- Cvs-hugs - Haskell.org

cvs commit: hugs98/src/unix convert_libraries hugs98/libraries/Hu gs IOExts.hs
by ross＠glass.cse.ogi.edu 11 Aug '03

11 Aug '03

ross 2003/08/11 06:18:25 PDT Modified files: src/unix convert_libraries libraries/Hugs IOExts.hs Log: get unsafeInterleaveIO from fptools/libraries Revision Changes Path 1.75 +1 -1 hugs98/src/unix/convert_libraries 1.15 +0 -4 hugs98/libraries/Hugs/IOExts.hs

1 0

RE: unsafeInterleaveIO
by Simon Marlow 11 Aug '03

11 Aug '03

> On Mon, Aug 11, 2003 at 12:15:57PM +0100, Simon Marlow wrote: > > I'm sure you're right, but I don't understand why an unsafePerformIO > > has to have its own thread at all. > > In Hugs, threads exist only in the IO monad. > > > It's fine for an unsafePerformIO to run to completion before > > anything else happens - in fact that's exactly what I'd expect in a > > single-threaded system. > > Sure, but is that what you'd expect of unsafeInterleaveIO? Certainly! Consider the common use of unsafeInterleaveIO to return the contents of a file lazilly. You would write something like this: getContents = unsafeInterleaveIO f where f = do c <- getChar xs <- unsafeInterleaveIO f return (c:xs) as you can see, it's fine for each invocation of f to run to completion, because it uses nested unsafeInterleaveIO to get the interleaving. Cheers, Simon

1 0

RE: unsafeInterleaveIO
by Simon Marlow 11 Aug '03

11 Aug '03

> That's exactly what's there now, and it doesn't produce interleaving. > Hugs doesn't have a global thread pool -- each instance of > unsafePerformIO > has its own. Once the side computation starts, it runs to completion > before returning to the main thread. Even if it yields, only threads > spawned inside the unsafePerformIO can take over. A global > thread pool > doesn't seem feasible with the current implementation of concurrency > in Hugs. Concurrency is implemented in the libraries; a > blocked thread > is a continuation, and it's not clear how to construct that > continuation in a pure context. I'm sure you're right, but I don't understand why an unsafePerformIO has to have its own thread at all. It's fine for an unsafePerformIO to run to completion before anything else happens - in fact that's exactly what I'd expect in a single-threaded system. Cheers, Simon

2 1

RE: unsafeInterleaveIO
by Simon Marlow 11 Aug '03

11 Aug '03

> I submit that unsafeInterleaveIO hasn't worked on Hugs since > at least July > 2000 (when the implementation of concurrency was changed), and that it > doesn't seem feasible to implement it with the current > concurrency model. > So I propose that we drop it from Hugs. Can't you implement it like this: unsafeInterleaveIO io = return (unsafePerformIO io) I seem to recall some reason why we couldn't do this on GHC, it might have been something to do with the fact that this allows the argument to unsafeInterleaveIO to be performed *before* the unsafeInterleaveIO if the strictness analyser got hold of it, but this won't be a problem for Hugs. Cheers, Simon

3 2

unsafeInterleaveIO
by Ross Paterson 11 Aug '03

11 Aug '03

I submit that unsafeInterleaveIO hasn't worked on Hugs since at least July 2000 (when the implementation of concurrency was changed), and that it doesn't seem feasible to implement it with the current concurrency model. So I propose that we drop it from Hugs.

