From rodney at elego.de Tue Jul 1 00:01:05 2014 From: rodney at elego.de (Rodney M. Bates) Date: Tue, 1 Jul 2014 0:01:05 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140630220117.CE2385DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/01 00:01:05 Modified files: cm3/m3-libs/vararray/src/: VarArray.ig VarArray.mg Log message: Consistify comments. From hosking at elego.de Mon Jul 7 23:47:16 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:47:16 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707214716.EF0165DE10C@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:47:16 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Fix fork bug in which there are outstanding pthread mutexes held at the time of the fork (in the form of M3 MUTEXes). The fix is to make M3 MUTEX not identical to pthread mutex. Downside is that MUTEX acquire/release is now more expensive. Upside is that performance improvements to MUTEX can more easily now be incorporated using wait-free techniques (such as lightweight and biased locking). From hosking at elego.de Mon Jul 7 23:54:35 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:54:35 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707215435.CA8BE9BA010@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:54:35 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Need to protect UnlockMutex as well. From hosking at elego.de Mon Jul 7 23:55:13 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:55:13 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707215513.C69F95DEA8F@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:55:13 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Spacing From hosking at elego.de Mon Jul 7 23:56:45 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:56:45 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707215646.093CD5DEA8F@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:56:45 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Spacing From rodney at elego.de Thu Jul 10 02:06:17 2014 From: rodney at elego.de (Rodney M. Bates) Date: Thu, 10 Jul 2014 2:06:17 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140710000617.E8A465DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/10 02:06:17 Modified files: cm3/m3-sys/m3linker/src/: MxIn.m3 Log message: Some groundwork to improve m3linker messages. Some of these are intimidating. They can be improved by putting some identifiers from source code into the messages, but this information is not available in the .mx or .m3x file. Putting it there requires incompatible changes. While improved messages are low priority fixes, right now happens to be a rare opportunity to make such changes to .mx/.m3x files with minimal incompatability problems. The addition of full Unicode support has already necessitated one such change, and its detection mechanism can handle others too. This change just adds the capability to read and ignore some new operators in these files. Later, code to generate and to use these can be added without creating additional version problems. There is still a distant possibility of problems. If you try to use a compiler with Unicode-width WIDECHAR but pre this change, to recompile or link to, code produced by some later compiler, after the new operators are being generated, you will get compile failures. Hopefully, making this change now will make this unlikely. From rodney at elego.de Mon Jul 14 01:44:27 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 1:44:27 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140713234427.D6FB09BA003@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 01:44:27 Modified files: cm3/m3-sys/cm3/src/: Builder.m3 ./: Builder.m3 cm3/m3-sys/m3front/src/misc/: M3Compiler.i3 ./: M3Compiler.i3 cm3/m3-sys/m3front/src/types/: OpaqueType.m3 ./: OpaqueType.m3 cm3/m3-sys/m3linker/src/: Mx.i3 MxCheck.m3 MxIn.i3 MxIn.m3 MxOut.m3 Log message: Add the offending opaque type name to m3linker messages about wrong numbers of revelations. From rodney at elego.de Mon Jul 14 20:07:28 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 20:07:28 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714180728.B64FF5DEA96@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 20:07:28 Modified files: cm3/m3-libs/libm3/src/rw/: Rd.i3 Log message: Document additional procedures that fail if stream is closed. From rodney at elego.de Mon Jul 14 20:23:54 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 20:23:54 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714182354.C93305DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 20:23:54 Modified files: cm3/m3-sys/cm3/src/: Makefile.m3 Log message: Systematically rename identifier that clashed with the enclosing module name Makefile. From rodney at elego.de Mon Jul 14 21:06:40 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 