simonpj [Tue, 21 Dec 2004 12:12:40 +0000 (12:12 +0000)]
[project @ 2004-12-21 12:12:40 by simonpj]
Add GHC.Prim to base package modules
panne [Sat, 18 Dec 2004 15:38:08 +0000 (15:38 +0000)]
[project @ 2004-12-18 15:38:08 by panne]
* Fixed Haddock comment. Please: Before committing, everybody should test with
"make html" if the documentation is syntactically OK. It is quite frustrating
to find a broken RPM build after several hours and 99% of the work done...
>:-(
* Fixed isPathSeparator comment.
* Cleanup
ross [Sat, 18 Dec 2004 00:45:27 +0000 (00:45 +0000)]
[project @ 2004-12-18 00:45:27 by ross]
Add system-dependent filename extensions:
exeExtension ("" or ".exe")
objExtension (".o" or ".obj")
dllExtension (".so" or ".dll")
ross [Sat, 18 Dec 2004 00:43:03 +0000 (00:43 +0000)]
[project @ 2004-12-18 00:43:03 by ross]
avoid a warning by defining getDrive only on Windows
simonmar [Fri, 17 Dec 2004 15:12:13 +0000 (15:12 +0000)]
[project @ 2004-12-17 15:12:13 by simonmar]
Add GHC.ConsoleHandler
simonmar [Tue, 14 Dec 2004 13:55:22 +0000 (13:55 +0000)]
[project @ 2004-12-14 13:55:22 by simonmar]
Add Text.Printf
malcolm [Tue, 14 Dec 2004 13:32:38 +0000 (13:32 +0000)]
[project @ 2004-12-14 13:32:38 by malcolm]
Plumb in the new Text/Printf module.
malcolm [Tue, 14 Dec 2004 13:31:44 +0000 (13:31 +0000)]
[project @ 2004-12-14 13:31:44 by malcolm]
For nhc98, get entities from H'98 libraries, rather than redefine them here.
simonmar [Tue, 14 Dec 2004 12:52:03 +0000 (12:52 +0000)]
[project @ 2004-12-14 12:52:03 by simonmar]
Some more Typeable instances, as requested on the ghc-users list a while back.
simonmar [Tue, 14 Dec 2004 12:44:52 +0000 (12:44 +0000)]
[project @ 2004-12-14 12:44:52 by simonmar]
Add instance Typeable MVar
simonmar [Tue, 14 Dec 2004 12:37:28 +0000 (12:37 +0000)]
[project @ 2004-12-14 12:37:28 by simonmar]
Add Lennart's Printf module, extended by me to include hPrintf.
simonmar [Thu, 9 Dec 2004 17:25:15 +0000 (17:25 +0000)]
[project @ 2004-12-09 17:25:15 by simonmar]
Haddock only: import Data.Array.IArray, to avoid linking to hidden Data.Array.Base.
simonmar [Thu, 9 Dec 2004 09:45:39 +0000 (09:45 +0000)]
[project @ 2004-12-09 09:45:39 by simonmar]
Add subRegex & splitRegex
Contributed by: John Goerzen <jgoerzen@complete.org>
simonmar [Wed, 8 Dec 2004 11:05:32 +0000 (11:05 +0000)]
[project @ 2004-12-08 11:05:31 by simonmar]
- Update docs on finalizers: we don't guarantee to run finalizers at all,
and you can't express finalizer ordering using touchForeignPtr.
- consistently rename finaliser to finalizer.
ross [Fri, 3 Dec 2004 14:08:07 +0000 (14:08 +0000)]
[project @ 2004-12-03 14:08:07 by ross]
added a simple-minded implementation of rawSystem for non-GHC implementations.
