Harry goes to Boston (again)
Jul. 17th, 2012 10:01 pmAnd I am back in Ann Arbor. This is going to have quite a few parts, so for once I'm going to take the trouble to separate them with <hr> tags... well, I'll try. This might not separate into discrete topics very well.
Edit several hours later: I just realized I left out an interesting conversation Josh and I had on the subject of golems. Oh well. Maybe I'll put that in a later entry.
So this past weekend I went up to Boston to go see Josh and Heidi and Erick and whoever might be around at PROMYS. Naturally several students from when I was a counselor are now themselves counselors -- Heidi (as mentioned earlier), Zev, Andrew Ardito, Michael Greenberg. Also Will Perry, who's now a head counselor. Heidi had kept referring to him as "Will the Brit" previously, so I didn't realize it was Will Perry who I already knew. The other head counselor is Charlotte Chan, who I'd briefly met due to her going to Michigan next year.
Actually it seems a bunch of old PROMYS people were visiting this weekend -- there was also Jeremy Booher, Yiwei, and, uh, I think one more, who I forget right now.
Josh was kind enough to let me stay with him; this meant I had to deal with one of BU's pointlessly strict sets of rules about outsiders in the dorms, but I had stayed with a counselor instead, I would have instead had to deal with their other pointlessly strict set of rules about outsiders in the dorms.
Heidi gave me a new duck. This one is made of socks and named Durrell. He now sits on my desk along with Dererrer. (This one she didn't make for me, he already existed. She thinks she's the one who named him.) Josh said that Durrell looked like a voodoo doll. Like an actual voodoo doll, he pointed out, not the Hollywood version.
I had also intended to meet up with Shuo Zhang. Shuo, you may recall, had kind of gone hermit some time back, so I was pretty surprised when she emailed me out of the blue some time ago to say, hey, I'm in the area of Boston, so if you're coming here maybe we should meet up. It wasn't until last Saturday night, as I was checking things on the internet in Josh's apartment before going to sleep, that she mentioned that actually she had moved back to Minnesota.
(Kate Harney was not around either, having gone to Guatemala.)
Actually Erick I didn't get to see very much, he was working or sleeping much of the time I was active...
Unrelated to anything: Stone B50 has a new periodic table, one which includes elements up to Copernicium -- except, as Heidi pointed out but neither Josh nor I noticed before then, it has quite a larger error: Platinum and Mercury are swapped. As in, not just the symbols and the names, the whole thing; thus it's very visibly an error, since the numbers go 77, 80, 79, 78, 81...
Anyway. Friday night we played AGoT. (First edition, of course, I don't think I know anyone who has second edition. Still want to try that sometime.) Players were me as Greyjoy, Josh as Lannister, Andrew Ardito as Baratheon, Heidi as Stark, and Eve (Josh's girlfriend) as Tyrell. This... was not well-planned. We played in Towers[0], meaning we'd get kicked out at midnight, but we didn't start till pretty late. Dito and Eve were new, and Heidi didn't remember the rules (kind of surprised she played again after how long the last game went), so yes we had to explain the rules too (with Dito and Heidi constantly interrupting, of course).
This was basically the dullest game of AGoT I've ever played. Even with Josh playing, there was no conspiring at all until turn 4 -- which is when the game ended.
(Also, I made some real boneheaded moves that game out of sheer not paying attention. Well, one mildly boneheaded move -- entirely forgetting about consolidate power one turn and thus having a useless unit defend instead -- and one really boneheaded move. Namely, I had managed to get a ship into Bay of Ice, and was intending to use it to hit Heidi hard before she could prepare, but I didn't do this because I somehow failed to notice that Greywater Watch borders on Bay of Ice. I had to save my attack for turn 4, by which time she'd prepared some. It would have worked anyway, except, of course, the game ended before then[6].)
