Newest at the top
| 2025-04-07 00:39:07 +0200 | jacopovalanzano | (~jacopoval@cpc151911-cove17-2-0-cust105.3-1.cable.virginm.net) (Quit: Client closed) |
| 2025-04-07 00:39:01 +0200 | <|cos|> | koala_man: understandable! i wasn't really expecting it to be possible. the wine-idea is a creative approach! |
| 2025-04-07 00:29:51 +0200 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn |
| 2025-04-07 00:19:15 +0200 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 268 seconds) |
| 2025-04-07 00:18:20 +0200 | Guest90 | (~Guest90@76.78.179.109) (Quit: Client closed) |
| 2025-04-07 00:16:30 +0200 | target_i | (~target_i@user/target-i/x-6023099) (Quit: leaving) |
| 2025-04-07 00:13:56 +0200 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn |
| 2025-04-07 00:11:02 +0200 | <hellwolf> | here you go: https://github.com/hellwolf/haskell-examples/blob/master/2025-04-07-nfa-unboxed-linearity/nfa-perf… will share to discourse tomorrow with a few more details. feedback welcome. |
| 2025-04-07 00:07:24 +0200 | ljdarj1 | ljdarj |
| 2025-04-07 00:07:24 +0200 | ljdarj | (~Thunderbi@user/ljdarj) (Ping timeout: 260 seconds) |
| 2025-04-07 00:04:05 +0200 | ljdarj1 | (~Thunderbi@user/ljdarj) ljdarj |
| 2025-04-07 00:03:12 +0200 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 252 seconds) |
| 2025-04-07 00:00:46 +0200 | gmg | (~user@user/gehmehgeh) gehmehgeh |
| 2025-04-07 00:00:04 +0200 | gmg | (~user@user/gehmehgeh) (Remote host closed the connection) |
| 2025-04-06 23:59:04 +0200 | <yahb2> | *** Exception: hmm ; ; HasCallStack backtrace: ; error, called at <interactive>:53:1 in interactive:Ghci12 |
| 2025-04-06 23:59:04 +0200 | <EvanR> | % error "hmm" >>= error "haw" :: IO Char |
| 2025-04-06 23:58:55 +0200 | <lambdabot> | <IO ()> |
| 2025-04-06 23:58:53 +0200 | <monochrom> | > putStrLn "<IO ()>" |
| 2025-04-06 23:58:44 +0200 | <monochrom> | You will need yahb2 for that. |
| 2025-04-06 23:58:37 +0200 | <EvanR> | lol |
| 2025-04-06 23:58:30 +0200 | <lambdabot> | <IO Char> |
| 2025-04-06 23:58:29 +0200 | <EvanR> | > error "hmm" >>= error "haw" :: IO Char |
| 2025-04-06 23:58:17 +0200 | <lambdabot> | Identity *Exception: haw |
| 2025-04-06 23:58:16 +0200 | <EvanR> | > error "hmm" >>= error "haw" :: Identity Char |
| 2025-04-06 23:58:09 +0200 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn |
| 2025-04-06 23:56:47 +0200 | <c_wraith> | (the big difference here is that pattern-matching a data constructor causes evaluation, but matching a newtype constructor is a no-op) |
| 2025-04-06 23:56:45 +0200 | <dolio> | I'm not certain that if you have multiple binds that the compiler can see, it isn't allowed to reorder them and stuff. |
| 2025-04-06 23:56:44 +0200 | <koala_man> | |cos|: it's hard enough to try to stay compatible with different GHC versions, let alone anything else. if you have Wine then you could try running the windows binary |
| 2025-04-06 23:56:13 +0200 | <c_wraith> | But it had better document how it's implemented. |
| 2025-04-06 23:55:57 +0200 | <c_wraith> | It depends on how >>= is implemented. If it pattern matches its left argument, it does guarantee sequential evaluation to at least that constructor. |
| 2025-04-06 23:55:26 +0200 | <dolio> | I think. |
| 2025-04-06 23:55:04 +0200 | <monochrom> | Also, I used to have much fun with left recursion with Control.Monad.State.Lazy. |
| 2025-04-06 23:54:08 +0200 | <dolio> | Data Identity doesn't really sequence things, either. At least, you aren't specified to be able to rely on it. |
| 2025-04-06 23:53:58 +0200 | forell | (~forell@user/forell) (Ping timeout: 268 seconds) |
| 2025-04-06 23:53:50 +0200 | <monochrom> | OK right right. |
| 2025-04-06 23:53:33 +0200 | <c_wraith> | Eh. It adds the opportunity for a data dependency |
| 2025-04-06 23:53:13 +0200 | <monochrom> | (But Tardis would be an example where data dependency does not imply control flow sequentiency.) |
| 2025-04-06 23:52:55 +0200 | <EvanR> | does it? |
| 2025-04-06 23:52:34 +0200 | <monochrom> | OK, what you want to state instead is that monads necessarily adds data dependency. |
| 2025-04-06 23:52:29 +0200 | <EvanR> | probably a bug in "what is sequentialize even" |
| 2025-04-06 23:51:26 +0200 | <c_wraith> | What's funny is that if you take Identity and change it from newtype to data, leaving all the function definitions the same... it suddenly starts strongly ordering in >>= |
| 2025-04-06 23:50:50 +0200 | <monochrom> | But one is really better off talking about one particular monad rather than general statements. |
| 2025-04-06 23:50:15 +0200 | <EvanR> | e.g. Par is a monad |
| 2025-04-06 23:50:13 +0200 | <monochrom> | What you want to state instead is that there is no guarantee that an arbitrary monad doesn't sequentialize things. |
| 2025-04-06 23:49:27 +0200 | <monochrom> | The Identity Monad doesn't sequentialize things. |
| 2025-04-06 23:49:06 +0200 | <EvanR> | no? |
| 2025-04-06 23:48:54 +0200 | <hellwolf> | monad necessarily sequentialize things. |
| 2025-04-06 23:48:17 +0200 | <EvanR> | linear base has several linear monads |
| 2025-04-06 23:47:36 +0200 | kathryn1024_ | (~kathryn10@user/kathryn1024) (Ping timeout: 264 seconds) |
| 2025-04-06 23:47:34 +0200 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 272 seconds) |