Also re-instated the doc comment that rawSystem lost in its travels.
ross [Thu, 2 Dec 2004 15:57:04 +0000 (15:57 +0000)]
[project @ 2004-12-02 15:57:02 by ross]
Hugs only: replace the CBITS pragma (files relative to cbits) with
CFILES (files relative to the root of the package).
ross [Thu, 2 Dec 2004 14:52:30 +0000 (14:52 +0000)]
[project @ 2004-12-02 14:52:30 by ross]
add Henrik Nilsson's combinators for composing with pure functions
ross [Wed, 1 Dec 2004 17:46:14 +0000 (17:46 +0000)]
[project @ 2004-12-01 17:46:14 by ross]
tweaks for portability
ross [Wed, 1 Dec 2004 17:45:28 +0000 (17:45 +0000)]
[project @ 2004-12-01 17:45:28 by ross]
markup
simonpj [Tue, 30 Nov 2004 10:04:31 +0000 (10:04 +0000)]
[project @ 2004-11-30 10:04:31 by simonpj]
Use C comment not Haskell comment in package.conf
simonmar [Fri, 26 Nov 2004 16:22:10 +0000 (16:22 +0000)]
[project @ 2004-11-26 16:22:09 by simonmar]
Further integration with the new package story. GHC now supports
pretty much everything in the package proposal.
- GHC now works in terms of PackageIds (<pkg>-<version>) rather than
just package names. You can still specify package names without
versions on the command line, as long as the name is unambiguous.
- GHC understands hidden/exposed modules in a package, and will refuse
to import a hidden module. Also, the hidden/eposed status of packages
is taken into account.
- I had to remove the old package syntax from ghc-pkg, backwards
compatibility isn't really practical.
- All the package.conf.in files have been rewritten in the new syntax,
and contain a complete list of modules in the package. I've set all
the versions to 1.0 for now - please check your package(s) and fix the
version number & other info appropriately.
- New options:
-hide-package P sets the expose flag on package P to False
-ignore-package P unregisters P for this compilation
For comparison, -package P sets the expose flag on package P
to True, and also causes P to be linked in eagerly.
-package-name is no longer officially supported. Unofficially, it's
a synonym for -ignore-package, which has more or less the same effect
as -package-name used to.
Note that a package may be hidden and yet still be linked into
the program, by virtue of being a dependency of some other package.
To completely remove a package from the compiler's internal database,
use -ignore-package.
The compiler will complain if any two packages in the
transitive closure of exposed packages contain the same
module.
You *must* use -ignore-package P when compiling modules for
package P, if package P (or an older version of P) is already
registered. The compiler will helpfully complain if you don't.
The fptools build system does this.
- Note: the Cabal library won't work yet. It still thinks GHC uses
the old package config syntax.
Internal changes/cleanups:
- The ModuleName type has gone away. Modules are now just (a
newtype of) FastStrings, and don't contain any package information.
All the package-related knowledge is in DynFlags, which is passed
down to where it is needed.
- DynFlags manipulation has been cleaned up somewhat: there are no
global variables holding DynFlags any more, instead the DynFlags
are passed around properly.
- There are a few less global variables in GHC. Lots more are
scheduled for removal.
- -i is now a dynamic flag, as are all the package-related flags (but
using them in {-# OPTIONS #-} is Officially Not Recommended).
- make -j now appears to work under fptools/libraries/. Probably
wouldn't take much to get it working for a whole build.
simonmar [Fri, 26 Nov 2004 16:00:06 +0000 (16:00 +0000)]
[project @ 2004-11-26 16:00:06 by simonmar]
more make -j fixing
ross [Fri, 26 Nov 2004 15:03:22 +0000 (15:03 +0000)]
[project @ 2004-11-26 15:03:22 by ross]
fix tycon name strings
simonmar [Fri, 26 Nov 2004 11:58:21 +0000 (11:58 +0000)]
[project @ 2004-11-26 11:58:18 by simonmar]
More 'import Prelude's to help make -j.
simonmar [Thu, 25 Nov 2004 15:01:24 +0000 (15:01 +0000)]
[project @ 2004-11-25 15:01:24 by simonmar]
Add explicit 'import Prelude' to fix dependencies (necessary for make -j)
simonmar [Mon, 22 Nov 2004 14:03:15 +0000 (14:03 +0000)]
[project @ 2004-11-22 14:03:15 by simonmar]
Plug a race condition in the IO manager
simonmar [Mon, 22 Nov 2004 10:26:46 +0000 (10:26 +0000)]
[project @ 2004-11-22 10:26:46 by simonmar]
More fixes to the sample code (thanks to Satnam Singh for pointing out
a problem).
