00:00:15 elliott_: since 32 has so many factors, the gcd rule is likely to cut that down more than 34 which is twice a prime. i think. 00:01:23 oerjan: would it be possible to make some sort of progress report output? :P 00:01:35 probably? 00:02:27 dunno 00:03:32 oerjan: well i could always report progress through the original list passed, right? 00:03:36 that would be rather non-finegrained though 00:04:45 um but those are done in parallel... 00:05:04 oerjan: oh right :D 00:05:54 [[Two things - firstly: "Every time you do any string handling with C's native string types and functions, you need to make sure you have allocated enough memory for each string" - that's right, and experienced C developers write the correct code naturally and without thinking.]] 00:05:55 hehehe 00:06:00 humans are infallible and busywork is good 00:06:49 as i've almost said before, with perfect people everything works, even nazism 00:07:16 oerjan: Even sparks can be destroyed by overhead, but 99% of the time, it "just works". 00:07:44 (those suckers are *really* cheap) 00:08:01 mhm 00:08:22 Understanding probability can help to play at pokemon card! 00:09:01 oerjan: indeed 00:10:01 "That's silly - almost all programs use strings heavily, considering that they're really the most natural way for programs to take input from and produce output for humans." 00:10:09 and that's why everyone uses text UIs, without even colour or cursor control 00:11:11 I am playing CLEFAIRY [Lv15] against DARK CLEFABLE [Lv33] and I think my chance of hitting them is 50%. Their chance of hitting me is probably lower, but I don't know the exact amount. It is not greater than 50% though. 00:12:42 Pokémon names aren't officially always-uppercased nowadays 00:12:59 zzo38: link? 00:13:00 win 4 00:13:02 although it's a relatively recent change 00:13:18 coppro: No link, sorry. 00:13:40 Bulbapedia has most of the older Pokémon cards, but they're sorted by expansion, rather than zzo38's naming scheme 00:13:42 ais523: I am uppercased the names of all the cards 00:14:07 ais523: Do you know how to play pokemon card? 00:14:12 yes 00:14:24 * coppro braces 00:14:28 See if you understand my puzzle games. 00:14:47 Cords are a nice structure; I wonder if they're efficient when implemented purely functionally. 00:14:59 yup, there it was 00:15:02 elliott_: Can you try it and figure out? 00:15:06 it'll be complicated by the fact that zzo38 likely plays a mix of old rulesets 00:15:16 -!- CakeProphet has quit (Read error: Connection reset by peer). 00:15:19 zzo38: I'm trying to. 00:15:44 http://zzo38computer.cjb.net/textfile/miscellaneous/pokemon_card/puzzle.1 00:15:46 http://zzo38computer.cjb.net/textfile/miscellaneous/pokemon_card/puzzle.2 00:15:51 -!- CakeProphet has joined. 00:15:51 -!- CakeProphet has quit (Changing host). 00:15:51 -!- CakeProphet has joined. 00:16:43 ah, DARK CLEFABLE [Lv33] is more commonly called Dark Clefable (Team Rocket's Ambition) or Dark Clefable (GB 2) 00:16:55 so that was easy enough to find: http://bulbapedia.bulbagarden.net/wiki/Dark_Clefable_%28GB_2%29 00:17:05 Yes it is the GB 2 version. 00:17:09 newtype Cord a = Leaf ![a] 00:17:09 | Branch !Int !(Maybe (Cord l a)) !(Maybe (Cord l a)) 00:17:11 kind of ugly, but oh well 00:17:31 Calling it [Lv33] should still be unambiguous, though. Also, the puzzle games list all the card texts. 