Home freenode/#haskell: Logs Calendar

Logs: freenode/#haskell

←Prev  Next→
Page 1 .. 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 .. 5022
— 502,152 events total
2020-11-13 16:07:11 × bifunc2 quits (bifunc2@gateway/vpn/protonvpn/bifunc2) (Quit: -)
2020-11-13 16:08:53 <ski> baidobz : fwiw, in general, instead of `sequence (fmap f xs)', you can do `mapM f xs'. but from what you're saying, it sounds like you want to use `ExceptT'. but let's see the code ?
2020-11-13 16:09:39 <tomsmeding> I suspect that the relevant code looks like 'sequence <$> mapM action list'
2020-11-13 16:09:42 <mtothem> How can i import a local haskell module to my xmonad.hs?
2020-11-13 16:09:48 <ski> (something like `runExceptT (mapM (ExceptT . f) xs)')
2020-11-13 16:09:59 → Ariakenom joins (~Ariakenom@h-82-196-111-82.NA.cust.bahnhof.se)
2020-11-13 16:10:07 <ski> mtothem : `import MyModule' ?
2020-11-13 16:10:15 × heatsink quits (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net) (Ping timeout: 265 seconds)
2020-11-13 16:10:57 × quinn quits (~quinn@c-73-223-224-163.hsd1.ca.comcast.net) (Ping timeout: 244 seconds)
2020-11-13 16:11:07 <ski> tomsmeding : hm, sounds likely
2020-11-13 16:11:51 <mtothem> I've been trying to do that. but it doesn't work for some reason. it does in other projects. so maybe xmonad does something differently i'm thinking
2020-11-13 16:12:02 <ski> what is the other file named ?
2020-11-13 16:12:19 <mtothem> `colorscheme.hs` ski
2020-11-13 16:12:36 <ski> try naming it `ColorScheme.hs', and do `import ColorScheme' ?
2020-11-13 16:12:41 → Sanchayan joins (~Sanchayan@171.61.102.109)
2020-11-13 16:13:04 <tomsmeding> (the initial letter of a module name must be a capital letter, the "S" doesn't matter)
2020-11-13 16:13:27 × Sanchayan quits (~Sanchayan@171.61.102.109) (Read error: Connection reset by peer)
2020-11-13 16:13:30 <mtothem> `Could not find module ColorScheme`
2020-11-13 16:13:31 <ski> (unless it's the main module of an executable)
2020-11-13 16:13:49 <tomsmeding> (or you're using a case-insensitive file system, like is the default on macOS)
2020-11-13 16:13:51 <ski> you renamed the file ?
2020-11-13 16:13:54 <ski> yea
2020-11-13 16:14:09 × berberman quits (~berberman@unaffiliated/berberman) (Quit: ZNC 1.7.5 - https://znc.in)
2020-11-13 16:14:38 → berberman joins (~berberman@unaffiliated/berberman)
2020-11-13 16:16:26 → quinn joins (~quinn@c-73-223-224-163.hsd1.ca.comcast.net)
2020-11-13 16:17:11 <baidobz> tomsmeding: ski: This is what I'm playing around with now: https://pastebin.com/A2MhUdC8
2020-11-13 16:18:06 <tomsmeding> baidobz: yes try ski's suggestion with ExceptT
2020-11-13 16:18:07 <baidobz> Basically I would like to short-circuit the processing of the list elements in IO upon any error
2020-11-13 16:18:29 ski low fours tomsmeding
2020-11-13 16:18:29 <tomsmeding> (import Control.Monad.Except from the 'mtl' package)
2020-11-13 16:18:56 <ski> baidobz : but first replace `sequence (map f xs)' by `mapM f xs'
2020-11-13 16:19:16 <tomsmeding> (not necessary, it means the same, but it's neater)
2020-11-13 16:19:20 <baidobz> Great I'll try that, thank you
2020-11-13 16:21:02 <ski> baidobz : you could change the signature to `runFoo :: Text -> ExceptT MyErr IO Text', if you like to
2020-11-13 16:21:52 × zaquest quits (~notzaques@5.128.210.178) (Quit: Leaving)
2020-11-13 16:22:12 <ski> (if it's likely that you'd want to call `runFoo' in a context where you'd want to continue doing some possibly-aborting execution (assuming the `runFoo' invocation didn't abort), then i'd do that)
2020-11-13 16:22:59 <baidobz> Indeed that worked! How come it does not work with a plain map and sequence? Where does the short-circuiting logic come from?
