February 2012 - jhc - Haskell.org

error messages and outdated packages
by Roman Cheplyaka 08 Feb '12

08 Feb '12

Hi John and others, I guess the following message tries to tell me that (<*) is not a member of the Applicative class: [1 of 1] Text.Regex.Applicative.Interface(.......jhci: sigFromClass: Control.Applicative.Applicative [(Control.Applicative.pure, forall f0 a . Control.Applicative.Applicative f0 => a -> f0 a),(Control.Applicative.<*>, forall f0 a b . Control.Applicative.Applicative f0 => f0 (a -> b) -> f0 a -> f0 b)] Control.Applicative.<* Could it be improved? Also, I'd like to update the 'applicative' package to match the current 'base', but before I spend time on that I'd like to check if there are no objections to merging the changes into the jhc repository. -- Roman I. Cheplyaka :: http://ro-che.info/

2 1

darcs patch: clean up foreignptr implementation some. (and 7 more)
by John Meacham 08 Feb '12

08 Feb '12

Mon Feb 6 10:11:22 PST 2012 John Meacham <john(a)repetae.net> * clean up foreignptr implementation some. Mon Feb 6 17:57:19 PST 2012 John Meacham <john(a)repetae.net> * add Jhc.Prim.Rts to access properties of the RTS Mon Feb 6 22:32:02 PST 2012 John Meacham <john(a)repetae.net> * add support for 'Coerce' grin primitive, add atomic allocs in grin. Tue Feb 7 02:55:10 PST 2012 John Meacham <john(a)repetae.net> * make foreign imports work with unboxed IO types, improve primitive operators. Tue Feb 7 03:56:05 PST 2012 John Meacham <john(a)repetae.net> * clean up handling of raw files, keep source file separate when compiling. include useful header files when compiling. Tue Feb 7 17:52:20 PST 2012 John Meacham <john(a)repetae.net> * fixes for distribution. Tue Feb 7 18:00:17 PST 2012 John Meacham <john(a)repetae.net> * added 0.8.0 announcment Tue Feb 7 18:00:26 PST 2012 John Meacham <john(a)repetae.net> tagged 0.8.0

1 0

Re: [jhc] almost to 0.8.0 - feedback welcome
by John Meacham 07 Feb '12

07 Feb '12

On Tue, Feb 7, 2012 at 4:53 AM, Henning Thielemann <lemming(a)henning-thielemann.de> wrote: >> Yeah, I discovered that bug too. Not sure what exactly introduced it but i >> was >> messing with the renamer. I temporarily disblede utility-ht until I get a >> chance to look at it again. > > I am innocent. Yes you are. :) >> Well, it isn't really base 1.0, it has no relation to the hackage/ghc >> version >> other than the unfortunate shared name and the fact they implement similar >> things which can be confusing. There is no correspondence between the >> version >> numbers but it does support many of the interfaces that newer versions of >> ghcs >> base has, so you can generally compile cabal projects by simply deleting >> their >> build-depends and letting jhc loose on it. > > > For me a package that tracks the various simple additions to Data.List and > Control.Monad, while omitting more complicated stuff like concurrency, would > be nice. I would really like to write packages that can be compiled on both > GHC and JHC without cabal switches. Yeah, something like that was my goal with 'base' originally, but I don't think it helped much because it wasn't clear which extensions to the haskell2010 libraries I included or didn't and people just had to try compiling things and see what worked. by supporting 'haskell2010' by default, at least something that compiles today will probably compile unchanged many versions from now. Though, it wouldn't be hard to present the view you want (it is more or less what I would use too), you would just create a library like name: haskell-plus synopsis: haskell 2010 with commonly accepted library extensions exposed-modules: - Ext.Data.Char as Data.Char - Ext.Foreign as Foreign .... then just compile with '-phaskell-plus' to get the extended version of the libraries. My goal is to make this sort of thing easy so a few different useful 'views' like this can be included for common workflows. John

1 0

Re: [jhc] almost to 0.8.0 - feedback welcome
by Henning Thielemann 07 Feb '12

07 Feb '12

On Tue, 7 Feb 2012, John Meacham wrote: > Yeah, I discovered that bug too. Not sure what exactly introduced it but i > was > messing with the renamer. I temporarily disblede utility-ht until I get a > chance to look at it again. I am innocent. > Well, it isn't really base 1.0, it has no relation to the hackage/ghc version > other than the unfortunate shared name and the fact they implement similar > things which can be confusing. There is no correspondence between the version > numbers but it does support many of the interfaces that newer versions of > ghcs > base has, so you can generally compile cabal projects by simply deleting > their > build-depends and letting jhc loose on it. For me a package that tracks the various simple additions to Data.List and Control.Monad, while omitting more complicated stuff like concurrency, would be nice. I would really like to write packages that can be compiled on both GHC and JHC without cabal switches.