21:06:40 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714190640.B8D955DEA96@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 21:06:40 Modified files: cm3/m3-libs/libm3/src/table/: Table.ig Log message: Document the problem with using a hash function that depends on the bits of a (GC-movable) pointer. From rodney at elego.de Mon Jul 14 21:43:46 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 21:43:46 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714194346.17F795DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 21:43:46 Modified files: cm3/doc/help/cm3/: quake.html Log message: Document that Quake and and or operators are not short-circuit. From rodney at elego.de Sun Jul 20 22:55:42 2014 From: rodney at elego.de (Rodney M. Bates) Date: Sun, 20 Jul 2014 22:55:42 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140720205546.D72105DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/20 22:55:42 Modified files: cm3/m3-libs/m3core/src/: Tag: devel_unicode m3makefile Log message: Put this in cm3-uniboot, just in case it ever gets used. From rodney at elego.de Mon Jul 21 04:19:58 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 21 Jul 2014 4:19:58 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140721021958.9A4CC5DE15C@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/21 04:19:58 Modified files: cm3/m3-sys/m3front/src/misc/: Scanner.m3 Log message: Better error message texts. From hosking at elego.de Sat Jul 26 21:10:00 2014 From: hosking at elego.de (Antony Hosking) Date: Sat, 26 Jul 2014 21:10:00 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140726191000.6C8E39BA00B@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/26 21:10:00 Modified files: cm3/m3-libs/m3core/src/runtime/common/: RTAllocator.m3 Log message: Even untraced objects should be allocated zeroed. From rodney at elego.de Sat Jul 26 22:15:42 2014 From: rodney at elego.de (Rodney M. Bates) Date: Sat, 26 Jul 2014 22:15:42 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140726201543.3204D9BA00B@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/26 22:15:42 Modified files: cm3/m3-libs/m3core/src/runtime/common/: RTAllocator.m3 Log message: calloc needs an element count parameter. From rodney at elego.de Wed Jul 30 04:11:04 2014 From: rodney at elego.de (Rodney M. Bates) Date: Wed, 30 Jul 2014 4:11:04 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140730021104.B04A59BA003@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/30 04:11:04 Modified files: cm3/m3-libs/m3core/src/text/: Text.i3 TextCat.m3 TextClass.i3 TextClass.m3 TextLiteral.m3 Log message: Change the behaviour of Text.GetChar and Text.SetChars when the TEXT parameter contains a character value outside of CHAR. Formerly, they would "CHOP" the value, i.e., just zero out the high order bits of the value. The new behaviour is to raise a range exception. It lies beyond my imagination why anybody would want the old behaviour. It would only make sense when the value is in CHAR, in which case, this change will have no effect. If you suddenly start getting these exceptions from code that was formerly working, I modestly suggest that perhaps it really wasn't. From rodney at elego.de Tue Jul 1 00:01:05 2014 From: rodney at elego.de (Rodney M. Bates) Date: Tue, 1 Jul 2014 0:01:05 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140630220117.CE2385DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/01 00:01:05 Modified files: cm3/m3-libs/vararray/src/: VarArray.ig VarArray.mg Log message: Consistify comments. From hosking at elego.de Mon Jul 7 23:47:16 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:47:16 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707214716.EF0165DE10C@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:47:16 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Fix fork bug in which there are outstanding pthread mutexes held at the time of the fork (in the form of M3 MUTEXes). The fix is to make M3 MUTEX not identical to pthread mutex. Downside is that MUTEX acquire/release is now more expensive. Upside is that performance improvements to MUTEX can more easily now be incorporated using wait-free techniques (such as lightweight and biased locking). From hosking at elego.de Mon Jul 7 23:54:35 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:54:35 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707215435.CA8BE9BA010@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:54:35 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Need to protect UnlockMutex as well. From hosking at elego.de Mon Jul 7 23:55:13 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:55:13 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707215513.C69F95DEA8F@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:55:13 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Spacing