00:17:44 and hmm, is exclusive to one of the Game Boy games, rather than being an actual physical card 00:18:44 oerjan: still running, btw 00:19:07 Yes it is, although it could be played with physical cards too I think, if the chance os 1/3 for -0, 1/3 for -1, 1/3 for -2, then you can use a dice however, which is not part of the standard game. 00:19:35 (These are also the numbers I used to estimate my probability of hitting) 00:21:12 huh 00:22:10 elliott_: your newtype has some spurious l's 00:22:23 oerjan: indeed, remnants from a previous, more general attempt that I'm delaying 00:22:28 oerjan: ([a] is not a very efficient short-string structure) 00:23:04 "Leaf nodes (as well as some single-child internal nodes) also contain a short string." 00:23:07 i love that unexplained as-well-as 00:23:42 "Seen from another perspective, the binary tree of a rope can be seen as several levels of nodes. The bottom level contains all the nodes that have a short string. Higher levels have fewer and fewer nodes, until finally there is just one root node in the top level. We can build the rope by putting all the nodes with short strings in the bottom level, then randomly picking half those nodes in order to form parent nodes in the next level. Nodes with no p 00:23:42 arent (for example, the two nodes storing the strings "my_" and "me_i" in the diagram above) become the right child of the node located immediately to their left, thus forming a chain. In this way, we can build higher levels one level at a time. We stop when there is only one node remaining." 00:23:43 this is stupid 00:23:51 it's like it's purposefully trying not to be an actual algorithm 00:24:00 The GameBoy computer opponents seem to be bad at defensive play and average at offensive play, and often do stupid things. 00:24:15 ais523: Can you figure out my puzzle games? Can you understand it? 00:24:36 zzo38: Referring to cards by level is a bad approach 00:24:46 zzo38: The cards are much more well-known by name 00:25:16 oerjan: hmm, looks like finger trees are actually pretty close to this 00:25:20 The card texts of the puzzles are typed on there 00:25:30 And anyways I only know them by level anyways 00:25:38 oerjan: hm but wait 00:25:43 oerjan: finger trees have O(n) indexing, right? 00:25:50 oh, O(log(min(i,n-i))), weird 00:26:00 i guess because it tries from the end too 00:26:03 or something 00:26:05 zzo38: I haven't looked at it, I'm watching a Pokémon Blue TAS at the moment 00:26:56 elliott_: finger trees have O(1) indexing at the end, is the major difference i think 00:27:24 oerjan: that's actually quite useful, for strings 00:27:32 yeah 00:27:42 a pokemon TAS? oh dear 00:27:56 oerjan: the problem is that you don't want (Seq Char), you want (Seq AboutSixteenOrSoChars) 00:28:00 oerjan: which just complicates every algorithm 00:28:09 sufficiently so that I think reinventing the wheel might be more expedient 00:28:39 elliott_: the O(1) also keeps holding a*ly as you traverse inwards 00:28:55 yes, that is definitely a sentence that makes sense. 00:29:04 good job for making it. 00:29:08 a* is here the word i keep forgetting 00:29:08 i am proud of you. 00:29:21 What is a Pokemon TAS? 00:29:24 oerjan: asymptomatically? 00:29:25 er oh er 00:29:29 a....stigma 00:29:33 apex 00:29:35 aperture 00:29:40 argh :P 00:30:47 i know what you mean :D 00:31:29 good, good 00:31:49 http://hackage.haskell.org/packages/archive/simple-rope/0.1/doc/html/Data-Rope.html ;; TYPE FAMILIES: TOTALLY SIMPLE 00:31:53 anyway, the first 16 chars take 16*O(1) to traverse 00:32:36 elliott_ hates history 00:32:41 and historical stuff 00:33:43 oerjan: hmm, http://upload.wikimedia.org/wikipedia/commons/c/c2/Rope_example.jpg confuses me 00:33:51 apart from being tilted for no apparent reason, the lengths of the strings all vary wildly 00:33:59 is it because of inserting "s" somewhere that there's a one-char "s"? 