1 0

almost to 0.8.0 - feedback welcome
by John Meacham 07 Feb '12

07 Feb '12

Hi all, I am happy to announce that we are almost upon the 0.8.0 release of jhc, which has quite a few changes from the past so I was hoping I could get some people to help test the newest darcs tree and give feedback. changes include. - A brand new library descriptor file format. Now it is a true YAML file. no more of the confusing quasi-cabal files that were not well defined. - the library description files now can affect pretty much any aspect of compilation, so it is much easier to make a self-contained build or for tools like configure/cabal to interface to jhc. They just need to spit out an appropirate yaml file rather than care about passing the right options to jhc in addition to creating a description file. - jhc transparently handles '.hsc' files (assuming you have hsc2hs installed). It even does the right thing when cross compiliing! using the target architectures types rather than the hosts. - the primtive libraries have been re-done, now there is a truely absolutely minimal jhc-prim library that is everything that must exist for jhc to work. everything else and i mean everything is implemented in normal haskell code. Want to create a varient of haskell that has 16 bit ints, ASCII 8 bit Chars and pervasively uses null terminated C style strings? just create a base-tiny library and link it against jhc-prim. - the standalone deriving extension has been implemented, this greatly improves the ability to re-arrange the libraries logically. - support the LANGUAGE pragma and the equivalant field in the library description files with ghc-compatibility. (ghc extensions are mapped to the closest matching jhc flags.) Things I am still working on but want for the release: - C code and header files will be wrapped up into the .hl file as part of a library build. so there is no need to coordinate native code and jhc libraries, installing a library will always be as simple as dropping an hl file in your library directory. (notably, this will allow transparent distribution of 'bytestring' which has C components) - full haskell 2010 compliance, getting closer, rewrote a lot of libraries, still some work to go. - newtype deriving, sort of hacky and unchecked right now but oh so very useful. What I can use help with: - documentation, anything obvious missing, out of date? - external libraries. in the lib/ext directory I have a few. I'd like to expand it and keep track of what people want. I'd like some better automated way to handle these libraries so they need not be integrated with the main darcs tree. probably post 0.8.0. - many error messages produced by jhc are inscrutable. I'd like to start keeping track of the most eggregious ones and working on fixes. John

3 5

darcs patch: allow extra C code to be bundled within .hl files, inc...
by John Meacham 07 Feb '12

07 Feb '12

Mon Feb 6 07:30:22 PST 2012 John Meacham <john(a)repetae.net> * allow extra C code to be bundled within .hl files, include bytestring in the standard external libraries.

1 0

darcs patch: remove old references to the haskell obj... (and 1 more)
by John Meacham 07 Feb '12

07 Feb '12

Mon Feb 6 04:24:44 PST 2012 John Meacham <john(a)repetae.net> * remove old references to the haskell object directory, made everything cache based. added --purge-cache Mon Feb 6 06:03:18 PST 2012 John Meacham <john(a)repetae.net> * pass -D flags to hsc2hs, add CLong and CULong instances, fill out Foreign.ForeignPtr

1 0

darcs patch: remove a lot of unneeded dependencies. (and 4 more)
by John Meacham 07 Feb '12

07 Feb '12

Sun Feb 5 22:56:36 PST 2012 John Meacham <john(a)repetae.net> * remove a lot of unneeded dependencies. Mon Feb 6 00:13:57 PST 2012 John Meacham <john(a)repetae.net> * move arrays to jhc-prim, remove special IORef specific primitives in favor of length 1 arrays Mon Feb 6 01:27:30 PST 2012 John Meacham <john(a)repetae.net> * include location of 'import' declaration for unknown imports, move Data.Int and Data.Word out of jhc Mon Feb 6 01:52:09 PST 2012 John Meacham <john(a)repetae.net> * remove redundant dependencies from 'base' Mon Feb 6 03:22:32 PST 2012 John Meacham <john(a)repetae.net> * update documentation, move Data.Char to base

1 0

darcs patch: don't require 'error' to be in scope for... (and 2 more)
by John Meacham 06 Feb '12

06 Feb '12

Sun Feb 5 22:11:51 PST 2012 John Meacham <john(a)repetae.net> * don't require 'error' to be in scope for where pattern desugaring, process renaming errors before kind inference Sun Feb 5 22:13:04 PST 2012 John Meacham <john(a)repetae.net> * add script to check for and remove unneeded imports from libraries, remove a bunch from lib/jhc/ Sun Feb 5 22:13:14 PST 2012 John Meacham <john(a)repetae.net> * create Jhc.Class.Real to hold more advance numeric classes.

1 0

darcs patch: use quoteName mechanism to pre-rename ce... (and 1 more)
by John Meacham 06 Feb '12

06 Feb '12

Sun Feb 5 16:57:46 PST 2012 John Meacham <john(a)repetae.net> * use quoteName mechanism to pre-rename certain names, fix class instance namespace bug Sun Feb 5 18:28:10 PST 2012 John Meacham <john(a)repetae.net> * seperate --stop parameter from compilation mode

1 0