2020-11-13 16:23:04 × christo quits (~chris@81.96.113.213) (Remote host closed the connection)
2020-11-13 16:23:39 <tomsmeding> ExceptT explicitly stops executing remaining monadic actions if a Left occurs
2020-11-13 16:23:43 <baidobz> The bind of MonadError?
2020-11-13 16:24:04 <tomsmeding> normally, the mapM (or in your original version, the 'sequence') would first evaluate all monadic actions, and only when those are complete return the resulting list
2020-11-13 16:24:05 <tomsmeding> yup
2020-11-13 16:24:06 → zaquest joins (~notzaques@5.128.210.178)
2020-11-13 16:24:10 × zaquest quits (~notzaques@5.128.210.178) (Read error: Connection reset by peer)
2020-11-13 16:24:30 <ski> it comes from using the `ExceptT' version of `sequence'/`mapM', which ultimately comes from the `Applicative' / `Monad' instance for `ExceptT' being short-circuiting
2020-11-13 16:25:15 → heatsink joins (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net)
2020-11-13 16:26:30 <ski> (`traverse' (or `sequenceA') is a version of `mapM' (or `sequence') which only requires `Applicative', not `Monad'. back when `sequence'/`mapM' was defined, `Applicative' didn't exist yet, so now you have both those, and `sequenceA'/`traverse')
2020-11-13 16:27:15 <ski> baidobz : `MonadError' doesn't come to play here
2020-11-13 16:27:46 → gproto023 joins (~gproto23@unaffiliated/gproto23)
2020-11-13 16:28:45 → LKoen joins (~LKoen@9.253.88.92.rev.sfr.net)
2020-11-13 16:29:43 → gproto0023 joins (~gproto23@unaffiliated/gproto23)
2020-11-13 16:30:38 × kav quits (~kari@dsl-hkibng42-56733f-225.dhcp.inet.fi) (Ping timeout: 264 seconds)
2020-11-13 16:30:40 → zaquest joins (~notzaques@5.128.210.178)
2020-11-13 16:31:01 → Martinsos joins (~user@cpe-188-129-116-164.dynamic.amis.hr)
2020-11-13 16:31:26 <baidobz> Super interesting. I'm looking at the Haddock and mtl's sources to try and find the relevant definition in the code now
2020-11-13 16:32:05 × gproto023 quits (~gproto23@unaffiliated/gproto23) (Ping timeout: 240 seconds)
2020-11-13 16:32:18 <tomsmeding> https://hackage.haskell.org/package/transformers-0.5.6.2/docs/src/Control.Monad.Trans.Except.html#line-212
2020-11-13 16:32:25 <baidobz> I guess it's in transformers
2020-11-13 16:32:37 <baidobz> Yes! Thanks :)
2020-11-13 16:33:20 <tomsmeding> it's refreshing how readable basic definitions like these are, the short-circuiting is actually evident in the definition of >>= (which is only 5 lines)
2020-11-13 16:33:36 <tomsmeding> contrast with e.g. C++ standard libraries...
2020-11-13 16:34:10 <baidobz> Indeed it's very simple and clear
2020-11-13 16:34:48 <baidobz> I used to do lots of template meta-programming in C++ and I don't miss it
2020-11-13 16:35:01 → Chi1thangoo joins (~Chi1thang@87.112.60.168)
2020-11-13 16:35:17 × conal quits (~conal@66.115.157.159) (Quit: Computer has gone to sleep.)
2020-11-13 16:35:49 <baidobz> Thanks a lot for your quick help!
2020-11-13 16:36:12 → Martinsos_ joins (~Martinsos@cpe-188-129-116-164.dynamic.amis.hr)
2020-11-13 16:36:55 <baidobz> Can I also ask: is there a situation (different Monad?) where my naive approach with sequence would work out similarly due to laziness or does sequence/mapM always evaluate all monadic actions?
