Well, I just won a game of Kongai with a totally unfamiliar deck, so I can't be *too* rusty

Background: Earlier I slapped together some counterpick decks as I said I would, but I never really tested them. Just now I tried one out - Yoshiro w/Crane, Tafari w/Feather, and Phoebe w/Stone - meant to counter Popo, though it wasn't Popo that did me in and he didn't use the same deck second time around - by actually playing it. I can't say how well it plays as a whole, though, seeing as my opponent conceded before I had reason to bring Phoebe out.

Meanwhile I have yet to try the deck with Ubuntu in place of the Marquis. Largely because I can't remember what my reason was for wanting to put him in.

-Harry

The difficulty non-ramp

...However, it would appear that there's nothing about sheaves or presheaves on next week's homework.

There is, however, one of the most useless hints ever. "Hint: requires some work." Nice. In essence "Hint: this is hard." I mean, I guess it's a helpful note, but I don't think "hint" is really the right word.

There is also the Grothendieck group of finitely generated A-modules, which I think is just funny. Take all isomorphism classes of finitely generated A-modules [yes this is set-sized, if you think about it, though Nori doesn't bother to note this in his definition], take the free abelian group with them as basis, then mod out by the relation M'+M''=M whenever there's a short exact sequence 0→M'→M→M''→0. Wow. You just made (finitely generated) A-modules into an abelian group, even if the modules themselves aren't the elements (the image of M in this group is denoted cl(M)). You know, I feel like it would somehow be more appropriate if you used the free A-module rather than the free abelian group, so as to make an A-module of A-modules, but it doesn't seem like you'd actually gain any additional information that way, and the fact that that isn't what's done suggests that my suspicion is right.

-Harry