1 0

(MUTUAL BENEFIT)
by FROM DR IBRAHIM AHMED 10 Aug '03

10 Aug '03

FROM: DR IBRAHIM AHMED . ROYAL TRUST & SECURITIES BANK LIMITED. VICTORIA ISLAND. LAGOS NIGERIA. DATE: 11th AUGUST, 2003 Dear Friend, I know this mail may come to you as a surprise, I got your email address through the web while in search of who will carefully handle this urgent and hitch free deal. I am Dr Ibrahim Ahmed a bank manager with ROYAL TRUST (&SECURITIES) BANK LIMITED Nigeria. I am contacting you based on trust and confidentiality that will be attached to this deal. The Management and the Legal department of our BANK (RTSBL) in a recent meeting recommended that one of my branch depositor (customer) account be declared dormant and confisticated, the deposited fund to the Bank Treasury according to Nigerian Banking and Financial law. He is a foreigner based in Nigeria, unfortunately he is late. He died on a fatal plane crash 2 years ago. He is an account holder in my branch, he owns a dollars account with the sum of Twenty five million four hundred thousand united state dollars ($25.400,000.00) deposited in the account. In fact it happened that since his death, the next of kin of the dollar account holder(the brother) has not show up for the claim and since then nobody has come up to make claim or to say he know anything about the deceased except the attorney. To cut the matter short, the attorney has no contact of any of the deceased relations. This is where I am interested on and this is where I want you to come in. I want you to come in as the relation of the deceased, I will give you the contact of the attorney, you will have to write to him tell him you are a relation to the deceased and that you have found out that Engr. Nicholine Jackson has account with RTSBL which you wants to lay claim through his legal authority. He is going to be happy to assist you. I will be the one to provide the vital documents for the claims of the money . Please include your telephone number when replying this mail and I will give you more information as soon as you indicate your willingness to assist in this transaction. After successful conclussion of this transaction we shall share the money 50/50%. Yours truly, DR IBRAHIM AHMED

1 0

Hugs .msi
by Sven Panne 10 Aug '03

10 Aug '03

Two questions about the .msi file(s) for Hugs on WinDoze: * How exactly are they built? Doing it via Visual Studio sounds like an extensive mouse orgy: http://msdn.microsoft.com/library/default.asp?url=/library/en-us/vsinstal/h… * Not surprisingly, I'd like to ship Hugs on WinDoze with OpenGL/GLUT support. Compiling is the easy part when you have the .h/.def files, but a normal system ships without a glut32.dll, which is annoying for an end user. Would it be possible to include the glut32.dll in the .msi? The missing parts are available at http://www.xmission.com/~nate/glut.html Cheers, S.

1 0

(MUTUAL BENEFIT)
by FROM DR IBRAHIM AHMED 10 Aug '03

10 Aug '03

FROM: DR IBRAHIM AHMED . ROYAL TRUST & SECURITIES BANK LIMITED. VICTORIA ISLAND. LAGOS NIGERIA. DATE: 11th AUGUST, 2003 Dear Friend, I know this mail may come to you as a surprise, I got your email address through the web while in search of who will carefully handle this urgent and hitch free deal. I am Dr Ibrahim Ahmed a bank manager with ROYAL TRUST (&SECURITIES) BANK LIMITED Nigeria. I am contacting you based on trust and confidentiality that will be attached to this deal. The Management and the Legal department of our BANK (RTSBL) in a recent meeting recommended that one of my branch depositor (customer) account be declared dormant and confisticated, the deposited fund to the Bank Treasury according to Nigerian Banking and Financial law. He is a foreigner based in Nigeria, unfortunately he is late. He died on a fatal plane crash 2 years ago. He is an account holder in my branch, he owns a dollars account with the sum of Twenty five million four hundred thousand united state dollars ($25.400,000.00) deposited in the account. In fact it happened that since his death, the next of kin of the dollar account holder(the brother) has not show up for the claim and since then nobody has come up to make claim or to say he know anything about the deceased except the attorney. To cut the matter short, the attorney has no contact of any of the deceased relations. This is where I am interested on and this is where I want you to come in. I want you to come in as the relation of the deceased, I will give you the contact of the attorney, you will have to write to him tell him you are a relation to the deceased and that you have found out that Engr. Nicholine Jackson has account with RTSBL which you wants to lay claim through his legal authority. He is going to be happy to assist you. I will be the one to provide the vital documents for the claims of the money . Please include your telephone number when replying this mail and I will give you more information as soon as you indicate your willingness to assist in this transaction. After successful conclussion of this transaction we shall share the money 50/50%. Yours truly, DR IBRAHIM AHMED

1 0

cvs commit: hugs98/src MkInstal.in hugs98 hugs98.spec
by Jeff Lewis 09 Aug '03

09 Aug '03

lewie 2003/08/09 07:34:13 PDT Modified files: src MkInstal.in . hugs98.spec Log: Don't `install_notes' - this offends the RH9 rpm build because it sees files that were installed, but not included in the RPM (nice sanity check tho') (files that are specified using %doc don't need to be explicitly `install'ed since the %doc rule handles that). Revision Changes Path 1.40 +5 -0 hugs98/src/MkInstal.in 1.26 +1 -1 hugs98/hugs98.spec