simonmar [Mon, 22 Nov 2004 10:16:42 +0000 (10:16 +0000)]
[project @ 2004-11-22 10:16:42 by simonmar]
Fix example code
ross [Fri, 19 Nov 2004 12:54:12 +0000 (12:54 +0000)]
[project @ 2004-11-19 12:54:12 by ross]
make instance Enum () conform to the Report (though I doubt anyone
will notice)
stolz [Thu, 18 Nov 2004 16:39:54 +0000 (16:39 +0000)]
[project @ 2004-11-18 16:39:54 by stolz]
Push down more feature-tests
tharris [Thu, 18 Nov 2004 09:56:59 +0000 (09:56 +0000)]
[project @ 2004-11-18 09:56:58 by tharris]
Support for atomic memory transactions and associated regression tests conc041-048
sof [Wed, 17 Nov 2004 19:07:38 +0000 (19:07 +0000)]
[project @ 2004-11-17 19:07:38 by sof]
Expose Win32 console event handling to the user.
Added RTS support for registering and delivering console events quite
a while ago (rts/win32/ConsoleHandler.c), but got bored with it before
completing the job. Here's the concluding commit; it does the following:
- new module, base/GHC/ConsoleHandler.hs which supports registering of
console event handlers (the null module on plats other than mingw).
- special handling of aborted async read()s on 'standard input' in
rts/win32/IOManager.c (together with GHC.Conc.asyncRead). See comments
in that IOManager.c as to why this is needed.
[ Any other code that performs blocking I/O on 'standard input' will
need to be tweaked too to be console event handler/signal friendly.]
- for now, disable the delivery of 'close' events (see
rts/win32/ConsoleHandler.c:generic_handler() for reasons why)
Feel free to hoik GHC/ConsoleHandler.hs around the lib hierarchy to wherever
is considered more fitting. Unifying functionality between System.Posix.Signals
and GHC.ConsoleHandler is one (obvious) thing to do.
-- Demonstrating GHC.ConsoleHandler use; win32 only
module Main(main) where
import GHC.ConsoleHandler
import System.IO (hFlush, stdout)
import GHC.Conc (threadDelay)
main :: IO ()
main = do
installHandler (Catch (\ _ -> putStrLn "Caught console event; ignoring" >> hFlush stdout))
loop
where
loop = do
threadDelay 100000
ls <- getLine
putStrLn ls
loop
--
ross [Tue, 16 Nov 2004 23:36:36 +0000 (23:36 +0000)]
[project @ 2004-11-16 23:36:36 by ross]
drop with a negative length should yield the whole list, not []
ross [Tue, 16 Nov 2004 18:02:07 +0000 (18:02 +0000)]
[project @ 2004-11-16 18:02:07 by ross]
spelling in error message
panne [Sun, 14 Nov 2004 20:25:54 +0000 (20:25 +0000)]
[project @ 2004-11-14 20:25:54 by panne]
Tiny refactoring, mostly used as a reminder that '#if blah_platform' tests are
evil and should be replaced by feature-based test (autoconf!) and concentrated
in few modules. System.FilePath is another great example for this, all #ifs
should be replaced by a function handling a possible drive letter + a few
separator constants.
ross [Sun, 14 Nov 2004 12:32:48 +0000 (12:32 +0000)]
[project @ 2004-11-14 12:32:48 by ross]
tweak imports for portability
malcolm [Sun, 14 Nov 2004 10:48:23 +0000 (10:48 +0000)]
[project @ 2004-11-14 10:48:23 by malcolm]
Plumb in new addition System.FilePath.
malcolm [Sun, 14 Nov 2004 10:47:26 +0000 (10:47 +0000)]
[project @ 2004-11-14 10:47:26 by malcolm]
Allow new additions to build with nhc98 too. (I suspect more work
is still needed to make them portable to Hugs as well.)
panne [Sat, 13 Nov 2004 14:37:18 +0000 (14:37 +0000)]
[project @ 2004-11-13 14:37:18 by panne]
Get rid of those ugly WinDoze CR/LF
krasimir [Sat, 13 Nov 2004 08:21:33 +0000 (08:21 +0000)]
[project @ 2004-11-13 08:21:32 by krasimir]
Added the proposed System.FilePath. I also added to System.Directory the
canonicalizePath and findExecutable functions.