I'm still not totally clear on exactly what happened. Dito, as Baratheon, basically just kind of swept the south[3], and Eve just kind of collapsed before him, and this was enough to win, with Lannister largely unable to intervene. On turn 4 Josh and I agreed that I would turn Sunset Sea over to him so he could get down there and help out, but Dito was able to get his 7 before then. (Also, we failed to coordinate on the details -- Josh, moving before me, intended to take the Sunset Sea on his second march and get down there on his third; but I didn't intend to vacate the Sunset Sea until my third march, since I meant to use it to convoy an attack against Heidi on my second -- the first march was gathering my forces for the attack. Again, it didn't matter because Dito had won before Josh's second march anyway.)
I gather that Eve didn't really pay attention to the rules, didn't really know what was going on, and didn't really try to defend herself. I.e. she didn't really want to play but somehow went along with it anyway, meaning... yeah, this was ill-planned.
We (by which I mean Heidi and Josh and I) also spent a lot of time playing Zendo. Actually before I get to anything specific I'd like to talk more generally about Zendo terminology.
The fact of the matter is that I hadn't actually looked at the standard Zendo terminology page in quite a long time. Doing so just now I find that in fact some of the terminology I've been using is in fact *not* standard.
Let's start with pointing. I thought the standard terminology was that pieces don't point through other pieces; actually, they do. Oops. This makes pointing potentially much more interesting. I mean, OK, changing terminology doesn't actually change what rules you can make, but in practice it does affect what rules people tend to make.
Furthermore, take a look at what the pointing rules say about how pointing interacts with the surface of the table. I just played it and explained it as, flat pieces point differently (or maybe just grounded flat pieces point differently -- I hadn't really thought about ungrounded flat pieces). The standard way is a bit weirder -- if you have an ungrounded weird piece, it point at pieces that it doesn't point at in any visually obvious way. Hm.
Also I'd generally avoided making rules involving towers or stacks due to not being clear on what the definition is. The page does include an official definition, which I think is a good one, and I will have to promulgate here. It does not include a definition of a "nest", which might be a good thing. (I thought nest was a standard term -- I'm pretty sure I at least got it from *something* related to Icehouse pieces.)
Interestingly, it mentions height, something I've certainly never mentioned. Someone here -- I think MVK -- once made a height-based rule[4], but I don't think I've seen any other instances of this.
It also mentions direction, as in the same direction, or opposite direction. This strikes me as a bad idea to use; how can you tell when two things are "the same direction"? OK, sure, master's judgment call, but usually we try to build things so that the master doesn't have to make judgment calls. But maybe using angles would be workable -- acute vs. obtuse? Except lots of times people might place things at what are supposed to be right angles. Pointing works because although the corner case of "this is pointing *right* at the edge of that" can come up, people don't construct koans like that. But people might well try to construct something that's supposed to be a right angle, and then you have a problem.
I should also note -- the people Josh plays with have long used the term "pile" to mean a connected component of the touching graph. Surprisingly, Josh wasn't aware that this term was nonstandard. (The reason this surprises me is that I was under the impression that someone in his group had come up with it.)
Anyway. Zendo in Boston. Heidi, Josh, and I played a game Friday night, during which Josh explained that his group often moves koans into columns rather than marking each one. We didn't have Go stones or anything on hand, so the best thing to mark them with was the black and clear pyramids; but we had a small playing surface that didn't leave much room for extra markers. I'm uncomfortable with moving koans, but in this case moving them seemed the sensible thing.
We played again Saturday night, with Yiwei and Qiaochu as well. (That's Qiaochu Yuan, as in, that guy from MathOverflow; he's a PROMYS counselor this year.) This time we marked them -- we still lacked stones, so we used the black and clear pyramids. Heidi thought doing it this way was clearly wrong and started moving them anyway. This quickly got the rest of us annoyed at her, while she repeatedly asked just what she was doing wrong, insisting this was clearly the right way. Finally Qiaochu just up and left. (Yiwei left a bit later, too, due to being tired.)
