June 2013 - Libraries

Proposal: Warn when using Enum instance of Float or Double
by harry 08 Jul '13

08 Jul '13

The Enum instances for Float and Double have dubious semantics which cause endless confusion, e.g. http://stackoverflow.com/questions/13203471/the-math-behind-1-0999999999999…, http://stackoverflow.com/questions/9810002/floating-point-list-generator, http://stackoverflow.com/questions/7290438/haskell-ranges-and-floats, http://stackoverflow.com/questions/10328435/how-to-solve-floating-point-num…, and many more. I would therefore like to propose that the usage of an Enum instance of Float or Double generate a compiler warning, such as "The Enum instance of Float is subject to rounding errors". Deadline: 2 weeks. Pedantic question: Should it be the Enum instance _of_ Float or _for_ Float? -- View this message in context: http://haskell.1045720.n5.nabble.com/Proposal-Warn-when-using-Enum-instance… Sent from the Haskell - Libraries mailing list archive at Nabble.com.

12 20

balanced fold for Data.List
by Brent Yorgey 29 Jun '13

29 Jun '13

How would people feel about adding a function foldb :: (a -> a -> a) -> a -> [a] -> a to Data.List which performs a *balanced* fold of a list? Right now I have this function in the Diagrams.Util module of diagrams-lib but it seems generally useful. Here's a (strawman) implementation: foldb :: (a -> a -> a) -> a -> [a] -> a foldb _ z [] = z foldb f _ as = foldb' as where foldb' [x] = x foldb' xs = foldb' (go xs) go [] = [] go [x] = [x] go (x1:x2:xs) = f x1 x2 : go xs I could also stick it in a separate package but it seems silly to have a whole package for such a small function. Also, does this already exist anywhere? Can it be implemented more simply in terms of existing machinery? etc.? -Brent

2 1

MonadPlus instance for ContT
by Ross Paterson 25 Jun '13

25 Jun '13

Alistair Lynn has proposed the following instance: instance (Monoid r, Monad m) => MonadPlus (ContT r m) where mzero = ContT $ const $ return mempty m `mplus` n = ContT $ \ c -> liftM2 mappend (runContT m c) (runContT n c) but this would also be possible: instance (MonadPlus m) => MonadPlus (ContT r m) where mzero = ContT $ const mzero m `mplus` n = ContT $ \ c -> runContT m c `mplus` runContT n c Is one of them better?

5 9

[ANNOUNCE] Formation of the Core Libraries Committee
by Edward Kmett 25 Jun '13

25 Jun '13

I am pleased to announce the creation of the Haskell Core Libraries Committee. The Core Libraries Committee was formed to give us a maintainer to many of the libraries that previously fell under "GHC HQ" in the core libraries section of the library submission guidelines. As a maintainer, this committee will follow the guidelines of the libraries submission process collectively: http://www.haskell.org/haskellwiki/Library_submissions The main effect of this is that it provides a body that is authorized to actually make decisions about how to proceed with changes to core libraries, and this should help with distributing development effort out into the community in the wake of Simon Marlow's departure from Microsoft Research. From the community's perspective, the creation of this committee provides a way to get more definitive feedback about proposals per the responsiveness guidelines: http://www.haskell.org/haskellwiki/Library_submissions#Responsiveness The inaugural composition of the committee is as follows: * Edward Kmett (chair) * Brent Yorgey * Michael Snoyman * Dan Doel * Doug Beardsley * Joachim Breitner * Mark Lentczner (ex-officio, as the Haskell Platform Maintainer) After the first year, as per the haskell.org committee, the core libraries committee will be seeking self-nominations for two of the members per year on a rolling basis. We'll internally determine which terms are ending after this year and next, and members thereafter will have a three year term, with the possibility to self-renominate. Thank you. We are looking forward to continuing to work with the community that we all know and love.

4 5

Re: Make Prelude's map = fmap
by Alec 25 Jun '13

25 Jun '13

On Sun, Jun 23, 2013 at 2:20 PM, Nikita Volkov <nikita.y.volkov(a)gmail.com>wrote: > That's exactly what I am doing now and the very existence of a dozen of > "prelude" projects on hackage suggests that I'm hardly alone in that. As the author of one of those projects wherein I rather joyously pollute the namespace, I still elected not to do the "map = fmap" thing. I had two main reasons: 1. fmap (or (<$>)) isn't a terribly onerous thing to type. (I did end up exporting (<>), because typing "mappend" is clearly an intolerable burden.) 2. Code that works with fmap in my prelude will work with fmap without my prelude. Changing map means a type error (rather than, at worst, an unbound identifier) should someone elect not to use my prelude one day. While generalizing map doesn't break old code, it makes it somewhat more annoying to backport new code. -A

1 0

Make Prelude's map = fmap
by Nikita Volkov 24 Jun '13

24 Jun '13

Since there's been so many suggestions concerning radically changing the Prelude recently, I'd like to start a poll on a subject. I know that "map" is limited to lists for the beginners, but then we're already making a lot of unbeginnerish changes to Prelude. Also it's not a monad transformer, but just a functor - how can a person learn Haskell without understanding what a Functor is?