simonpj [Fri, 12 Nov 2004 17:07:54 +0000 (17:07 +0000)]
[project @ 2004-11-12 17:07:54 by simonpj]
Missing #include following per-package config changes
simonmar [Fri, 12 Nov 2004 15:14:17 +0000 (15:14 +0000)]
[project @ 2004-11-12 15:14:17 by simonmar]
Note Haskell 98 divergence in isAlpha.
stolz [Fri, 12 Nov 2004 12:34:52 +0000 (12:34 +0000)]
[project @ 2004-11-12 12:34:52 by stolz]
Push FreeBSD-unicode-detection into package
simonpj [Thu, 11 Nov 2004 17:48:49 +0000 (17:48 +0000)]
[project @ 2004-11-11 17:48:49 by simonpj]
Fix version skew
simonmar [Thu, 11 Nov 2004 17:17:30 +0000 (17:17 +0000)]
[project @ 2004-11-11 17:17:30 by simonmar]
remove unnecessary #ifdef
simonmar [Wed, 10 Nov 2004 11:27:54 +0000 (11:27 +0000)]
[project @ 2004-11-10 11:27:54 by simonmar]
Move the compatibility code for rawSystem from libraries/base into
ghc/lib/compat.
simonmar [Tue, 9 Nov 2004 17:02:23 +0000 (17:02 +0000)]
[project @ 2004-11-09 17:02:23 by simonmar]
Make this compile with GHC < 6.3
simonmar [Tue, 9 Nov 2004 15:48:34 +0000 (15:48 +0000)]
[project @ 2004-11-09 15:48:34 by simonmar]
Adding Cabal to GHC, stage 1:
- Add Data.Version library as previously discussed on
libraries@haskell.org.
- import the Cabal library as a package under fptools/libraries,
build & install it by default.
- Instead of importing Cabal's version of ReadP, I added the extra
combinators to Text.ParserCombinators.ReadP. If anyone objects,
please speak up.
I made various changes to Cabal along the way, which I'll try to fold
back into the main Cabal sources in due course. The changes are
roughly these:
- Generic Version stuff removed from Distribution.Version (now in
Data.Version).
- Some modules were rearranged. Distribution.Misc was replaced by
Distribution.License & Distribution.Extension. Distribution.Package
split into D.PackageDescription and D.Package.
- modules under Compat.* renamed to Distribution.Compat.*. Hopefully
I've retained enough compatibility stuff so that this will still build
under nhc98 & Hugs.
- Some elaboration of InstalledPackageInfo, which was previously unused
in Cabal. I've updated the type definition with respect to changes
in PackageDescription, and added a parser/pretty-printer for it.
This is going to be used in the new ghc-pkg.
- Fixed a bug or two that I found along the way.
The next stage will be to integrate GHC's idea of packages with
Cabal's. That means ghc-pkg must grok InstalledPackageInfo, and we
must use all the new information (versions, exposed/unexposed modules etc.).
ross [Sun, 7 Nov 2004 10:17:22 +0000 (10:17 +0000)]
[project @ 2004-11-07 10:17:22 by ross]
move CONST_O_BINARY detection down to libraries/base
panne [Sat, 6 Nov 2004 17:03:43 +0000 (17:03 +0000)]
[project @ 2004-11-06 17:03:43 by panne]
* Issue a warning when compilation/execution fails during FP_READDIR_EOF_ERRNO
and assume a value of '0'.
* Stylistic changes
panne [Sat, 6 Nov 2004 16:34:25 +0000 (16:34 +0000)]
[project @ 2004-11-06 16:34:25 by panne]
'd->d_name' (where d is of type 'struct dirent*') should always return a pointer
to the directory entry's name, regardless if dirent is defined as
struct dirent { ... ; char d_name[...]; }
or
struct dirent { ... ; char* d_name; }
No need for autoconf trickery here.
panne [Sat, 6 Nov 2004 14:15:06 +0000 (14:15 +0000)]
[project @ 2004-11-06 14:15:06 by panne]
* Re-enable large file support
* No need to test for lchown here
panne [Sat, 6 Nov 2004 13:01:20 +0000 (13:01 +0000)]
[project @ 2004-11-06 13:01:18 by panne]
* Changed some '#include "ghcconfig.h"' to '#include "HsBaseConfig.h"' (or added
the latter), tracking the recent autoconf-related changes.
* Continued my crusade against CVS keywords.
