Blessed by the Potato

So I had a nice weekend in Niagara Falls. It was a little chilly, and a little rainy, but we never got soaked, and it was never so cold that we needed a scarf or hats. And, for the first time, I actually won money at the casino! I hit a $100 payout on one of the slot machines, and it rang the bell and everything.

The casinos are looking a little sad over there. Fallsview is brand new and all, but really all they have going for them is the slot machines: the tables are so expensive that I just simply can’t play them. We were hoping that a cool weekend in October, late at night and early Sunday morning would be quiet enough that they might try a $5 table, but the cheapest we saw was $15 roulette. There were lots of empty tables, so I don’t quite see why they didn’t bother to put someone there to take my money. They can easily get a few hundred an hour even from a $1 table, so the economics seem to be there. And if all there is to do are slot machines, then why go all the way out there? London has slot machines; Woodbine has slot machines… And speaking of slots, they’ve gotten more ridiculous since I was last in a casino. There were some that were so complicated that we just couldn’t figure out why we were or weren’t winning, and we had no idea what to hope would show up. Some had up to 20 different “lines” which is just getting ridiculous. I really didn’t find them fun at all.

I’d recommend the Hilton Garden Inn again: it was a bit out of the way, just about equidistant from St. Catherine’s, Niagara Falls, and Niagara-on-the-Lake. It was no problem to drive into Niagara Falls, park for free at the casino, and walk around from there. Plus, the hotel restaurant was actually great for breakfast, which is a pretty rare thing in a hotel restaurant. I had a waffle with fruit topping and whipped cream, and Wayfare got pancakes, bacon, and hash browns (lots of each), and our total bill came to under $10. Usually, I find hotel restaurants are really over priced because of the “convenience factor”, so this was a really nice surprise. Especially since we ate at the Golden Griddle the day before for twice as much, and it really sucked.

I also had a costume idea: dress up as a ghost and follow people when they turn their back, then look away when they turn to look at me.

I was just reading an article the other night about how random celebrity walk-ons usually signal the impending doom of a TV show, particularly when said celebrities aren’t even guest characters, but merely faces in the background, or themselves. Usually, I agree with that point of view: when you get down to throwing random people into scenes just to try to maintain interest, you’ve probably jumped the shark.

Corner Gas however, seems to be an exception to that, since it’s had random, self-conscious inserts of Canadians pretty much all along. Not an hour after I read that article, Adrienne Clarkson popped on the show to help tear down a barn. After all, what self-respecting Canadian wouldn’t want to help wreck a barn? Anyhow, it just really worked for me, in a way celebrity cameos never really have before. I laughed and laughed, and had just barely stopped laughing by the time they cut back to the former Governor General with a sledgehammer in hand…

I ran into this logic puzzle the other day, and have been thinking about it a lot between Nyquil comas. I will try to explain it as best I can here, but be warned that there are pages of discussion and argument on this as people try to wrap their brains around it. If you want to think about it on your own for a while, then stop reading here.

So, to start with, I’m going to drop the numbers down from what is presented on the xkcd page: let’s say there are 20 people in this weird tribe of logicians, and 5 of them have blue eyes. This better matches the forum discussion, and makes the numbers easier to follow.

A group of people with assorted eye colors live on an island. They are all perfect logicians — if a conclusion can be logically deduced, they will do it instantly. No one knows the color of their eyes. Every night at midnight, a ferry stops at the island. If anyone has figured out the color of their own eyes, they [must] leave the island that midnight. Everyone can see everyone else at all times and keeps a count of the number of people they see with each eye color (excluding themselves), but they cannot otherwise communicate. Everyone on the island knows all the rules in this paragraph.

On this island there are 5 blue-eyed people, 15 brown-eyed people, and the Guru (she happens to have green eyes). So any given blue-eyed person can see 15 people with brown eyes and 4 people with blue eyes (and one with green), but that does not tell him his own eye color; it could be 16 brown and 4 blue. Or 15 brown, 4 blue, and he could have red eyes.

The Guru is allowed to speak once (let’s say at noon), on one day in all their endless years on the island. Standing before the islanders, she says the following:

“I can see someone who has blue eyes.”

Who leaves the island, and on what night?

The answer, [Spoilers!] to start with, is that the blue eyed people would all leave the island together on the 5th night.

The logic looks inductive: if there was one person with blue eyes, they would see a sea of brown eyes, and the gurus words would speak to him, and him alone: he does not see anyone with blue eyes, so therefore he must be the one with blue eyes the guru was speaking about; he would leave the first night. If no one leaves the first night, then everyone knows that everyone else saw at least one other person with blue eyes, so they could not conclude that they were the only one and leave. If there is anyone who only sees one person with blue eyes, then they know that they must also have blue eyes, because otherwise that lone blue eyed person would have left (i.e.: the blue eyed person must have seen someone else with blue eyes, namely them). For every extra blue eyed person, the departure date is delayed one day. If no one left the second night, then there must be three blue eyed people, so anyone who only saw two blued eyed people would leave on the third night.

There are then a few more questions. The first: what information did the guru really add? The guru made it so that not only did everyone see people with blue eyes, she made it so everyone knew everyone knew that there were blue eyes.