I don't recall how it got to this point -- probably via Josh saying something to the effect of "There's only some finite amount you care about this. For instance, you wouldn't shave your head to get this" -- but eventually Josh stated that if Heidi were to shave her head, we would play it her way.
Sunday rolled around. I was hanging around with Heidi while she worked on putting together her prize for the talent show (not going to say anything about what it might be). It came up that she was considering actually shaving her head. I was checking things on the internet and noted that![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
squid314 had just shaved his head, pointing out the caption he used. (By the way, are you reading squid314? You should read squid314.)
And so eventually Heidi decided that she was going to actually shave her head. After finishing her work on her talent show prize and eating dinner, she immediately went into the bathroom in Warren Towers to do this. I wouldn't see the results until later that night, when she showed up at Josh's apartment intending to play Zendo. (She didn't realize the apartments had buzzers -- she didn't realize those were real, common things, actually -- and so filled up Josh's phone with way more text messages than he was expecting to have to pay for...) And yes, Josh kept his word and we played it her way.
(Also, she braided her hair before cutting it off. As of when I left, she was still keeping this braid in her backpack.)
Heidi and I played one more game on Monday, in the half hour before her lab. I know there are rules for two players where both are master and each is trying to figure out the other's rule, but I don't know the actual rules for that, so she just played master, I played student, and we didn't bother with mondo or guessing stones.
I couldn't get it in the half hour we had; her rule turned out to be, "The number of grounded pieces is bigger than the largest number of pieces in a pile". I was kind of taken aback; that is, I think, a very hard rule, by pretty much anyone's yardstick. And she had played a few games by now, so I was quite surprised that she insisted that that was an easy rule and I should have been able to get it in half an hour. I don't know if Julian could get it in all of an hour.
Actually, due to time constraints or people just giving up, it seemed quite a few of our games ended in stumpers. Partly this seems to be due to a difference in what rules are common in Josh's group vs. with my friends at Michigan. I used "every flat piece points at something", which stumped Josh and Heidi; very surprising, as that seems to me almost a canonical example of a simple rule. Josh's group, apparently, doesn't use pointing very often. This surprises me because it seems to me that pointing is the most interesting thing to use, due to its asymmetry; not using it seems highly constraining. (And as I said above, the standard notion of pointing seems even more interesting, since then nodes in the pointing graph will no longer be constrained to have outdegree at most 1.)
One thing that Josh's group seems to use a lot that we don't is numerical properties. Well, of course we use numbers, but -- aside from the none-vs.-any distinction which isn't really numerical -- what we mostly do with them is mostly restricted to comparisons or occasionally parity. And I don't think I've seen any of us ever make a pip-count-based rule[5]. Heidi, Josh, and I played one game with Josh as master where the rule was "The number of pieces is a power of 2". Heidi won, as I remained stumped the whole game. OK, it sure seemed like all koans with the same numbers of pieces were getting marked the same -- but someone wouldn't *really* make a rule that depended only on the number of pieces, would they? Apparently, they would.
So it happened that I ended up discussing wireheading with Heidi. To my shock, she was in favor of it. I told Josh about this, but initially -- well, when you talk about the problem of wireheading in constructing an AGI, there's two different problems you could be talking about. The first potential problem is that the AI wireheads itself. (But don't think that means it's safe -- it isn't using a human motivational system, it's going to try to now maximize the probability that it can continue to wirehead itself, and it can determine that there's a nonzero probability that you might interefere with that...) The second potential problem is that the AI tries to wirehead everyone else. But I guess when you say a person is in favor of wireheading, you usually just mean the first.
But I didn't explicitly say that when I initially related this, because I'd forgotten that distinction at the time, and so Josh was kind of horrified to learn that Heidi was in favor of forcibly wireheading everyone on the planet. Yes, even if it means all of humanity starves to death soon afterwards.