00:34:02 elliott_: one hour 00:34:06 oerjan: :D 00:37:08 and the numbers only apply to the part vertically below ignoring the tilting... 00:38:01 oerjan: I know 00:38:07 oerjan: I just mean, why would all the strings not be equally-sized? 00:38:10 (Or one off from that) 00:38:15 So presumably it isn't a newly-created string, but a modified one 00:38:30 * oerjan doesn't actually know ropes 00:38:48 i think i briefly considered if they were useful for implementing dupdog 00:39:24 ok i guess if there isn't more to them than that, i know them 00:39:36 It is impossible to pull the ropes regardless of the number of your tentacles, I think. 00:40:28 maybe zzo38 is a great old one 00:41:20 Actually I am not one million years old. Nobody lives that long, I think. 00:44:17 -!- ralc has quit (Quit: Leaving). 00:47:24 hmm, I wonder what I should use as the target length of the small strings in a rope 00:47:32 there is probably some obvious answer based on keeping the tree shallow 00:47:34 but I don't know what it is 00:47:43 well, shallow but still... foresty enough 00:50:54 oerjan: still giong 00:50:54 going 00:51:53 RIP Bonobo Conspiracy 00:52:59 -!- azaq23 has quit (Quit: Leaving.). 00:53:52 * elliott_ was hoping oerjan would answer :D 00:54:01 4004 elliott 20 0 48740 3984 1260 S 109 0.1 124:00.04 nsearch.optim 00:54:02 two hours 00:55:01 wat 00:55:07 only in Haskell would you rotate a list by extending it to be infinitely long and then throwing out most of it ;P 00:55:08 oerjan: ? 00:55:41 * Phantom_Hoover → sleep 00:55:49 -!- Phantom_Hoover has quit (Quit: Leaving). 00:56:15 > let rot n l = take (length l) . drop n $ cycle l in rot 5 "abracadabra" 00:56:16 "adabraabrac" 00:56:19 LinuxJournal: Is there a world outside of computers? 00:56:19 -- http://www.linuxjournal.com/article/5048 00:56:25 "No." 00:56:39 elliott_: * elliott_ was hoping oerjan would answer :D 00:56:52 oerjan: hmm, I wonder what I should use as the target length of the small strings in a rope 00:56:53 there is probably some obvious answer based on keeping the tree shallow 00:56:53 but I don't know what it is 00:56:53 well, shallow but still... foresty enough 00:57:02 oerjan: also, the reason no haskell compiler does automatic memoisation is because of space leaks, right? 00:57:14 i think so 00:58:52 http://www.reddit.com/r/programming/comments/ifq59/a_regular_expression_matcher_in_30_lines_of_c/c23fz1d?context=3 hey oerjan, tell me whether my response is bullshit or not :P 00:58:55 ;D ;D ;D 01:00:05 -!- jcp has joined. 01:01:39 sounds fair. although i vaguely recall at one time reading that main = putStrLn "hello" >> main 01:01:40 ?hoogle Int -> a -> [[a]] 01:01:40 Prelude replicate :: Int -> a -> [a] 01:01:40 Data.List replicate :: Int -> a -> [a] 01:01:40 Control.Monad replicateM :: Monad m => Int -> m a -> m [a] 01:01:46 ?hoogle Int -> [a] -> [[a]] 01:01:47 Control.Monad replicateM :: Monad m => Int -> m a -> m [a] 01:01:47 Prelude drop :: Int -> [a] -> [a] 01:01:47 Prelude take :: Int -> [a] -> [a] 01:01:49 grr 01:01:52 does leak space 01:02:01 elliott_: how dare you paste in the middle of my broken line 01:02:03 oerjan: hmm, I don't think so 01:02:05 main is shared, after all 01:02:13 anyway, it won't blow the stack 01:02:14 elliott_: er, *did leak space 01:02:16 even if it does eat all your ram 01:02:18 at 01:02:19 [asterisk]ah 01:02:31 this was in some ancient version 01:03:02 elliott_: what do you want for that ?hoogle function? 