The logic proceeds thusly: A person with blue eyes (say #5) sees 4 other people with blue eyes, but may assume that they have brown eyes. So #5 thinks that #4 sees 3 other people with blue eyes, and thinks that #4 thinks #3 sees 2 others; #5 thinks that #4 thinks that #3 thinks that #2 only sees one other, and thus #5 thinks that #4 thinks that #3 thinks that #2 thinks that #1 sees only brown eyed logicians.

The chain of logic is absolutely insane, since #5 knows, directly, that each of #’s 1-4 must see at least 3 others, because he can see all 4 of them. But nonetheless, he knows that each makes those assumptions down the chain. The trick is that at the end of the chain, someone thinks that someone only sees brown, which doesn’t help them guess their eye colour, since they don’t necessarily know that there is at least one blue present [or rather, they don’t think that someone else thinks that they think that they know… ok, it’s making my head hurt follow those chains of reasoning]. Once the guru speaks up, everyone knows that everyone at the bottom of their chains of reasoning knows that blue exists, and thus would deduce their eye colour if the twisted logic played out. It starts the countdown so that by day five, the 5 blue eyed people up and board the ferry.

It really makes my head spin to try to follow all that, especially since I sometimes have problem with the “a pencil” problems to begin with. But here’s the thing that I wonder: following the chain of logic, you know that if no one leaves the first night, it means that the blue eyed people all saw at least one other blue eyed person. Following the chain of guessing what you think the next guy thinks the next guy thinks, you get to the point where the last person hanging in that chain might think they were the only ones and actually leave the first night. So if no one does, that’s new information to act on. But, you also know, in addition to the crazy deduction, that everyone else sees at least one other blue. So no one expects anyone to leave that first night; in that case, is the fact that no one does leave actually new information? Does the chain of reason really continue night after night until they call all leave on the fifth? In fact, with 5 blues there are enough that you can know that not only does everyone else see at least one other blue, but you can know that they know everyone else can see at least one other. Can you move to the next step of waiting for someone to leave on day 4 if you knew nothing would ever happen on day 1?

About a year ago, Rogers cut its newsgroup offerings from the high speed internet service, first using the retarded excuse that it was a den of child porn (perhaps trying to shame people into not complaining?). They quickly retracted that explanation from their website, and instead said that usenet is an outdated, arcane corner of the internet, and that by cutting it they could bring us more high quality services such as free Yahoo 360 blogs (which would have been free even if we weren’t Rogers members).

Anyhow, putting the acrimony in the past, us Rogers users were stuck trying to find an alternate newsgroup provider. I looked around for a while and found that most services were expensive. Some offered unlimited downloads at speeds that could max out the cable connection, but rivalled my cable as a monthly expense. Others offered limited downloads, or limited speed (or both) for less. After looking around for a while, I settled on Astraweb. They seemed to have the best value for money at the time, and most suited my needs: I could buy a block of data transfer, and use it until it ran out, no matter how many months that would take me. Their reputation for reliability and retention was fairly middle-of-the-road on the reviews I read, which was good enough for me.

It’s been over 10 months now, and I’m just coming to the end of my first block of data: I bought 90 GB of transfer for $25 (US), which works out to an average of 8.2 GB/month (I’m on track to finish it at around the 11 month mark), at a cost of about $2.30 (US)/month. That’s pretty damned decent if you’ve tried looking at some of the monthly billed “light” plans that have roughly the same transfer: I think the up-front billing saves them some administrative costs.

I’m also now in a good position to comment on the service. In case you couldn’t tell by my tone above, I think it’s been great value-for-money, and when I use up my last few bytes in the next few weeks, I’ll be buying another 90 GB package (they also have a 280 GB package for $100 US, and I was briefly tempted to jump on that and not have to worry about a usenet server for years… until I did the math and saw that buying 4 90 GB packages is beter. That mathematical oddity is due to the fact that the 90 GB package is usually a 65 GB one, but there’s been a “special” on for the last year or so).

However, saying it’s great value-for-money isn’t quite the same as saying it’s fanstastic. The uptime is very good, there was only one big outage I was aware of, and it only lasted for a day or so. The transfer speeds are adequate, but if you’re impatient or a big user then they will probably not work out for you: I average about 50 kB/s, and let most of my downloads do their thing overnight. At peak times the transfer can dip below 20 kB/s, but that could also be a problem on Rogers’ end. Either way, that’s good enough to get a CD-sized chunk of data before morning, and a DVD-sized chunk of data in about a day or two. In addition to the steady-state speed, there’s also a noticeable delay when opening a connection that’s been idle (loggin in?). They advertise their retention as being 25 days, and it appears to be at least that good. There are some issues with missing parts, a bit worse than I remember on Rogers’ Giganews service. It’s never been so bad that a par2 file or two couldn’t patch it up though.

All-in-all, a good service that should be adequate for most people’s needs, but there are some limitations there for really serious users.

Update: I feel kind of dumb for not thinking about it before, but you have the ability to run more than one thread with some nntp software. I just opened two connections to Astraweb, and they’re both running smoothly at 60 kB/s at the moment, so using more connections may be a way to get better speeds out of Astraweb.

The leaves are barely halfway through changing for the fall, and already we’ve got a snowstorm here:

I actually took the photo fairly early on in the snow, when I thought it was a freak thing we’d never see again. Since then, it’s gotten dark, and the snow is accumulating out there.