1 0

cvs commit: hugs98/src/unix convert_libraries
by ross＠glass.cse.ogi.edu 08 Aug '03

08 Aug '03

ross 2003/08/08 18:15:14 PDT Modified files: src/unix convert_libraries Log: convert OpenGL and GLUT packages iff hopengl was enabled in configure Revision Changes Path 1.74 +6 -3 hugs98/src/unix/convert_libraries

1 0

cvs commit: hugs98/src/unix aclocal.m4
by ross＠glass.cse.ogi.edu 08 Aug '03

08 Aug '03

ross 2003/08/08 13:26:27 PDT Modified files: src/unix aclocal.m4 Log: for older autoconf's, use a dumb definition of AC_HELP_STRING Revision Changes Path 1.17 +5 -0 hugs98/src/unix/aclocal.m4

1 0

release?
by Ross Paterson 08 Aug '03

08 Aug '03

Can we do a release of Hugs over the next month or two?

4 10

cvs commit: hugs98/src/unix configure configure.in hugs98/docs/users_guide Makefile.in Makefile
by ross＠glass.cse.ogi.edu 08 Aug '03

08 Aug '03

ross 2003/08/08 11:30:22 PDT Modified files: src/unix configure configure.in Added files: docs/users_guide Makefile.in Removed files: docs/users_guide Makefile Log: docbook stuff Revision Changes Path 1.68 +177 -13 hugs98/src/unix/configure 1.73 +6 -1 hugs98/src/unix/configure.in

1 0

cvs commit: hugs98/src config.h.in hugs98/src/unix acconfig.h
by ross＠glass.cse.ogi.edu 08 Aug '03

08 Aug '03

ross 2003/08/08 10:11:35 PDT Modified files: src config.h.in src/unix acconfig.h Log: put USE_QUARTZ_OPENGL in acconfig.h, so it will survive autoheader Revision Changes Path 1.29 +6 -6 hugs98/src/config.h.in 1.14 +4 -2 hugs98/src/unix/acconfig.h

1 0

cvs commit: hugs98/docs/users_guide commands.sgml running.sgml
by ross＠glass.cse.ogi.edu 08 Aug '03

08 Aug '03

ross 2003/08/08 05:03:43 PDT Modified files: docs/users_guide commands.sgml running.sgml Log: document defaulting at the evaluation prompt, and a few minor things. Revision Changes Path 1.7 +39 -0 hugs98/docs/users_guide/commands.sgml 1.14 +14 -8 hugs98/docs/users_guide/running.sgml

1 0

cvs commit: hugs98/src MkInstal.in hugs98 hugs98.spec
by Sven Panne 08 Aug '03

08 Aug '03

panne 2003/08/08 04:57:24 PDT Modified files: src MkInstal.in . hugs98.spec Log: Back to the roots: Let's gzip via the .spec file Revision Changes Path 1.39 +0 -1 hugs98/src/MkInstal.in 1.25 +2 -1 hugs98/hugs98.spec

1 0

cvs commit: hugs98 hugs98.spec
by Sven Panne 08 Aug '03

08 Aug '03

panne 2003/08/08 04:51:04 PDT Modified files: . hugs98.spec Log: * Fixed RPM_BUILD_ROOT handling * Synched file list with reality Revision Changes Path 1.24 +10 -5 hugs98/hugs98.spec

1 0

cvs commit: hugs98 hugs98.spec hugs98/src MkInstal.in
by Sven Panne 08 Aug '03

08 Aug '03

panne 2003/08/08 03:49:02 PDT Modified files: . hugs98.spec src MkInstal.in Log: To conform to the filesystem hierarchy standard, install manual pages to %{_mandir}, not simply to %{prefix}/man. Revision Changes Path 1.23 +2 -4 hugs98/hugs98.spec 1.38 +1 -0 hugs98/src/MkInstal.in

