- Haskell-Cafe - Haskell.org

uninstalling libraries
by Evan Laforge 15 Nov '17

15 Nov '17

For a very long time, I've used a local script to uninstall libraries. Initially it was very simple: use ghc-pkg field to find and remove library-dirs, import-dirs, and haddock-html, and call ghc-pkg unregister. It served well for a long time, but eventually I got tired of copy paste games and extended it to be able to recursively delete dependents too. Unfortunately now it's no longer so simple. The main problem is that the only way I know to find dependents is to ghc-pkg unregister, and see what the error message complains about. That's obviously pretty bad, since by that time you've already unregistered, so it's too late to back out. The non-atomic nature of unregister+rm has always caused problems anyway, since if unregister succeeds, but the remove fails, we are stuck with a partial install. Recursive delete is too error prone, so I reverted that. And then I discovered that hmatrix helpfully includes /opt/local/lib and /usr/local/lib in its library-dirs, so clearly just deleting whatever the package tells me is not very safe. All this has led me to believe uninstalling packages is not so simple, and maybe there should be an actual real way to do it, not everyone hacking up their own dangerous shell scripts. So... is there? Should there be one? Is there interest in one? How do other people uninstall libraries? Is there a better interface to the pkg db than ghc-pkg? Is there a better way to find dependents than ghc-pkg unregister?

7 12

how to make a KnownNat at runtime?
by Johannes Waldmann 15 Nov '17

15 Nov '17

Dear Cafe, I am using type-level numbers for representing dimensions of vectors and matrices (similar to https://github.com/ekmett/ersatz/blob/master/tests/Z001.hs ) Now when I indeed know all the dimensions at compile time, (in the example code, line 20) everything works nicely. But what do I do when the dimensions will only become available at run time? Here is a simplified example: print the null vector where the size is read from the command line - of course in my application "handle" does more, but it essentially has the type given here. I can extend the "case" in the "main" function to include all the values I want to handle. Looks ugly. Is there a better way? - J. {-# language KindSignatures, RankNTypes, LambdaCase, TypeApplications, DataKinds, ScopedTypeVariables #-} import GHC.TypeLits import Data.Proxy import System.Environment main :: IO () main = getArgs >>= \ case [ "2" ] -> handle (Proxy :: Proxy 2) [ "3" ] -> handle (Proxy :: Proxy 3) handle :: forall (n::Nat) . KnownNat n => Proxy n -> IO () handle (_ :: Proxy n) = print (zero :: V n Int) data V (n::Nat) a = V [a] deriving Show zero :: forall (n::Nat) a . (KnownNat n, Num a) => V n a zero = V $ replicate (fromIntegral $ natVal (Proxy :: Proxy n)) 0

4 4

[ANN] Book About Simulation
by David Sorokin 11 Nov '17

11 Nov '17

Hi Cafe, I am delightful to inform that I wrote a book about simulation, Haskell and my Aivika libraries. http://aivikasoft.com/downloads/aivika/aivika.pdf It depends on point of view how to concern with this book. It can be a book about discrete event simulation per se. Also it can be considered as Aivika User’s Manual, or as a book about that how we can apply Haskell for simulation. I would be very grateful if someone could review my book, especially the English text as I’m a Russian native speaker. Best regards, David Sorokin

3 3

[ANN] Haskell Theory Exploration scripts
by Chris Warburton 10 Nov '17

10 Nov '17

Hello all, I thought I'd make a public announcement about some scripts I've built up over the last few years for "exploring" Haskell code. They can be found at: http://chriswarbo.net/git/haskell-te With mirrors at /ipns/QmWv958VzBJQchGjYKiSaxLC9ugrjvXkqMpVrmjp9AonXq and https://github.com/Warbo/haskell-te These use the excellent QuickSpec library from: https://hackage.haskell.org/package/quickspec This empirically (through brute force and QuickCheck testing) conjectures equations about given Haskell function definitions. I think it's a very under-utilised library, perhaps due to a couple of reasons: - It requires hand-holding from the user, e.g. installing packages, importing modules, naming functions, monomorphising types, picking an arity, etc. - It can be very resource-intensive! I know there's some work to address the efficiency concerns in a newer version, but that's not made its way to Hackage, Nixpkgs, etc. yet: https://github.com/nick8325/quickspec My scripts try to address the usability issues. More details are in the README, but basically we can run: quickspec myHaskellDir | renderEqs This will (hopefully) output conjectures found by QuickSpec about the functions defined in myHaskellDir. This automatically finds all function definitions, looks up their types, arities, etc. and invokes QuickSpec in a subprocess (with optional timeouts and memory limits). If we leave off the 'renderEqs', we get an easily-parsed JSON format, which we can feed into other tools. Note that I make *heavy* use of the Nix language, and it's quite hit-or-miss whether a package will work. My hope is that more widespread knowledge of these tools will alleviate some of the existing pain points. In particular, many packages define `Arbitrary` instances for their types, but they do so in their test suites rather than their libraries, so nobody else can make use of them (I assume this is to keep down library dependencies?) Suggestions and contributions are welcome! -- Thanks, Chris Warburton