From hosking at elego.de Mon Jul 7 23:56:45 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:56:45 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707215646.093CD5DEA8F@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:56:45 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Spacing From rodney at elego.de Thu Jul 10 02:06:17 2014 From: rodney at elego.de (Rodney M. Bates) Date: Thu, 10 Jul 2014 2:06:17 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140710000617.E8A465DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/10 02:06:17 Modified files: cm3/m3-sys/m3linker/src/: MxIn.m3 Log message: Some groundwork to improve m3linker messages. Some of these are intimidating. They can be improved by putting some identifiers from source code into the messages, but this information is not available in the .mx or .m3x file. Putting it there requires incompatible changes. While improved messages are low priority fixes, right now happens to be a rare opportunity to make such changes to .mx/.m3x files with minimal incompatability problems. The addition of full Unicode support has already necessitated one such change, and its detection mechanism can handle others too. This change just adds the capability to read and ignore some new operators in these files. Later, code to generate and to use these can be added without creating additional version problems. There is still a distant possibility of problems. If you try to use a compiler with Unicode-width WIDECHAR but pre this change, to recompile or link to, code produced by some later compiler, after the new operators are being generated, you will get compile failures. Hopefully, making this change now will make this unlikely. From rodney at elego.de Mon Jul 14 01:44:27 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 1:44:27 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140713234427.D6FB09BA003@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 01:44:27 Modified files: cm3/m3-sys/cm3/src/: Builder.m3 ./: Builder.m3 cm3/m3-sys/m3front/src/misc/: M3Compiler.i3 ./: M3Compiler.i3 cm3/m3-sys/m3front/src/types/: OpaqueType.m3 ./: OpaqueType.m3 cm3/m3-sys/m3linker/src/: Mx.i3 MxCheck.m3 MxIn.i3 MxIn.m3 MxOut.m3 Log message: Add the offending opaque type name to m3linker messages about wrong numbers of revelations. From rodney at elego.de Mon Jul 14 20:07:28 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 20:07:28 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714180728.B64FF5DEA96@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 20:07:28 Modified files: cm3/m3-libs/libm3/src/rw/: Rd.i3 Log message: Document additional procedures that fail if stream is closed. From rodney at elego.de Mon Jul 14 20:23:54 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 20:23:54 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714182354.C93305DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 20:23:54 Modified files: cm3/m3-sys/cm3/src/: Makefile.m3 Log message: Systematically rename identifier that clashed with the enclosing module name Makefile. From rodney at elego.de Mon Jul 14 21:06:40 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 21:06:40 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714190640.B8D955DEA96@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 21:06:40 Modified files: cm3/m3-libs/libm3/src/table/: Table.ig Log message: Document the problem with using a hash function that depends on the bits of a (GC-movable) pointer. From rodney at elego.de Mon Jul 14 21:43:46 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 21:43:46 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714194346.17F795DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 21:43:46 Modified files: cm3/doc/help/cm3/: quake.html Log message: Document that Quake and and or operators are not short-circuit. From rodney at elego.de Sun Jul 20 22:55:42 2014 From: rodney at elego.de (Rodney M. Bates) Date: Sun, 20 Jul 2014 22:55:42 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140720205546.D72105DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/20 22:55:42 Modified files: cm3/m3-libs/m3core/src/: Tag: devel_unicode m3makefile Log message: Put this in cm3-uniboot, just in case it ever gets used. From rodney at elego.de Mon Jul 21 04:19:58 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 21 Jul 2014 4:19:58 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140721021958.9A4CC5DE15C@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/21 04:19:58 Modified files: cm3/m3-sys/m3front/src/misc/: Scanner.m3 Log message: Better error message texts. From hosking at elego.de Sat Jul 26 21:10:00 2014 From: hosking at elego.de (Antony Hosking) Date: Sat, 26 Jul 2014 21:10:00 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140726191000.6C8E39BA00B@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/26 