01:03:20 oerjan: group into lists of length N 01:03:32 "abcdefgh" grouped by three -> ["abc","def","gh"] 01:03:35 :t chunks 01:03:36 elliott_: Data.Split, i think 01:03:36 :t chunk 01:03:36 Not in scope: `chunks' 01:03:37 Not in scope: `chunk' 01:03:40 oerjan: yeah :( 01:03:53 not in standard libraries 01:04:58 > takeWhile (not.null) . map(take 3) . iterate(drop 3) $ "abcdefgh" 01:04:59 ["abc","def","gh"] 01:09:49 -!- Sgeo has quit (Ping timeout: 255 seconds). 01:10:33 -!- Sgeo has joined. 01:11:30 algorithms suck 01:13:58 ?pl \d -> f d . g 01:13:58 (. g) . f 01:15:10 http://hackage.haskell.org/packages/archive/ListLike/3.1.1/doc/html/Data-ListLike.html#t:ListLike ugliest typeclass 01:15:28 elliott_: Should I feel bad for having made sabotage build dynamically? 01:15:34 pikhq: yes 01:15:40 -!- Sgeo has quit (Ping timeout: 255 seconds). 01:15:49 Well, that was straightforward. 01:15:56 oerjan: still going, btw 01:16:42 elliott_: It also seems to have cut the size of X11 by 95%. 01:17:59 Does the Linux kernel have support for error emulation? 01:19:48 -!- jcp|other has joined. 01:19:55 -!- jcp|other has quit (Read error: Connection reset by peer). 01:20:14 -!- jcp|other has joined. 01:23:48 -!- Sgeo has joined. 01:26:58 -!- CakeProphet has quit (Ping timeout: 260 seconds). 01:33:04 oerjan: COUGH COUGH 01:33:07 eek 01:33:12 TWO HOURS 01:33:32 oerjan: that was actually to answer my implicit question IN RETURN FOR ALL MY CPU TIME 01:33:32 also you beat me by seconds there 01:33:44 oerjan: hmm, I wonder what I should use as the target length of the small strings in a rope 01:33:44 there is probably some obvious answer based on keeping the tree shallow 01:33:44 but I don't know what it is 01:33:44 well, shallow but still... foresty enough 01:33:44 oh hm 01:34:01 i think i can say with some confidence that i dunno 01:34:06 I dunno whether you want it to be some fraction of the total initial string length or some constant :D 01:34:08 oerjan: >:( 01:34:29 elliott_: square root or logarithm perhaps? 01:34:40 hm but you want a lower limit too 01:34:44 yes, let's just throw random functions at the problem :D 01:34:58 RANDOM, THAT'S IT! 01:35:11 gaussian, to be precise 01:35:21 oerjan: :D 01:35:25 you are joking right 01:35:40 defaultChunkSize :: Int 01:35:40 defaultChunkSize = 64 01:35:40 pack :: (StringLike a) => Int -> a -> Rope a 01:35:40 pack cs s | size s < 2*cs = Leaf cs s 01:35:40 | otherwise = let (a,b) = splitAt cs s in pack cs a <> pack cs b 01:35:42 HMMMMMMMmmmmmmmmmmmmmmmmmm 01:35:47 _finally_ elliott_ is getting a clue 01:35:53 :( 01:36:31 -- | O(n) Convert a list of 'Word8' into a 'Rope' 01:36:32 pack::[Word8]->Rope 01:36:32 pack s= 01:36:32 if Prelude.length s<=leafSize then unsafePerformIO $ do 01:36:32 contents<-mallocForeignPtrArray leafSize 01:36:34 i0<-mallocForeignPtr 01:36:36 withForeignPtr i0 $ \i->poke i $ Prelude.length s 01:36:38 withForeignPtr contents $ \c->fillPtr c s 01:36:40 return $ String { contents, i0, offset=0, length_=Prelude.length s } 01:36:42 else 01:36:44 let (a,b)=Prelude.splitAt leafSize s in 01:36:46 append (pack a) (pack b) 01:36:48 where 01:36:50 fillPtr::Storable a=>Ptr a->[a]->IO () 01:36:52 fillPtr _ []=return () 01:36:54 fillPtr ptr (h:t)=do 01:36:56 poke ptr h 01:36:58 fillPtr (plusPtr ptr 1) t 01:37:00 cool 01:37:04 \end{code} Il n'est pas clair qu'on ne puisse pas faire beaucoup mieux que ce pack. Par exemple, vu qu'on calcule la taille de la liste, on pourrait sortir un arbre directement \'equilibr\'e. Vu qu'on ne fait jamais pack avec des 'ByteString's (sauf ghc quand il compile des IsString), \c ca ne vaut sans doute pas le co\^ut. Par contre, append est utilise presque partout. Il est crucial que son implementation soit extremement efficace. \begin{code} 01:37:09 cool 01:37:11 ok its leaf size is 65,536 01:37:13 so the ideal chunk size is between 64 and 65536 01:37:15 :D 01:39:56 wait this is the kind of "you need no know cache" stuff, isn't it. 01:40:01 *to know 01:40:27 oerjan: i'm not sure :D 01:40:32 HEY VOOOOORPAAAAAAAAAL 01:40:35 HEY GREEEEEGOOOOOOOOOR 01:41:33 Newspaper works as a mousepad! 