* Made CTypes.h self-contained.
* Protected headers against multiple inclusions.
panne [Sat, 6 Nov 2004 11:10:18 +0000 (11:10 +0000)]
[project @ 2004-11-06 11:10:18 by panne]
Moved test for C/ISO types to base package, hopefully testing for all
needed headers first.
panne [Sat, 6 Nov 2004 10:45:47 +0000 (10:45 +0000)]
[project @ 2004-11-06 10:45:46 by panne]
Push down the tests for errno values to the base package. There's really
no need testing for this when e.g. building Happy. :-)
ross [Wed, 27 Oct 2004 15:47:23 +0000 (15:47 +0000)]
[project @ 2004-10-27 15:47:23 by ross]
adjustments to doc comments
simonmar [Mon, 25 Oct 2004 13:47:34 +0000 (13:47 +0000)]
[project @ 2004-10-25 13:47:34 by simonmar]
hDuplicate, hDuplicateTo: add finalizers
simonmar [Mon, 25 Oct 2004 13:40:08 +0000 (13:40 +0000)]
[project @ 2004-10-25 13:40:08 by simonmar]
Doc update: notes on finalization of Handles
ross [Sun, 17 Oct 2004 00:22:03 +0000 (00:22 +0000)]
[project @ 2004-10-17 00:22:03 by ross]
tighten imports
ross [Sun, 17 Oct 2004 00:09:59 +0000 (00:09 +0000)]
[project @ 2004-10-17 00:09:58 by ross]
move some GHC-specific implementations into GHC.*
ross [Sun, 17 Oct 2004 00:08:09 +0000 (00:08 +0000)]
[project @ 2004-10-17 00:08:08 by ross]
markup
simonmar [Thu, 14 Oct 2004 14:58:51 +0000 (14:58 +0000)]
[project @ 2004-10-14 14:58:50 by simonmar]
Threaded RTS improvements:
- Unix only: implement waitRead#, waitWrite# and delay# in Haskell,
by having a single Haskell thread (the IO manager) performing a blocking
select() operation. Threads communicate with the IO manager
via channels. This is faster than doing the select() in the RTS,
because we only restart the select() when a new request arrives,
rather than each time around the scheduler.
On Windows we just make blocking IO calls, we don't have a fancy IO
manager (yet).
- Simplify the scheduler for the threaded RTS, now that we don't have
to wait for IO in the scheduler loop.
- Remove detectBlackHoles(), which isn't used now (not sure how long
this has been unused for... perhaps it was needed back when main threads
used to be GC roots, so we had to check for blackholes manually rather
than relying on the GC.)
Signals aren't quite right in the threaded RTS. In fact, they're
slightly worse than before, because the thread receiving signals might
be blocked in a C call - previously there always be another thread
stuck in awaitEvent() that would notice the signal, but that's not
true now. I can't see an easy fix yet.
simonpj [Wed, 13 Oct 2004 10:24:16 +0000 (10:24 +0000)]
[project @ 2004-10-13 10:24:16 by simonpj]
c_execvpe & pPrPr_disableITimers: remove these in a Windows build.
ross [Tue, 12 Oct 2004 17:45:44 +0000 (17:45 +0000)]
[project @ 2004-10-12 17:45:44 by ross]
unused line
simonmar [Mon, 11 Oct 2004 10:54:57 +0000 (10:54 +0000)]
[project @ 2004-10-11 10:54:57 by simonmar]
genericRaise() hack needed on FreeBSD too.
panne [Sat, 9 Oct 2004 07:51:06 +0000 (07:51 +0000)]
[project @ 2004-10-09 07:51:06 by panne]
Unbreak Hugs by moving pPrPr_disableITimers and execvpe to System.Posix.Internals
(base package) and use it from System.Posix.Process (unix package).
malcolm [Fri, 8 Oct 2004 21:21:08 +0000 (21:21 +0000)]
[project @ 2004-10-08 21:21:08 by malcolm]
Fix layout to Haskell'98.