2 1

cvs commit: hugs98 RPM.mk
by Sven Panne 08 Aug '03

08 Aug '03

panne 2003/08/08 03:42:29 PDT Modified files: . RPM.mk Log: Added comment about rpmbuild vs. rpm Revision Changes Path 1.13 +4 -4 hugs98/RPM.mk

1 0

cvs commit: hugs98/docs hugs.1.in hugs98/docs/users_guide commands.sgml hugs98/src MkDepend.in hugs.c module.c module.h hugs98/src/dotnet Makefile hugs98/src/msc Makefile hugs98/src/winhugs Makefile
by ross＠glass.cse.ogi.edu 07 Aug '03

07 Aug '03

ross 2003/08/07 06:14:35 PDT Modified files: docs hugs.1.in docs/users_guide commands.sgml src MkDepend.in hugs.c module.c module.h src/dotnet Makefile src/msc Makefile src/winhugs Makefile Log: Change the behaviour of the :browse command (if "all" not specified): formerly listed names defined in the module, now lists exported names. Since the libraries now use imports much more, with the old behaviour things like ":b List" listed no names, which is both useless and confusing to beginners. Now it lists the exports of List (which is also much closer to the output before we switched to the hierarchical libraries). Also added a few rev's to keep the export lists in the order they were specified. Revision Changes Path 1.18 +1 -1 hugs98/docs/hugs.1.in 1.6 +6 -3 hugs98/docs/users_guide/commands.sgml 1.5 +2 -2 hugs98/src/MkDepend.in 1.127 +16 -50 hugs98/src/hugs.c 1.9 +54 -0 hugs98/src/module.c 1.3 +1 -0 hugs98/src/module.h 1.11 +2 -2 hugs98/src/dotnet/Makefile 1.36 +2 -2 hugs98/src/msc/Makefile 1.25 +2 -2 hugs98/src/winhugs/Makefile

1 0

cvs commit: hugs98 Defs.mk hugs98/src/unix convert_libraries test_libraries
by ross＠glass.cse.ogi.edu 05 Aug '03

05 Aug '03

ross 2003/08/05 10:28:44 PDT Modified files: . Defs.mk src/unix convert_libraries test_libraries Log: add newly split QuickCheck package Revision Changes Path 1.8 +2 -1 hugs98/Defs.mk 1.73 +1 -1 hugs98/src/unix/convert_libraries 1.12 +1 -1 hugs98/src/unix/test_libraries

1 0

cvs commit: hugs98/src storage.c
by ross＠glass.cse.ogi.edu 05 Aug '03

05 Aug '03

ross 2003/08/05 10:26:33 PDT Modified files: src storage.c Log: rename strScript (defined but not used) to avoid confusion. Revision Changes Path 1.72 +3 -3 hugs98/src/storage.c

1 0

echoing of stdin
by Ross Paterson 03 Aug '03

03 Aug '03

Hugs currently turns off echoing of stdin before evaluation, and turns it on again afterwards. In partial compensation, getChar echoes chars to stdout (which is wierd if either stdin or stdout is redirected) but getContents doesn't. Is there any reason to retain this behaviour, rather than treating stdin like any other handle? For applications that don't want echoing, we could provide implementations of the System.IO functions hSetEcho, hGetEcho and hIsTerminalDevice.

2 2

cvs commit: hugs98/src/unix config.guess config.sub
by ross＠glass.cse.ogi.edu 01 Aug '03

01 Aug '03

ross 2003/08/01 05:56:42 PDT Modified files: src/unix config.guess config.sub Log: updates Revision Changes Path 1.3 +43 -61 hugs98/src/unix/config.guess 1.3 +26 -7 hugs98/src/unix/config.sub

1 0

cvs commit: hugs98/libraries/Hugs ForeignPtr.hs hugs98/src iomonad.c
by ross＠glass.cse.ogi.edu 01 Aug '03

01 Aug '03

ross 2003/08/01 03:02:00 PDT Modified files: libraries/Hugs ForeignPtr.hs src iomonad.c Log: swapped argument order to addForeignPtrFinalizer, as per latest FFI draft. Revision Changes Path 1.6 +2 -2 hugs98/libraries/Hugs/ForeignPtr.hs 1.52 +5 -5 hugs98/src/iomonad.c

1 0