There is a surprising amount price variation on the auction closing price for the same piece of good on ebay. Certainly most of the time this includes consideration for the reputation of the sellers and other such miscellany, but even when the same seller sells, over time, multiple pieces of the same good, at the same time of the day every day, the price fluctuates by a least a few dollars each way around the mean, and occasionally by five dollars or more.

I am surprised that this happens. The distribution of closing prices should be sharply cut off at the lower end, because any good selling at substantially below the mean is subject to arbitrage on the price distribution. This works best for abstract goods like gift certificates, where there is no shipping cost to complicate matters. For example, a gift certificate sells for $100 on average, so you bid on all auctions for it at $95. If there is enough price fluctuation, you will get a few at that price. Now you relist them. There are more sophisticated options but let’s say you just relist them as auctions. Then on average, you will get $100. So you make $5 for each one sold.

If everybody thinks this can be done, they will all go do this and there will be no auctions that close below $95.

This shows closing prices of equally valued amazon gift certificates sold on ebay within the last two days. There are a lot of “buy it now” transactions listed, and those tend to be at the higher end. It seems like an excellent arbitrage opportunity right here to bid on auctions selling at below $80 and relist as “buy it now” for, say, $88. Ebay and paypal fees will eat about $6, so there is still a profit of at least $2 each time this happens.

Comments(0)

Boy, this one may need a table of contents…

Let’s see, it all started with somebody wondering if you can get a letter back from the postal service once it has been mailed, but before it has been delivered. Maybe you changed your mind about sending the letter, for example. I still don’t know the answer, but I’m guessing if there is no return address on it, forget it. If there is a return address, however, it ought to be possible, right? The sender will get the letter back normally in the case that it is undeliverable, so the sender is essentially a secondary recipient. What does the postal service do with undeliverable mail that has no return address anyway? Shred it? Anyway, this doens’t seem like a satisfactory conclusion in any case, that the return address should play an unrelated role in the mail interception problem.

Which brings me to the second topic. (Read the article)

Comments(6)

Old news but news to me.

Cingular, which just a year or two ago acquired AT&T Wireless and changed everything to the orange color scheme, will change everything back to AT&T again.

What?

The newly reborn AT&T, fresh off of digesting BellSouth, plans to turn off the lights on their Cingular Wireless brand by next year and name it AT&T Wireless.

What?

Cingular originally was unconnected to AT&T; it was a joint venture between SBC and BellSouth.

Subsequently, AT&T sold its wireless assets to Cingular; AT&T proper remained a separate company, and “AT&T Wireless” ceased to exist.

Last year, SBC acquired what was left of AT&T, but they adopted the AT&T corporate name (due to its much stronger brand recognition). This made Cingular a joint venture of AT&T and BellSouth.

Now AT&T intends to acquire BellSouth. Once again, the AT&T brand will survive the takeover, meaning that AT&T will own 100% of Cingular.

What??

Thirty-odd years after the break-up of AT&T, it is piecing itself back together… hoho, the regulators of the 70s must be rolling in their graves… or wheelchairs, whatever.

Comments(0)

Memory management: Despite running on 256MB (actually 274MB is what it’s set to, to be precise), memory management in Vista seems to be working well. The paging policy is persistently keeping physical RAM usage at around 200MB, +/- 20MB. This isn’t too different from XP.

Disk management: There appears to be some rudimentary non-destructive repartitioning functions like “extend” and “shrink,” much like gpart. The disk is also versioned. Creating and destroying symbolic links, however, is still not exposed in the shell.

Network management: A crapload of changes in network management — too much to figure out what’s going on there right now. Most notable is probably exposing IPv6 support.

Privilege escalation: Windows itself now makes the request for administrative privilege if it is needed, instead of saying the current user privileges are insufficient. So far this is saving a lot of time.

Foreign languages: Display of East Asian languages are enabled by the default installation (I guess that just means the fonts and codepages are installed by default). The input methods I use are still the same, although there are a lot more input methods now, including a half dozen minority languages of China. That’s amazing. But, I can’t believe that the language bar is still having issues, not knowing whether to hide itself on the taskbar or how it’s supposed to be aligned.