ross [Fri, 8 Oct 2004 12:07:02 +0000 (12:07 +0000)]
[project @ 2004-10-08 12:07:02 by ross]
Hugs can't handle System.Process yet
ross [Fri, 8 Oct 2004 12:04:48 +0000 (12:04 +0000)]
[project @ 2004-10-08 12:04:48 by ross]
revert previous change, so now these includes don't define PACKAGE_*
simonmar [Fri, 8 Oct 2004 10:28:37 +0000 (10:28 +0000)]
[project @ 2004-10-08 10:28:37 by simonmar]
- make sure we don't overflow the table
- use a checked array index just in case we *do* try to overflow
- fix a harmless off-by-one
stolz [Thu, 7 Oct 2004 13:13:22 +0000 (13:13 +0000)]
[project @ 2004-10-07 13:13:22 by stolz]
Bring declaration of runProcess back in line with prototype:
pid_t (ProcHandle) isn't compatible with int on SunOS.
malcolm [Thu, 7 Oct 2004 09:42:28 +0000 (09:42 +0000)]
[project @ 2004-10-07 09:42:28 by malcolm]
Oops, fix impdecl/fundecl ordering.
dons [Wed, 6 Oct 2004 23:45:37 +0000 (23:45 +0000)]
[project @ 2004-10-06 23:45:37 by dons]
Fix order of PACKAGE_* #undefs. They must appear before HsBaseConfig.h
malcolm [Wed, 6 Oct 2004 15:02:41 +0000 (15:02 +0000)]
[project @ 2004-10-06 15:02:41 by malcolm]
Dummy implementations of openBinaryFile and hSetBinaryMode for nhc98.
ross [Wed, 6 Oct 2004 11:11:35 +0000 (11:11 +0000)]
[project @ 2004-10-06 11:11:34 by ross]
Add getEnvironment from hslibs/lang/SystemExts. This differs from the
System.Posix.Env version in not failing if an entry lacks an '=' sign.
ross [Wed, 6 Oct 2004 10:13:07 +0000 (10:13 +0000)]
[project @ 2004-10-06 10:13:07 by ross]
make the evil PACKAGE_* hacks consistent
simonmar [Tue, 5 Oct 2004 15:56:17 +0000 (15:56 +0000)]
[project @ 2004-10-05 15:56:17 by simonmar]
If we try to use a finalized handle, then throw a useful exception.
Before, the thread would just block forever on the MVar or get a
BlockedOnDeadMVar exception.
The fact that you can actually get into a situation where a finalized
handle can be accessed is arguably a bug itself, but at least now
we'll be able to quickly see when it is happening which should cut
down on debugging time.
simonmar [Tue, 5 Oct 2004 12:51:33 +0000 (12:51 +0000)]
[project @ 2004-10-05 12:51:33 by simonmar]
inputReady foreign import should be "safe.
mthomas [Tue, 5 Oct 2004 07:43:07 +0000 (07:43 +0000)]
[project @ 2004-10-05 07:43:07 by mthomas]
Nonexistent file stopped entire install for libraries.
dons [Sat, 2 Oct 2004 07:34:38 +0000 (07:34 +0000)]
[project @ 2004-10-02 07:34:38 by dons]
Bind raiseSignal to genericRaise, on OpenBSD only, atm.
dons [Sat, 2 Oct 2004 07:14:38 +0000 (07:14 +0000)]
[project @ 2004-10-02 07:14:38 by dons]
Undef PACKAGE_NAME PACKAGE_STRING PACKAGE_BUGREPORT PACKAGE_TARNAME
coming in from ghcconfig.h, so that we can use HsBaseConfig.h versions
without screenfulls of cpp warnings.
malcolm [Thu, 30 Sep 2004 10:01:46 +0000 (10:01 +0000)]
[project @ 2004-09-30 10:01:46 by malcolm]
Add a type signature to help nhc98 out.
simonmar [Thu, 30 Sep 2004 09:42:17 +0000 (09:42 +0000)]
[project @ 2004-09-30 09:42:17 by simonmar]
Haddock fixes
simonmar [Thu, 30 Sep 2004 08:54:00 +0000 (08:54 +0000)]
[project @ 2004-09-30 08:54:00 by simonmar]
Update the signature of execvpe() to match the one in HsBase.
krasimir [Wed, 29 Sep 2004 22:49:38 +0000 (22:49 +0000)]
[project @ 2004-09-29 22:49:36 by krasimir]
Change file extension. We don't need hsc2hs here.
krasimir [Wed, 29 Sep 2004 22:48:07 +0000 (22:48 +0000)]
[project @ 2004-09-29 22:48:06 by krasimir]
compile execvpe only under Unix. The mingw's process.h header also contains
definition for execvpe which conflicts with our definition.
simonmar [Wed, 29 Sep 2004 15:50:53 +0000 (15:50 +0000)]
[project @ 2004-09-29 15:50:51 by simonmar]
Process reorganisation: the System.Process library moves into base,
and System.Cmd is re-implemented in terms of it.