01:41:41 Sgeo: Warren considers my ardent anti-historism beautiful. 01:43:31 lol 01:43:52 oerjan: 208:18.92 nsearch.optim 01:46:22 are you _sure_ you didn't accidentally write 43 or something 01:51:02 main = print (nsearch 34 1 0 [1..33]) 01:51:34 oerjan: heh, fromList is a painful function to write :( 01:51:42 because just getting a list of chunks isn't enough, i have to all spindly them down into a tree 01:51:43 im lazy 01:55:02 -!- Lymia has changed nick to Lymee. 01:56:19 :t Node 01:56:20 forall a. a -> Forest a -> Tree a 01:57:11 oerjan: what of it 01:57:52 :t Leaf 01:57:52 Not in scope: data constructor `Leaf' 01:58:14 > return True :: Tree Bool 01:58:15 Node {rootLabel = True, subForest = []} 01:58:27 aha 02:00:44 hm that's annoying 02:01:06 234:22.89 i give up 02:01:14 $ time ./nsearch.optim +RTS -N 02:01:14 ^C 02:01:14 real146m53.288s 02:01:14 user208m42.860s 02:01:14 sys25m50.830s 02:01:15 NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 02:01:23 two seconds later: OMG ITS FOUND 02:01:48 > undefined :: [Int] 02:01:49 *Exception: Prelude.undefined 02:02:02 why doesn't it print the initial [, anyway 02:02:23 for undefined 02:02:29 yeah 02:02:29 [undefined] would produce [ 02:02:43 well sure 02:02:51 > [undefined] :: [Int] 02:02:51 [*Exception: Prelude.undefined 02:03:01 but even without the brackets 02:03:26 > undefined :: String 02:03:27 "*Exception: Prelude.undefined 02:03:44 hm 02:03:57 that's the reason why fix show works 02:04:26 > fix show 02:04:27 "\"\\\"\\\\\\\"\\\\\\\\\\\\\\\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\"\\\\\\\\\\\\... 02:06:01 :t var 02:06:02 forall a. String -> Sym a 02:06:30 > fix (return.var.show) :: [Sym ()] 02:06:31 [[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[... 02:10:19 > fix (return.var.show) :: Maybe (Sym ()) 02:10:21 Just Just Just Just Just Just Just Just Just Just Just Just Just Just Just ... 02:10:43 hm 02:11:00 > Just (Just (Just True)) 02:11:00 Just (Just (Just True)) 02:16:05 -!- pikhq_ has joined. 02:16:09 -!- pikhq has quit (Ping timeout: 240 seconds). 02:16:16 elliott_: have you tried timing the parallel one for smaller values? that was some leap from 32 to 34... 02:16:37 yes i tried thirty 02:16:50 how long did that take? 02:19:53 -!- copumpkin has joined. 02:19:53 -!- copumpkin has quit (Changing host). 02:19:53 -!- copumpkin has joined. 02:20:39 realI1m31.738s 02:20:45 was that for 30? 02:20:52 yes 02:21:25 -!- copumpkin has quit (Client Quit). 02:21:31 that's more than 100 times up to when you stopped 34 :( 02:21:44 er, *almost 02:21:54 its obviously massively exponential 02:21:55 or worse 02:22:03 perhaps factorial 02:22:32 or rather, i cannot see how it could be _more_ than factorial. well maybe some of the large artithmetic on n. 02:22:44 *-t 02:23:22 wow, Slashdot has given me a total of 30 moderator points in a week 02:23:36 well you are a very moderate guy 02:24:52 according to the FAQ, it implies I'm in the top percentile of moderators, or something like that 02:25:16 (moderator selection is random, but apparently when you're selected, you get to do more moderation if the metamoderators think you've been doing well) 02:25:18 I mostly read via rss now 02:25:28 as far as I can tell, it just means most Slashdotters are really biased 02:25:38 and I read Slashdot for the comments 02:26:58 I have sometimes read the comments on Slashdot, too. 02:26:59 Are you able to read pokemon card puzzle by now? 02:27:06 ah, I forgot about it 02:27:11 let me look 02:27:45 I don't understand your notation for energy 02:28:08 does @ mean purple/psychic energy? 