Josh and I then had to try to explain to her that most humans do, in fact, value things other than happiness. This didn't go very well. I would personally find it a bit incredible if Heidi *actually* had no other values beyond human happiness, but she certainly claims not to.
Due to poor planning I ended up having to leave kind of abruptly, didn't really get a chance to say goodbye to people. Oops.
I also tried too late to print out my Michigan Flyer ticket, finding that firstly the way I was trying to print things wouldn't work, and secondly it wouldn't let me print it because it said it had already been used! I mean, I did already use it once, because it was a round trip ticket, so... perhaps I should call them and complain? Well, ultimately it didn't matter, because the Michigan Flyer guy didn't check tickets. Actually I don't think I've ever seen them actually check tickets, but for a while I forgot this and was afraid I would end up stranded in Detroit airport...
Josh had suggested we might brainstorm some integer complexity stuff, but we never found the time for that. Dunno that it would have gone much of anywhere, since we've gone in pretty different directions with it. (And he is actually writing up his upper bound result now!)
OK. Is that everything? Well, it'll do for now.
-Harry
[0]PROMYS has moved back to Towers now, but they still eat in Warren Towers. Although the PROMYS students also use the Warren Towers lounges, including Melville; apparently the counselors are a bit overworked trying to man all this space.
[3]Oddly, he didn't take King's Landing until pretty late. Maybe he didn't realize how easy it is to do so?
[4]Specifically, IIRC, it was that there must exist a unique highest piece. Though I'm not sure if his notion of height exactly matches the standard one.
[5]Terrible idea that came up while we were discussing rules before the game with Qiaochu: Instead of adding up the sizes of the pieces used, xor them!
[6]Let me add in one more boneheaded move I made, which I don't bring up above because I feel like it's the sort of mistake I could have made at any time. I (Greyjoy) attacked Heidi (Stark), with an overall advantage of +1. As I was first on the Fiefdoms track, this should have been an easy win. I decided to make it certain by playing Balon Greyjoy. Except she played Catelyn Stark, giving herself an additional +2 in a way that Balon doesn't counter. Oops. I was able to use the Valyrian steel blade and win anyway, but...
Edit several hours later: I just realized I left out an interesting conversation Josh and I had on the subject of golems. Oh well. Maybe I'll put that in a later entry.
So this past weekend I went up to Boston to go see Josh and Heidi and Erick and whoever might be around at PROMYS. Naturally several students from when I was a counselor are now themselves counselors -- Heidi (as mentioned earlier), Zev, Andrew Ardito, Michael Greenberg. Also Will Perry, who's now a head counselor. Heidi had kept referring to him as "Will the Brit" previously, so I didn't realize it was Will Perry who I already knew. The other head counselor is Charlotte Chan, who I'd briefly met due to her going to Michigan next year.
Actually it seems a bunch of old PROMYS people were visiting this weekend -- there was also Jeremy Booher, Yiwei, and, uh, I think one more, who I forget right now.
Josh was kind enough to let me stay with him; this meant I had to deal with one of BU's pointlessly strict sets of rules about outsiders in the dorms, but I had stayed with a counselor instead, I would have instead had to deal with their other pointlessly strict set of rules about outsiders in the dorms.
Heidi gave me a new duck. This one is made of socks and named Durrell. He now sits on my desk along with Dererrer. (This one she didn't make for me, he already existed. She thinks she's the one who named him.) Josh said that Durrell looked like a voodoo doll. Like an actual voodoo doll, he pointed out, not the Hollywood version.
I had also intended to meet up with Shuo Zhang. Shuo, you may recall, had kind of gone hermit some time back, so I was pretty surprised when she emailed me out of the blue some time ago to say, hey, I'm in the area of Boston, so if you're coming here maybe we should meet up. It wasn't until last Saturday night, as I was checking things on the internet in Josh's apartment before going to sleep, that she mentioned that actually she had moved back to Minnesota.
(Kate Harney was not around either, having gone to Guatemala.)