21:10:00 Modified files: cm3/m3-libs/m3core/src/runtime/common/: RTAllocator.m3 Log message: Even untraced objects should be allocated zeroed. From rodney at elego.de Sat Jul 26 22:15:42 2014 From: rodney at elego.de (Rodney M. Bates) Date: Sat, 26 Jul 2014 22:15:42 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140726201543.3204D9BA00B@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/26 22:15:42 Modified files: cm3/m3-libs/m3core/src/runtime/common/: RTAllocator.m3 Log message: calloc needs an element count parameter. From rodney at elego.de Wed Jul 30 04:11:04 2014 From: rodney at elego.de (Rodney M. Bates) Date: Wed, 30 Jul 2014 4:11:04 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140730021104.B04A59BA003@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/30 04:11:04 Modified files: cm3/m3-libs/m3core/src/text/: Text.i3 TextCat.m3 TextClass.i3 TextClass.m3 TextLiteral.m3 Log message: Change the behaviour of Text.GetChar and Text.SetChars when the TEXT parameter contains a character value outside of CHAR. Formerly, they would "CHOP" the value, i.e., just zero out the high order bits of the value. The new behaviour is to raise a range exception. It lies beyond my imagination why anybody would want the old behaviour. It would only make sense when the value is in CHAR, in which case, this change will have no effect. If you suddenly start getting these exceptions from code that was formerly working, I modestly suggest that perhaps it really wasn't. From rodney at elego.de Tue Jul 1 00:01:05 2014 From: rodney at elego.de (Rodney M. Bates) Date: Tue, 1 Jul 2014 0:01:05 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140630220117.CE2385DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/01 00:01:05 Modified files: cm3/m3-libs/vararray/src/: VarArray.ig VarArray.mg Log message: Consistify comments. From hosking at elego.de Mon Jul 7 23:47:16 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:47:16 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707214716.EF0165DE10C@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:47:16 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Fix fork bug in which there are outstanding pthread mutexes held at the time of the fork (in the form of M3 MUTEXes). The fix is to make M3 MUTEX not identical to pthread mutex. Downside is that MUTEX acquire/release is now more expensive. Upside is that performance improvements to MUTEX can more easily now be incorporated using wait-free techniques (such as lightweight and biased locking). From hosking at elego.de Mon Jul 7 23:54:35 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:54:35 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707215435.CA8BE9BA010@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:54:35 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Need to protect UnlockMutex as well. From hosking at elego.de Mon Jul 7 23:55:13 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:55:13 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707215513.C69F95DEA8F@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:55:13 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Spacing From hosking at elego.de Mon Jul 7 23:56:45 2014 From: hosking at elego.de (Antony Hosking) Date: Mon, 7 Jul 2014 23:56:45 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140707215646.093CD5DEA8F@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/07 23:56:45 Modified files: cm3/m3-libs/m3core/src/thread/PTHREAD/: ThreadPThread.m3 Log message: Spacing From rodney at elego.de Thu Jul 10 02:06:17 2014 From: rodney at elego.de (Rodney M. Bates) Date: Thu, 10 Jul 2014 2:06:17 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140710000617.E8A465DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/10 02:06:17 Modified files: cm3/m3-sys/m3linker/src/: MxIn.m3 Log message: Some groundwork to improve m3linker messages. Some of these are intimidating. They can be improved by putting some identifiers from source code into the messages, but this information is not available in the .mx or .m3x file. Putting it there requires incompatible changes. While improved messages are low priority fixes, right now happens to be a rare opportunity to make such changes to .mx/.m3x files with minimal incompatability problems. The addition of full Unicode support has already necessitated one such change, and its detection mechanism can handle others too. This change just adds the capability to read and ignore some new operators in these files. Later, code to generate and to use these can be added without creating additional version problems. There is still a distant possibility of problems. If you try to use a compiler with Unicode-width WIDECHAR but pre this change, to recompile or link to, code produced by some later compiler, after the