Thanks to Krasimir Angelov, we have a version of System.Process that
doesn't rely on the unix or Win32 libraries. Normally using
unix/Win32 would be the right thing, but since we want to implement
System.Cmd on top of this, and GHC uses System.Cmd, we can't introduce
a bunch of .hsc dependencies into GHC's bootstrap libraries.
So, the new version is larger, but has fewer dependencies. I imagine
it shouldn't be too hard to port to other compilers.
simonmar [Wed, 29 Sep 2004 15:46:53 +0000 (15:46 +0000)]
[project @ 2004-09-29 15:46:53 by simonmar]
Move this to the Attic
ross [Wed, 29 Sep 2004 10:29:15 +0000 (10:29 +0000)]
[project @ 2004-09-29 10:29:13 by ross]
haddock markup
ross [Tue, 28 Sep 2004 23:34:26 +0000 (23:34 +0000)]
[project @ 2004-09-28 23:34:26 by ross]
unbreak for non-GHC implementations
simonmar [Tue, 28 Sep 2004 12:38:55 +0000 (12:38 +0000)]
[project @ 2004-09-28 12:38:55 by simonmar]
Add update, and improve documentation of insert.
simonmar [Tue, 28 Sep 2004 11:29:29 +0000 (11:29 +0000)]
[project @ 2004-09-28 11:29:29 by simonmar]
- export foldl1'
- specialise minimum/maximum for Integer too
simonmar [Tue, 28 Sep 2004 09:02:13 +0000 (09:02 +0000)]
[project @ 2004-09-28 09:02:13 by simonmar]
- Move foldl1 to Data.List
- Provide a strict version of foldl1, namely foldl1'
- Move minimum, maximum to Data.List
- Provide specialised versions of minimum & maximum for Int, which use foldl1'
ross [Mon, 27 Sep 2004 09:04:17 +0000 (09:04 +0000)]
[project @ 2004-09-27 09:04:15 by ross]
update documentation of memory allocation
panne [Wed, 22 Sep 2004 08:37:01 +0000 (08:37 +0000)]
[project @ 2004-09-22 08:37:01 by panne]
* Replace obsolete macro AC_TRY_RUN with AC_RUN_IFELSE
* Same for AC_TRY_LINK and AC_LINK_IFELSE
* Minor cleanup
sof [Mon, 20 Sep 2004 16:35:12 +0000 (16:35 +0000)]
[project @ 2004-09-20 16:35:12 by sof]
openFile: have Haddock comments mention openBinaryFile
panne [Sat, 18 Sep 2004 12:49:59 +0000 (12:49 +0000)]
[project @ 2004-09-18 12:49:59 by panne]
Make autoupdate 2.52 happy, mainly by using the new formats of AC_INIT and
AC_OUTPUT. This has the nice side effect that all "packages" have now a name, a
version, a bug-report address, and a tar name, yielding better output with
"configure --help=recursive". Nuked an unused AC_STRUCT_ST_BLKSIZE test on the
way.
stolz [Wed, 15 Sep 2004 13:51:00 +0000 (13:51 +0000)]
[project @ 2004-09-15 13:51:00 by stolz]
- Remove configure tests for SIG{POLL,INFO,WINCH}:
Testing via #ifdef SIGFOO should be sufficient.
- Change #if HAVE_SIGPOLL to #ifdef SIGPOLL
- Remove SIGINFO/WINCH from package base: they'll reappear in package unix
in System/Posix/Signals.Exts.
ross [Fri, 10 Sep 2004 22:43:20 +0000 (22:43 +0000)]
[project @ 2004-09-10 22:43:20 by ross]
doc tweaks
ross [Fri, 10 Sep 2004 20:38:53 +0000 (20:38 +0000)]
[project @ 2004-09-10 20:38:53 by ross]
docs