Actually Erick I didn't get to see very much, he was working or sleeping much of the time I was active...
Unrelated to anything: Stone B50 has a new periodic table, one which includes elements up to Copernicium -- except, as Heidi pointed out but neither Josh nor I noticed before then, it has quite a larger error: Platinum and Mercury are swapped. As in, not just the symbols and the names, the whole thing; thus it's very visibly an error, since the numbers go 77, 80, 79, 78, 81...
Anyway. Friday night we played AGoT. (First edition, of course, I don't think I know anyone who has second edition. Still want to try that sometime.) Players were me as Greyjoy, Josh as Lannister, Andrew Ardito as Baratheon, Heidi as Stark, and Eve (Josh's girlfriend) as Tyrell. This... was not well-planned. We played in Towers[0], meaning we'd get kicked out at midnight, but we didn't start till pretty late. Dito and Eve were new, and Heidi didn't remember the rules (kind of surprised she played again after how long the last game went), so yes we had to explain the rules too (with Dito and Heidi constantly interrupting, of course).
This was basically the dullest game of AGoT I've ever played. Even with Josh playing, there was no conspiring at all until turn 4 -- which is when the game ended.
(Also, I made some real boneheaded moves that game out of sheer not paying attention. Well, one mildly boneheaded move -- entirely forgetting about consolidate power one turn and thus having a useless unit defend instead -- and one really boneheaded move. Namely, I had managed to get a ship into Bay of Ice, and was intending to use it to hit Heidi hard before she could prepare, but I didn't do this because I somehow failed to notice that Greywater Watch borders on Bay of Ice. I had to save my attack for turn 4, by which time she'd prepared some. It would have worked anyway, except, of course, the game ended before then[6].)
I'm still not totally clear on exactly what happened. Dito, as Baratheon, basically just kind of swept the south[3], and Eve just kind of collapsed before him, and this was enough to win, with Lannister largely unable to intervene. On turn 4 Josh and I agreed that I would turn Sunset Sea over to him so he could get down there and help out, but Dito was able to get his 7 before then. (Also, we failed to coordinate on the details -- Josh, moving before me, intended to take the Sunset Sea on his second march and get down there on his third; but I didn't intend to vacate the Sunset Sea until my third march, since I meant to use it to convoy an attack against Heidi on my second -- the first march was gathering my forces for the attack. Again, it didn't matter because Dito had won before Josh's second march anyway.)
I gather that Eve didn't really pay attention to the rules, didn't really know what was going on, and didn't really try to defend herself. I.e. she didn't really want to play but somehow went along with it anyway, meaning... yeah, this was ill-planned.
We (by which I mean Heidi and Josh and I) also spent a lot of time playing Zendo. Actually before I get to anything specific I'd like to talk more generally about Zendo terminology.
The fact of the matter is that I hadn't actually looked at the standard Zendo terminology page in quite a long time. Doing so just now I find that in fact some of the terminology I've been using is in fact *not* standard.
Let's start with pointing. I thought the standard terminology was that pieces don't point through other pieces; actually, they do. Oops. This makes pointing potentially much more interesting. I mean, OK, changing terminology doesn't actually change what rules you can make, but in practice it does affect what rules people tend to make.
Furthermore, take a look at what the pointing rules say about how pointing interacts with the surface of the table. I just played it and explained it as, flat pieces point differently (or maybe just grounded flat pieces point differently -- I hadn't really thought about ungrounded flat pieces). The standard way is a bit weirder -- if you have an ungrounded weird piece, it point at pieces that it doesn't point at in any visually obvious way. Hm.
Also I'd generally avoided making rules involving towers or stacks due to not being clear on what the definition is. The page does include an official definition, which I think is a good one, and I will have to promulgate here. It does not include a definition of a "nest", which might be a good thing. (I thought nest was a standard term -- I'm pretty sure I at least got it from *something* related to Icehouse pieces.)