1 0

Re: [Haskell-cafe] Proposal: Instance apartness guards
by Anthony Clayden 09 Nov '17

09 Nov '17

Surprisinlgy little comment on this proposal. Perhaps it landed when yous were busy elsewhere. I'd like to push it to the committee soon; this is an invitation for more feedback. Richard E wrote a nice brief summary https://github.com/ghc-proposals/ghc-proposals/pull/56#issuecomment-3114214… Thank you AntC ----- Original Message Follows ----- > After years of pondering this idea (in various forms), > and several rounds of discussion on several forums, > I've written it up. > > "This proposal tackles the thorny topic of Overlapping > instances, > for both type classes and Type Families/Associated types, > by annotating instance heads with type-level apartness > Guards. > Type-level disequality predicates appear in Sulzmann & > Stuckey 2002; > in the type-level case selection in HList 2004; > and in various guises in Haskell cafe discussions in > following years. > This proposal builds on the apartness testing implemented > as part of the Closed Type Families work." > > All feedback welcome. > > https://github.com/AntC2/ghc-proposals/blob/instance-apartness-guards/propo… > > AntC

2 1

Functional programming job opportunity
by Nathan van Doorn 08 Nov '17

08 Nov '17

Myrtle is looking for skilled developers to join our growing development team and assist with the development of our core compiler technology. Ideally candidates would have an interest in either efficient hardware design and/ or machine learning models and frameworks. We use the functional language, Haskell, for every aspect of our work and developers will get the chance to be involved in every part of the project. Familiarity and interest with hardware design would be viewed positively but is not essential. Applicants should be familiar with Unix based systems as all our development is done on Ubuntu or NixOS. More information at https://www.myrtlesoftware.com/vacancies/

1 0

ANN: nice-html, an all-around "nice" html templating library, similar but different to lucid, blaze, type-of-html, etc.
by Mike Ledger 08 Nov '17

08 Nov '17

Hi Haskell-Cafe, I've been working on and off on a HTML templating library that is slightly novel for a few months. Here's the good parts: * it's faster than lucid and blaze * it has a valid monad instance * it has another valid monad instance, that is slightly different! * it paramaterises templates by the data they require * the HTML5 combinators don't rely on any highly overloaded/overlapping typeclasses; type errors in nice-html templates tend to be nice monomorphic type mismatches To achieve this, it has a distinct compilation phase (at runtime) for templates. This means the actual rendering function does very little beyond concatenate Text and escape dynamic input. It also severely restricts usage to the built-in combinators available -- at least, for dynamic data -- and can make writing a nice-html template difficult, though littering _ throughout type signatures helps. Check it out! Benchmark results, a README, and an example: https://github.com/TransportEngineering/nice-html Hackage: https://hackage.haskell.org/package/nice-html The only thing on the roadmap right now is to have some nice :-) way to assign JavaScript event handlers to Markup. This is something that I really appreciate when using React.js, so my eventual aim is for nice-html to be like a (more) server-side version of React. Right now, you can keep track of element ids with `Text.Html.Writer.note`, but it's neither a very good nor very useful interface. Cheers, Mike

3 5

Re: [Haskell-cafe] First annual Haskell users survey
by Taylor Fausak 07 Nov '17

07 Nov '17

Today is the last day to fill out the survey! To the over 1,200 of you have already filled it out: thank you! For those of you that haven't, please take 10 minutes today to fill it out. Thanks! https://haskellweekly.news/surveys/2017.html

1 0

[ANNOUNCE] GHC 8.2.2 release candidate 2
by Ben Gamari 07 Nov '17

07 Nov '17

Hello everyone, The GHC team is very pleased to announce the second candidate of the 8.2.2 release of the Glasgow Haskell Compiler. Source and binary distributions are available at https://downloads.haskell.org/~ghc/8.2.2-rc2/ This is the second and hopefully last of two release candidates leading up the final 8.2.2 release. This release fixes approximately fifty bugs present in 8.2.1. These most notably include, * A correctness issue resulting in segmentation faults in some FFI-users (#13707, #14346) * A correctness issue resulting in undefined behavior in some programs using STM (#14171) * A bug which may have manifested in segmentation faults in out-of-memory condition (#14329) * clearBit of Natural no longer bottoms (#13203) * A specialisation bug resulting in exponential blowup of compilation time in some specialisation-intensive programs (#14379) * ghc-pkg now works even in environments with misconfigured NFS mounts (#13945) * GHC again supports production of position-independent executables (#13702) * Better error messages around kind mismatches (#11198, #12373, #13530, #13610) See [1] for the full list. As always, please report any issues you encounter. Happy testing, - Ben [1] https://ghc.haskell.org/trac/ghc/query?status=closed&milestone=8.2.2&col=id…

7 10

Replace list item with list.
by mirone 06 Nov '17

06 Nov '17

Hello everyone! If I have a list A : ["x", "y", "z"] and a list B: ["f", "g"]. Here is my question: What's the simplest way to replace "z" in the list A by list B, and get ["x", "y", "f", "g"] ?

4 4