2020-11-13 16:37:28 <comerijn> baidobz: The ordering of effects is independent of laziness
2020-11-13 16:37:34 <ski> it always executes all the actions (performing all the effects)
2020-11-13 16:37:59 <comerijn> baidobz: The Monad/Applicative instance of, say, IO are always in sequential order
2020-11-13 16:38:01 <ski> but yes, non-strictness is a separate thing
2020-11-13 16:38:48 <solonarv> well, no. sequence [System.Exit.exit, putStrLn "hello"] will not output hello to the terminal, for a counterexample
2020-11-13 16:38:49 × Martinsos_ quits (~Martinsos@cpe-188-129-116-164.dynamic.amis.hr) (Client Quit)
2020-11-13 16:38:59 <comerijn> baidobz: So yes, sequence and mapM guarantee that all the actions are done by the time the sequence/mapM is done
2020-11-13 16:39:03 <comerijn> solonarv: >.>
2020-11-13 16:39:15 ← Martinsos parts (~user@cpe-188-129-116-164.dynamic.amis.hr) ("ERC (IRC client for Emacs 27.1)")
2020-11-13 16:39:21 <comerijn> solonarv: That's not a contradiction, anyway
2020-11-13 16:39:31 hackage polysemy-methodology-composite 0.1.0.0 - Functions for using polysemy-methodology with composite. https://hackage.haskell.org/package/polysemy-methodology-composite-0.1.0.0 (locallycompact)
2020-11-13 16:39:47 <comerijn> solonarv: exit happens in sequential order with the rest :p
2020-11-13 16:39:49 <solonarv> depending on which Monad instance is being used, the later ones may or may not be exexucted, or evaluated
2020-11-13 16:40:03 <solonarv> (however the *order* does not depend on which instance is in play)
2020-11-13 16:40:29 <ski> oh, good example. i suppose one should say that each action is executed, provided all the previous ones terminated normally
2020-11-13 16:40:44 → nbloomf joins (~nbloomf@2600:1700:ad14:3020:b423:9089:de01:217e)
2020-11-13 16:41:01 hackage polysemy-methodology-composite 0.1.1.0 - Functions for using polysemy-methodology with composite. https://hackage.haskell.org/package/polysemy-methodology-composite-0.1.1.0 (locallycompact)
2020-11-13 16:41:04 → xerox_ joins (~xerox@unaffiliated/xerox)
2020-11-13 16:41:14 <solonarv> they're executed in order, but an action might say "exit immediately, nothing after me gets executed" or even "everything after me happens twice" or "everything after me happens on another thread"
2020-11-13 16:41:32 <ski> (obviously when executing `do foo; undefined; bar', `bar' will not be executed, at least in general)
2020-11-13 16:42:00 <ski> yes, which is exactly what `ExceptT' does (the "nothing after me gets executed" part)
2020-11-13 16:42:31 hackage polysemy-methodology-composite 0.1.2.0 - Functions for using polysemy-methodology with composite. https://hackage.haskell.org/package/polysemy-methodology-composite-0.1.2.0 (locallycompact)
2020-11-13 16:42:41 <solonarv> yep; the other two can be accomplished by Cont / ContT, and probably also by a number of other monads
2020-11-13 16:42:56 <ski> yep (i see how one could accomplish it)
2020-11-13 16:43:25 × avdb quits (~avdb@ip-213-49-124-119.dsl.scarlet.be) (Ping timeout: 240 seconds)
2020-11-13 16:43:28 <baidobz> Great points, that makes sense
2020-11-13 16:43:31 hackage cabal-file 0.1.1 - Cabal file access https://hackage.haskell.org/package/cabal-file-0.1.1 (JensPetersen)
2020-11-13 16:45:46 → Martinsos joins (~user@cpe-188-129-116-164.dynamic.amis.hr)
2020-11-13 16:46:06 ← Martinsos parts (~user@cpe-188-129-116-164.dynamic.amis.hr) ()
2020-11-13 16:46:25 <motte> hekkaidekapus: >[[(1 :: Int) .. 4], [5 .. 13]]
2020-11-13 16:46:29 <ski> in `do n <- [1 .. 9]; d <- [2 .. n-1]; guard (n `mod` d == 0); return (n,d)', for some choice of `n' (namely `1'), everything after `d <- [2 .. n-1]' will be executed zero times (iow, we abort that choice of `n'), and for some other choices of `n', what's after will be executed several times
2020-11-13 16:47:31 <ski> (and then, similarly, for different choices of `d', what comes after the `guard' invocation may be executed once or not at all)
←Prev  Next→
Page 1 .. 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 .. 5022

All times are in UTC.