Interestingly, it mentions height, something I've certainly never mentioned. Someone here -- I think MVK -- once made a height-based rule[4], but I don't think I've seen any other instances of this.
It also mentions direction, as in the same direction, or opposite direction. This strikes me as a bad idea to use; how can you tell when two things are "the same direction"? OK, sure, master's judgment call, but usually we try to build things so that the master doesn't have to make judgment calls. But maybe using angles would be workable -- acute vs. obtuse? Except lots of times people might place things at what are supposed to be right angles. Pointing works because although the corner case of "this is pointing *right* at the edge of that" can come up, people don't construct koans like that. But people might well try to construct something that's supposed to be a right angle, and then you have a problem.
I should also note -- the people Josh plays with have long used the term "pile" to mean a connected component of the touching graph. Surprisingly, Josh wasn't aware that this term was nonstandard. (The reason this surprises me is that I was under the impression that someone in his group had come up with it.)
Anyway. Zendo in Boston. Heidi, Josh, and I played a game Friday night, during which Josh explained that his group often moves koans into columns rather than marking each one. We didn't have Go stones or anything on hand, so the best thing to mark them with was the black and clear pyramids; but we had a small playing surface that didn't leave much room for extra markers. I'm uncomfortable with moving koans, but in this case moving them seemed the sensible thing.
We played again Saturday night, with Yiwei and Qiaochu as well. (That's Qiaochu Yuan, as in, that guy from MathOverflow; he's a PROMYS counselor this year.) This time we marked them -- we still lacked stones, so we used the black and clear pyramids. Heidi thought doing it this way was clearly wrong and started moving them anyway. This quickly got the rest of us annoyed at her, while she repeatedly asked just what she was doing wrong, insisting this was clearly the right way. Finally Qiaochu just up and left. (Yiwei left a bit later, too, due to being tired.)
I don't recall how it got to this point -- probably via Josh saying something to the effect of "There's only some finite amount you care about this. For instance, you wouldn't shave your head to get this" -- but eventually Josh stated that if Heidi were to shave her head, we would play it her way.
Sunday rolled around. I was hanging around with Heidi while she worked on putting together her prize for the talent show (not going to say anything about what it might be). It came up that she was considering actually shaving her head. I was checking things on the internet and noted that
squid314 had just shaved his head, pointing out the caption he used. (By the way, are you reading squid314? You should read squid314.)And so eventually Heidi decided that she was going to actually shave her head. After finishing her work on her talent show prize and eating dinner, she immediately went into the bathroom in Warren Towers to do this. I wouldn't see the results until later that night, when she showed up at Josh's apartment intending to play Zendo. (She didn't realize the apartments had buzzers -- she didn't realize those were real, common things, actually -- and so filled up Josh's phone with way more text messages than he was expecting to have to pay for...) And yes, Josh kept his word and we played it her way.
(Also, she braided her hair before cutting it off. As of when I left, she was still keeping this braid in her backpack.)
Heidi and I played one more game on Monday, in the half hour before her lab. I know there are rules for two players where both are master and each is trying to figure out the other's rule, but I don't know the actual rules for that, so she just played master, I played student, and we didn't bother with mondo or guessing stones.
I couldn't get it in the half hour we had; her rule turned out to be, "The number of grounded pieces is bigger than the largest number of pieces in a pile". I was kind of taken aback; that is, I think, a very hard rule, by pretty much anyone's yardstick. And she had played a few games by now, so I was quite surprised that she insisted that that was an easy rule and I should have been able to get it in half an hour. I don't know if Julian could get it in all of an hour.
Actually, due to time constraints or people just giving up, it seemed quite a few of our games ended in stumpers. Partly this seems to be due to a difference in what rules are common in Josh's group vs. with my friends at Michigan. I used "every flat piece points at something", which stumped Josh and Heidi; very surprising, as that seems to me almost a canonical example of a simple rule. Josh's group, apparently, doesn't use pointing very often. This surprises me because it seems to me that pointing is the most interesting thing to use, due to its asymmetry; not using it seems highly constraining. (And as I said above, the standard notion of pointing seems even more interesting, since then nodes in the pointing graph will no longer be constrained to have outdegree at most 1.)