02:28:13 and if so, what is %? 02:28:57 ais523: What part are you having trouble with about the energy? Yes, @ means purple/psychic energy although you can still solve it without that information (all the card texts use the same notation too, in that file). % means leaf energy (green) 02:28:57 also, how many Prizes does each player have? are they called something else? 02:29:10 "Side cards remaining". 02:29:54 Also called Prizes also the name "side cards" is also used sometimes, and that is what I am using there. 02:29:58 and numbers on their own in Pokémon descriptions are damages in multiples of 10? 02:30:06 Yes. 02:30:23 -!- copumpkin has joined. 02:32:30 What is called ten damage is called 1 here; it is more logical. 02:32:32 you need to point out that the opponent has at least one card in draw pile 02:32:43 otherwise you can win by passing the turn :) 02:32:52 you have [irrelevant] there atm 02:36:10 I don't think it is necessary but OK I will change that 02:37:50 I fixed it. 02:38:12 -!- zzo38 has quit (Quit: Unfortunately I have to leave now, sorry). 02:40:18 I'm not convinced the first puzzle is possible; there seems to be no way to get a draw from the deck other than itemfindering Bill in any of the cards there, and if you do that, you have no way to refill your deck with Nightly Garbage Run 02:40:28 other than that, it's reasonably obvious what you have to do 02:41:45 ais523: it's not possible according to modern rules 02:42:06 it /may/ have been possible according to older rules, although I don't recall them well enough 02:42:20 in any case, I'd forgotten the rule that makes it impossible 02:42:21 coppro: under zzo38's ruleset, weakness is almost certainly *2 damage rather than a fixed bonus amount of damage like it is nowadays 02:42:56 I'm assuming you're meant to hit a weakness (using goop gas attack) in order to deal 60 damage to get one prize, and 10 to a benched to get the other 02:43:30 oh, you can just itemfinder for goop gas attack and ignore all the other cards in the discard 02:44:28 move a counter from dark muk to porygon, retreat Gengar for Spearow, GoW Dark Muk, retreat Spearow for Hypno, itemfinder for Goop Gas Attack, play it, then hit for 60+10 benched damage for the double KO 02:44:45 normally it's bad style to write a puzzle that has that much irrelevant stuff in... 02:46:00 ais523: I cannot find any evidence to support your assertion that weakness is not *2 damage nowadays, and that is the correct solution although it's illegal under modern rules as you can only retreat once per turn 02:46:10 note that I define 'modern rules' as 'the rulebook on the website' 02:46:15 you can only retreat once per turn nowadays? 02:46:19 I stopped playing before that was introduced 02:46:33 and nowadays, weakness is printed on the cards; it's sometimes *2, but more commonly +20 or +30 02:46:43 ah ok 02:46:57 whereas in the first few sets, weakness was always printed as *2, making it look like a rules reminder rather than actually necessary 02:47:03 quick scan shows that most recent expansion is almost always *2 02:48:31 although it appears to be relatively simple cards generally 02:48:35 hmm, so it's changed again? 02:48:44 as I said, I'm a little out of the loop 02:48:53 I play the videogames competitively, but not the cards 02:48:59 ah ok 02:49:13 the videogames are much cheaper, a sort of one-off £40 or so for each version, and they come out once every few years 02:49:25 right 02:49:38 -!