new operators are being generated, you will get compile failures. Hopefully, making this change now will make this unlikely. From rodney at elego.de Mon Jul 14 01:44:27 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 1:44:27 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140713234427.D6FB09BA003@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 01:44:27 Modified files: cm3/m3-sys/cm3/src/: Builder.m3 ./: Builder.m3 cm3/m3-sys/m3front/src/misc/: M3Compiler.i3 ./: M3Compiler.i3 cm3/m3-sys/m3front/src/types/: OpaqueType.m3 ./: OpaqueType.m3 cm3/m3-sys/m3linker/src/: Mx.i3 MxCheck.m3 MxIn.i3 MxIn.m3 MxOut.m3 Log message: Add the offending opaque type name to m3linker messages about wrong numbers of revelations. From rodney at elego.de Mon Jul 14 20:07:28 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 20:07:28 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714180728.B64FF5DEA96@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 20:07:28 Modified files: cm3/m3-libs/libm3/src/rw/: Rd.i3 Log message: Document additional procedures that fail if stream is closed. From rodney at elego.de Mon Jul 14 20:23:54 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 20:23:54 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714182354.C93305DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 20:23:54 Modified files: cm3/m3-sys/cm3/src/: Makefile.m3 Log message: Systematically rename identifier that clashed with the enclosing module name Makefile. From rodney at elego.de Mon Jul 14 21:06:40 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 21:06:40 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714190640.B8D955DEA96@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 21:06:40 Modified files: cm3/m3-libs/libm3/src/table/: Table.ig Log message: Document the problem with using a hash function that depends on the bits of a (GC-movable) pointer. From rodney at elego.de Mon Jul 14 21:43:46 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 14 Jul 2014 21:43:46 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140714194346.17F795DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/14 21:43:46 Modified files: cm3/doc/help/cm3/: quake.html Log message: Document that Quake and and or operators are not short-circuit. From rodney at elego.de Sun Jul 20 22:55:42 2014 From: rodney at elego.de (Rodney M. Bates) Date: Sun, 20 Jul 2014 22:55:42 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140720205546.D72105DEA93@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/20 22:55:42 Modified files: cm3/m3-libs/m3core/src/: Tag: devel_unicode m3makefile Log message: Put this in cm3-uniboot, just in case it ever gets used. From rodney at elego.de Mon Jul 21 04:19:58 2014 From: rodney at elego.de (Rodney M. Bates) Date: Mon, 21 Jul 2014 4:19:58 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140721021958.9A4CC5DE15C@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/21 04:19:58 Modified files: cm3/m3-sys/m3front/src/misc/: Scanner.m3 Log message: Better error message texts. From hosking at elego.de Sat Jul 26 21:10:00 2014 From: hosking at elego.de (Antony Hosking) Date: Sat, 26 Jul 2014 21:10:00 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140726191000.6C8E39BA00B@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: hosking at birch. 14/07/26 21:10:00 Modified files: cm3/m3-libs/m3core/src/runtime/common/: RTAllocator.m3 Log message: Even untraced objects should be allocated zeroed. From rodney at elego.de Sat Jul 26 22:15:42 2014 From: rodney at elego.de (Rodney M. Bates) Date: Sat, 26 Jul 2014 22:15:42 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140726201543.3204D9BA00B@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/26 22:15:42 Modified files: cm3/m3-libs/m3core/src/runtime/common/: RTAllocator.m3 Log message: calloc needs an element count parameter. From rodney at elego.de Wed Jul 30 04:11:04 2014 From: rodney at elego.de (Rodney M. Bates) Date: Wed, 30 Jul 2014 4:11:04 () Subject: [M3commit] CVS Update: cm3 Message-ID: <20140730021104.B04A59BA003@birch.elegosoft.com> CVSROOT: /usr/cvs Changes by: rodney at birch. 14/07/30 04:11:04 Modified files: cm3/m3-libs/m3core/src/text/: Text.i3 TextCat.m3 TextClass.i3 TextClass.m3 TextLiteral.m3 Log message: Change the behaviour of Text.GetChar and Text.SetChars when the TEXT parameter contains a character value outside of CHAR. Formerly, they would "CHOP" the value, i.e., just zero out the high order bits of the value. The new behaviour is to raise a range exception. It lies beyond my imagination why anybody would want the old behaviour. It would only make sense when the value is in CHAR, in which case, this change will have no effect. If you suddenly start getting these exceptions from code that was formerly working, I modestly suggest that perhaps it really wasn't.