Other stuff: There is this “Windows Cardspace” thing which seems to be a online accounts manager. There is also a “People Near Me” function that uses the Messenger social network. Some lame games and a “game manager.” BitLocker and ReadyBoost are nice, but kind of over-engineered. I doubt these will be used extensively.

In general, I think there are many good and needed changes here, but very little that I find compelling. From 2000 to XP, remote desktop, multiple user logon, system restore, and wireless support were compelling. From XP to Vista, the only thing I see is Media Center. But that isn’t in most versions of Vista. Add packet writing of optical discs, also, that might come in handy. Some of the other changes might have been compelling years ago (like Sidebars), but at this point are too little too late. If new applications turn up either from Microsoft or others to make a compelling case for the new graphics subsystem or anything else that has been included (pen input? speech recognition? text-to-speech? imaging/color codecs?), things may be different.

In other trials:

• RDP 6 works fine. Sound quality seems better. Not much else exciting going on here. I thought there might be application publishing support, but that requires the server OS.
• Office 12 is fine. XML file format and some UI changes. The thing seems to be the same to me. Outlook doesn’t use the new UI but has improved IMAP support including remote sent-mail folder and auto-purge support, but I had those working with scripts anyway.

Comments(0)

Back during RC1, I requested a key but never got around to installing it. Then RC2 came out. Now of course Vista has RTM’d and the official MIT release is supposed to be coming at the end of this month, but I finally decided to see what’s what and gave RC2 a go. I was not about to blow away any production machine, but there was already a Linux host machine running some OS’s via VMWare, so that’s where the install went.

I had only two small problems. One, I had to repartition the only drive because I believed Vista docs, which said it required a min of 15GB. (The nice tool “gparted” did the trick of non-destructive repartitioning — when it didn’t crap out!) Turns out 15GB is total BS. A clean install of Vista Ultimate took 4-5GB on the disk. (I don’t remember having any choice over which version to install, strange!)

The second problem: The installer also refused to begin on a machine with less than 512MB of RAM since it’s the “minimum requirement.” I was poking around for a workaround online and saw people asking the same question. No answer was ever given (no command switch that I am aware of), only a swarm of trolls boasting about their new machines and how Vista could not possibly be useful on anything less than 1GB. Well, utter BS. It’s running right now on 256MB of RAM … Inside VMWare … On a physical machine that actually only has 512MB of RAM … Rendered over a remote desktop connection with all graphics turned on …. And tunneled over an uncompressed PPTP link. The machine is otherwise a P4 1.7GHz. It does just fine. I’m writing this in Vista right this moment and I wouldn’t be doing it if I felt the slightest bit of inconvenience. On a clean boot, the system eats around 160MB of RAM. That’s a lot more than the typical 60MB/80MB of 2000/XP, but it isn’t bad. The way I got it to install was this: I had to set 512MB of RAM for the virtual machine just to let the setup start (and the machine thrashed a bit — due to VMWare paging, not even due to the setup program), but as soon as setup rebooted for the first time, I switched the VM back to 256MB.

It is working well enough that I’m thinking of putting this on a real machine. The usability improvements are good and the sort that exercise the hardware improvements over the years — the Start menu search among them. The metaphors and idioms are still very much what was seen in XP though, so there is much continuity here. Maybe that’s why people say it’s XP Service Pack 11. But I think that’s a good thing in this case.

Next up, installing RDP 6 client for XP and Office 12 beta.

Comments(0)

Already here in a Dell? That didn’t take long, did it. True, it doesn’t take much to replicate 2 cores to 4, but huh, Intel is really not letting up on AMD. Intel is promising an 80-core teraflop processor within 5 years, with cache stack-intereconnected, or perhaps with optical interconnects. Hardware is moving fast these days. Too bad compilers aren’t doing too well in extracting thread level parallelism yet. They had better catch up soon or they’d be the next bottleneck.

Chip multiprocessors have definitely caught on, realizing the predictions of the Hydra project from some years back. Come to think of it, I was pretty close to going into computer architecture… for old time’s sake, here’s a nice review from Olukotun.