- copumpkin has quit (Quit: Computer has gone to sleep.). 02:49:53 I suspect that the TCG adopts a similar approach to Magic of making on expansion a year much simpler 02:50:19 although it looks like they've been revising the rules a lot too 02:50:25 Magic's reason is also because there are a lot of boring cards that are nonetheless necessary to round out decks 02:50:37 and yes, I coincidentally quit Magic at the time of the latest big rules change 02:50:59 not because of the rules change itself (although I don't really see the point of some of the more controversial changes, and think they're mostly negative), but because I disliked the sets that they'd been making 02:51:25 The core sets were never really about printing utility cards 02:51:48 They need to be in every set since otherwise the limited format would be impossible to play 02:52:08 The new core set model is very awesome 02:52:16 yep 02:52:31 but it was the difference between core sets using basic utility cards everyone gets, and expansion sets having quirky ones 02:52:57 and often the basic ones were a better fit if you weren't going heavily themed 02:59:14 It appears that the Pokemon TCG has redone supporters and stadia again; now they are subtypes of Trainer 02:59:48 supporters were always subtypes of trainer, weren't they? 02:59:51 just with a 1/turn restriction? 03:00:16 yeah 03:00:23 except for a while when they were a card type 03:00:42 but they were originall cobbled on with annoyingly-present reminder text 03:00:58 the reminder text was needed 03:01:07 the new Trainer template has room for Item/Supporter/Stadium in the top-right corner and has a nice box at the bottom for the reminder text 03:01:12 because people kept trying to bring them to people using only old-style cards, when they first came out 03:01:49 the invention of supporters was needed, anyway; under the original rules, there was never a reason not to play 4x Bill and 4x Professor Oak 03:01:56 yeah 03:02:10 definitely agree there 03:02:43 (bill was completly safe, professor oak was a gamble but one that had incredibly good return) 03:02:47 there's now also basic metal and dark energy, but without the special effects 03:02:53 that was needed too, I think 03:03:02 Just wait until your hand is low, then play Oak 03:03:14 coppro: I mean, you might draw Oak at a time when it isn't useful 03:03:23 worst case is getting two Oaks from Bill 03:03:27 suppose 03:03:34 but then you just play one and the problem goes away 03:03:46 yep, it's just a waste more than anything else 03:03:50 it's a very good worst case :) 03:09:11 -!- copumpkin has joined. 03:09:11 -!- copumpkin has quit (Changing host). 03:09:11 -!- copumpkin has joined. 03:21:33 -!- jcp|1 has joined. 03:21:54 -!- jcp has quit (Ping timeout: 258 seconds). 03:23:28 -!- jcp|other has quit (Ping timeout: 255 seconds). 03:27:19 -!- jcp|other has joined. 03:29:25 -!- jcp|1 has quit (Ping timeout: 264 seconds). 03:29:27 -!- TeruFSX has quit (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.). 03:29:34 -!- jcp has joined. 03:29:41 -!- BeholdMyGlory has quit (Remote host closed the connection). 03:38:08 -!- pikhq has joined. 03:41:10 -!- pikhq_ has quit (Ping timeout: 244 seconds). 03:41:51 -!- copumpkin has quit (Quit: Computer has gone to sleep.). 03:45:40 hi 03:53:39 -!- quintopia has quit (Ping timeout: 250 seconds). 03:55:38 -!- jcp|1 has joined. 03:56:19 -!- jcp has quit (Ping timeout: 255 seconds). 03:56:58 -!- jcp|other has quit (Ping timeout: 258 seconds). 03:58:58 -!- zzo38 has joined. 03:59:35 -!- quintopia has joined. 03:59:41