBL RCDR CARD
Memento of an earlier age,
When something goes wrong
in a 5 Crossbar,
it spits one out.
Manila cardboard
cheap grey ink
dozens of meaningful perforations.

It's the chad--the holes--that count.
The writing just shows
which hole
is which
relay
circuit
selector.
The machine would do fine with blank paper.

What happens when you run out?
They aren't printing any more.
Will it matter by then?

This is the first in a possible series of intermittent blogs about my experience at Microsoft as a High School Intern. I will be working today, June 28th, through September 2.

08:00 -- 16:58

I show up and went through New Employee Orientation. The first form they want is the I-9, which I fill out. Slightly confused because the box to write in is below the label, I cross out a few letters. The lady asks me to initial and date with the same pen, then circle it all. She takes the form.

I have a thick packet with a ring-bound initiation book. There are also three labeled folders in the packet: green, for Resources; blue, for Benefits; and red-orange, for Forms. Several of the sheets in the Resources folder are labeled ``Microsoft Confidential''. I also am given a red T-shirt that says ``Microsoft Intern 2005'' in white.

Two other Garfield students, Phun and some boy I don't know, are interns. A total of eighty interns are being ``onboarded'' today, approximately 35 from high school. (Microsoft employs 800 college interns yearly.) There's a talk about practices and legal issues, including a talk by Mr. Steve Ballmer (who looks slightly like a frog).

A Microsoft lawyer gives a talk to us. He addresses some important issues such as intellectual property and (vs.?) Open Source, confidentiality, and so forth. I learn that it is best not to bring Free Software onto campus in general, to avoid ``source taint''. Bringing source code to anything, especially Free Software, is especially discouraged. He mentions the various types of licenses, which ``run the gamut from the Berkeley license, which only requires attribution, to the GPL, which is the sort of license we really want to watch out for'' (not his exact words). Check with your manager if you want or need to use Open Source products on campus. He asks how many people have ``reviewed'' Open Source code. About a third of those present raise hands, myself included. According to that presentation, there are two confidentiality cultures. Information is public unless it isn't, and information isn't public unless it is. Microsoft is the latter, in case you didn't already know that.

On the shuttle to Red West B, where I will be working for the rest of this summer, I talk to a fellow intern and learn that I am not the only Linux-user in the group. It seems that there are 3--6 habitual (obsessive? compulsory?) Linuxers, yet we all applied for (and got!) internships at Microsoft.

I'm already privy to some ``confidential information''. As I intend to not get fired tomorrow morning, it would not be best to plan on obtaining interesting tech news. Accordingly, as I get deeper into interesting projects, I will probably write less.