Comments(0)

This paragraph caught my eye:

Some music theorists have criticized serialism on the basis that the compositional strategies employed are often incompatible with the way information is extracted by the human mind from a piece of music. Nicolas Ruwet (1959) was one of the first to criticise serialism through a comparison with linguistic structures. Henri Pousseur (1959) questioned the equivalence made by Ruwet between phoneme and the single note, and suggested that analyses of serial compositions that Ruwet names as exceptions to his criticisms might “register the realities of perception more accurately.” Later writers have continued Ruwet’s line of reasoning. Fred Lerdahl, for example, outlines this subject further in his essay “Cognitive Constraints on Compositional Systems” (Lerdahl 1988). Lehrdahl has in turn been criticized for excluding “the possibility of other, non-hierarchical methods of achieving musical coherence,” and for concentrating on the audibility of tone rows (Grant 2001, 219), and the portion of his essay focussing on Boulez’s “multiplication” technique (exemplified in three movements of Le Marteau sans maître) has been challenged on perceptual grounds by Stephen Heinemann (1998).

Although the above paragraph refers to “the way information is extracted by the human mind,” I think the problem is not that the serialist information is encoded in a way that is difficult for the human mind to extract (it may well be), so much as there is possibly insufficient information encoded to begin with, by any reasonable measure. Certainly, the compositional strategies of serialism call for much randomization and uniform dithering, such that most of what appears to be informative content is in fact common randomness coupled with very very little actual musical idea. I mean that’s how this music gets written right? A tiny bit of innovation, then mechanically amplified by pseudorandomness.

I don’t know enough about this, so just pure speculation here. Note to self, read: Cognitive Constraints on Compositional Systems

Contemporary Music Review
Publisher: Routledge, part of the Taylor & Francis Group
Issue: Volume 6, Number 2 / 1992
Pages: 97 – 121
URL: Linking Options
DOI: 10.1080/07494469200640161

Cognitive constraints on compositional systems

Fred Lerdahl
Columbia University, New York City

Comments(0)

This is old, but the humor is timeless.

Just look at these traffic patterns. Madness, or stroke of genius?
(Read the article)

Comments(2)

This is interesting, this self-parking car thing. Well, it’s not completely automatic, but it is supposed to take care of the most difficult part. Now I know my driving test required a demonstration of parallel parking ability – not sure about other states, so it is natural to ask, if somebody had this car, should they turn this off for the test?

Are there existing conventions dealing with technological aid in various kinds of skills testing? I can think of some, and they generally seem to fall on the side of accepting technology, with restrictions that can sometimes be arbitrary. For example, on the SAT you could use a graphing calculator, but not one with a “QWERTY” keyboard, so a TI-89 was okay but the TI-92 was not, even though they ran the exact same firmware. Not sure if they changed this. Guess not. The College Board still appears to be living in the stone age with regard to some of these banned calculators:

• calculators with QWERTY (typewriter-like) keypads arbitrary
• calculators that contain electronic dictionaries they all do now or can
• calculators with paper tape or printers cash register? lol
• calculators that “talk” or make noise right
• calculators that require an electrical outlet haha
• etc…

That aside, the car that drives itself has got to be the most popular civilian application touted by sensor networks people. So far, most automated components of the car are not fully automated. There is always some human element in some key part of the chain, unlike in airplanes. There is some human resistance toward giving up control on this matter. The trend toward more automation may be unstoppable, though, if automation creeps in a bit at a time like this.

Comments(0)

This jar of preserved bamboo shoots from the Chinese supermarket here… Comrade Food brand. Hahaha, you chuckle knowingly. Cue the Internationale. Right?

Wrong. Because Comrade Food is made in Taiwan.

“What…?” You say.

Look, on the jar it says it’s made in Taizhong City. (It’s hard to see in this blurry photo, but there it is.) Now, Taizhong is hardly known to be a hotbed of camaraderie. But then again, “comrade” means something else on the renegade island, namely, “gay”. Well, whatever. In the company of “Darkie/Darlie” brand toothpaste, also sold in Taiwan (look it up), I must say branding with Comrade Food is nothing unusual.

Comments(0)