12 15

ANN: time-1.4.1
by Ashley Yakeley 24 Jun '13

24 Jun '13

time-1.4.1 is now available from Hackage. <http://hackage.haskell.org/package/time> This update is three contributed improvements: * Derive Typeable instances, instead of writing them out explicitly, from Pedro Magalhães * Fix a bug where getCurrentTimeZone would ignore putEnv "TZ=...", from Russell O'Connor * Use throwErrnoIfMinus1 for gettimeofday, from Dylan Simon Thanks! -- Ashley Yakeley

1 0

Re: Make Prelude's map = fmap (Nikita Volkov)
by Gabriel Gonzalez 24 Jun '13

24 Jun '13

+1 I'm in favor of generalizing `map`. At first I was wary of less monomorphic types, but now I've found that it's not that hard to just lie and pretend the type is more specialized when teaching. Plus, `map` is a much nicer name than `fmap`. Since there's been so many suggestions concerning radically changing the > Prelude recently, I'd like to start a poll on a subject. > I know that "map" is limited to lists for the beginners, but then we're > already making a lot of unbeginnerish changes to Prelude. Also it's not a > monad transformer, but just a functor - how can a person learn Haskell > without understanding what a Functor is?

3 2

Proposal: Add forkOSWithUnmask to Control.Concurrent
by Joey Adams 23 Jun '13

23 Jun '13

We have forkIOWithUnmask[1] and forkOnWithUnmask[2], but we don't have forkOSWithUnmask. This proposal would add it: -- | Like 'forkIOWithUnmask', but the child thread is a bound thread, -- as with 'forkOS'. forkOSWithUnmask :: ((forall a . IO a -> IO a) -> IO ()) -> IO ThreadId forkOSWithUnmask io = forkOS (io unsafeUnmask) For GHC < 7.8, forkOSWithUnmask can be achieved using forkOS and block<http://hackage.haskell.org/packages/archive/base/4.6.0.1/doc/html/Control-E…> [3], but block is removed in GHC 7.8. unsafeUnmask can still be imported from GHC.IO, however. Impact: virtually no compatibility break (just adds a new definition), but we will probably want to update the async package[4] at some point, adding asyncBoundWithUnmask and withAsyncBoundWithUnmask. Such an addition would not require base >= 4.7 in async, as we could use the `block` workaround described above. The biggest objection I can imagine would be that we now have six basic ways to fork a thread: forkIO, forkOn, forkOS, and their fork*WithUnmask variants. Libraries like async end up having to replicate all these. Perhaps in the future, we should consolidate forkIO, forkOn, and forkOS: data ThreadMode = ForkIO | ForkOn Int | ForkBound forkMode :: ThreadMode -> IO () -> IO () forkModeWithUnmask :: ThreadMode -> ((forall a. IO a -> IO a) -> IO ()) -> IO ThreadId But for now, adding forkOSWithUnmask mirrors forkIOWithUnmask and forkOnWithUnmask, which are already in place. Ticket: http://hackage.haskell.org/trac/ghc/ticket/8010 . The title "Add forkOSUnmasked" is a mistake (should be "Add forkOSWithUnmask"), but the patch should be right. [1]: http://hackage.haskell.org/packages/archive/base/latest/doc/html/Control-Co… [2]: http://hackage.haskell.org/packages/archive/base/latest/doc/html/Control-Co… [3]: http://hackage.haskell.org/packages/archive/base/4.6.0.1/doc/html/Control-E… [4]: http://hackage.haskell.org/packages/archive/async/latest/doc/html/Control-C…

1 0

Re: Make Prelude's map = fmap
by wren ng thornton 22 Jun '13

22 Jun '13

On 6/21/13 5:12 PM, Matthias Kilian wrote: > On Fri, Jun 21, 2013 at 04:53:49PM -0400, Carter Schonwald wrote: >> I'd support the map=fmap thing, > > Does it cause any consequences for existing libraries and programs? > Does it break any existing Haskell98 (or Haskell2010) code? > > If not: fine. But if yes: please don't do it. As with the Foldable/Traversable stuff, there's the impediment of ensuring that list fusion still works right. This is hopefully a minor issue and can be easily resolved by inlining/monomorphizing at the appropriate stage; however, it is a technical issue that needs to be addressed by the time we actually implement the change. -- Live well, ~wren

2 1