One thing that Josh's group seems to use a lot that we don't is numerical properties. Well, of course we use numbers, but -- aside from the none-vs.-any distinction which isn't really numerical -- what we mostly do with them is mostly restricted to comparisons or occasionally parity. And I don't think I've seen any of us ever make a pip-count-based rule[5]. Heidi, Josh, and I played one game with Josh as master where the rule was "The number of pieces is a power of 2". Heidi won, as I remained stumped the whole game. OK, it sure seemed like all koans with the same numbers of pieces were getting marked the same -- but someone wouldn't *really* make a rule that depended only on the number of pieces, would they? Apparently, they would.
So it happened that I ended up discussing wireheading with Heidi. To my shock, she was in favor of it. I told Josh about this, but initially -- well, when you talk about the problem of wireheading in constructing an AGI, there's two different problems you could be talking about. The first potential problem is that the AI wireheads itself. (But don't think that means it's safe -- it isn't using a human motivational system, it's going to try to now maximize the probability that it can continue to wirehead itself, and it can determine that there's a nonzero probability that you might interefere with that...) The second potential problem is that the AI tries to wirehead everyone else. But I guess when you say a person is in favor of wireheading, you usually just mean the first.
But I didn't explicitly say that when I initially related this, because I'd forgotten that distinction at the time, and so Josh was kind of horrified to learn that Heidi was in favor of forcibly wireheading everyone on the planet. Yes, even if it means all of humanity starves to death soon afterwards.
Josh and I then had to try to explain to her that most humans do, in fact, value things other than happiness. This didn't go very well. I would personally find it a bit incredible if Heidi *actually* had no other values beyond human happiness, but she certainly claims not to.
Due to poor planning I ended up having to leave kind of abruptly, didn't really get a chance to say goodbye to people. Oops.
I also tried too late to print out my Michigan Flyer ticket, finding that firstly the way I was trying to print things wouldn't work, and secondly it wouldn't let me print it because it said it had already been used! I mean, I did already use it once, because it was a round trip ticket, so... perhaps I should call them and complain? Well, ultimately it didn't matter, because the Michigan Flyer guy didn't check tickets. Actually I don't think I've ever seen them actually check tickets, but for a while I forgot this and was afraid I would end up stranded in Detroit airport...
Josh had suggested we might brainstorm some integer complexity stuff, but we never found the time for that. Dunno that it would have gone much of anywhere, since we've gone in pretty different directions with it. (And he is actually writing up his upper bound result now!)
OK. Is that everything? Well, it'll do for now.
-Harry
[0]PROMYS has moved back to Towers now, but they still eat in Warren Towers. Although the PROMYS students also use the Warren Towers lounges, including Melville; apparently the counselors are a bit overworked trying to man all this space.
[3]Oddly, he didn't take King's Landing until pretty late. Maybe he didn't realize how easy it is to do so?
[4]Specifically, IIRC, it was that there must exist a unique highest piece. Though I'm not sure if his notion of height exactly matches the standard one.
[5]Terrible idea that came up while we were discussing rules before the game with Qiaochu: Instead of adding up the sizes of the pieces used, xor them!
[6]Let me add in one more boneheaded move I made, which I don't bring up above because I feel like it's the sort of mistake I could have made at any time. I (Greyjoy) attacked Heidi (Stark), with an overall advantage of +1. As I was first on the Fiefdoms track, this should have been an easy win. I decided to make it certain by playing Balon Greyjoy. Except she played Catelyn Stark, giving herself an additional +2 in a way that Balon doesn't counter. Oops. I was able to use the Valyrian steel blade and win anyway, but...
