Diff for /gforth/ChangeLog between versions 1.3 and 1.4

version 1.3, 2001/01/09 16:11:35 version 1.4, 2003/02/03 08:59:28
Line 1 Line 1
   2003-02-02  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * vmgen-ex/support.c, vmgen-ex2/support.c, README, engine/signals.c, test/signals.fs:
           portability bugfixes
           use SA_ONSTACK for all program-generated signals
           updated README
   2003-02-01  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * vmgen-ex/Makefile, vmgen-ex2/Makefile, configure.in, testdist:
           updated version number and date
           minor changes
           * testdist, vmgen-ex/mini.l, vmgen-ex2/mini.l:
           portability bugfixes, minor changes
           * kernel/int.fs, vmgen-ex/Makefile, vmgen-ex/profile.c, vmgen-ex2/Makefile, vmgen-ex2/profile.c, arch/hppa/cache.c, arch/hppa/machine.h, engine/main.c, engine/signals.c, testall, Makefile.in, prims2x.fs:
           portability changes
           bugfixes in prims2x.fs, kernel/int.fs
   2003-01-31  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * arch/hppa/machine.h, configure.in, extend.fs, testall:
           fixed some portability bugs
   2003-01-30  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs: bugfix in prims2x.fs sync line handling
           * vmgen-ex2/Makefile, arch/hppa/cache.c, arch/hppa/machine.h, engine/main.c, prims2x.fs, vmgen-ex/Makefile, configure.in:
           eliminated some (hopefully outdated) hppa special treatments
           rewrote hppa cacheflush
           prims2x can now process CRLF inputs (but the output is partly unixified)
           prims2x can now process several sync lines in sequence
           minor fixes
           * vmgen-ex/Makefile, vmgen-ex2/Makefile, engine/Makefile.in, engine/main.c, Makefile.in, configure.in, testall:
           fixed some portability bugs and other minor bugs
   2003-01-27  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * Makefile.in, configure.in: some changes for cygwin
           * Makefile.in, configure.in, engine/Makefile.in, mkdosmf.sed, mkos2mf.sed:
           replaced configure variable EXE with EXEEXT (provided by autoconf by default)
           * engine/Makefile.in, engine/engine.c, engine/forth.h, engine/main.c, engine/support.c:
           minor changes (64-bit-cleanups)
           * configure.in: updated version number and date
   2003-01-26  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, engine/memcmpc.c:
           removed engine/memcmpc.c (now in support.c)
           removed unused stuff
           * engine/Makefile.in, engine/engine.c, engine/forth.h, engine/memcmpc.c, engine/support.c, kernel/int.fs, test/string.fs, configure.in, ds2texi.fs, extend.fs, float.fs, history.fs, prim, see.fs, stuff.fs, Makefile.in:
           use AC_SYS_LARGEFILE instead of declaring this stuff unconditionally
           moved memcasecmp() into support.c
           eliminated -TEXT (all uses replaced, except within prim)
           converted CAPSCOMP and -TRAILING into high-level words
           folded C code for more complex primitives into support.c
           Fliterals are now decompiled ok
           f.rdp etc. documentation changes
           added test cases for SEARCH
           * Makefile.in, engine/main.c: minor changes
   2003-01-25  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * test/gforth.fs: bugfix in test/gforth.fs
           * ans-report.fs, stuff.fs: converted ans-report.fs to absolute branches
           changed comments for f.rdp etc.
           * Makefile.in, test/gforth.fs: added test cases for f>str-rdp
           * prim, stuff.fs:
           bugfixes and improvements for f.rdp, represent, and friends
   2003-01-24  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prim, startup.fs, stuff.fs: added F.RDP F>STR-RDP F>BUF-RDP
   2003-01-23  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prim, quotes.fs: REPRESENT bugfix
           new escape sequence \q (equivalent to \").
   2003-01-22  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/int.fs, stuff.fs: added ]] ... [[
           fixed compile-only error message
           * INSTALL, configure.in:
           Declared some environment variables in configure.in
           cleaned up PEEPHOLE configuration options (no longer used in C code)
           related documentation changes in INSTALL
           * configure.in, kernel/input.fs, kernel/require.fs, stamp-h.in, stuff.fs:
           disabled long long on PPC (buggy in some gcc versions).
           implemented EXECUTE-PARSING-FILE (new-input only).
           * kernel/input.fs: added EXECUTE-PARSING
           * kernel/require.fs, blocks.fs, kernel/files.fs, kernel/int.fs, source.fs, wf.fs:
           factored out saving and restoring loadfilename (not relevant for new-input)
   2003-01-21  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/gforth.ds, source.fs: loadfilename#>str is now safer
           documented the interaction of markers and ~~ and assertions
   2003-01-20  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/require.fs, source.fs:
           replaced "*a file*" etc. with "*somewhere*"
           * blocks.fs, kernel/files.fs, kernel/input.fs, kernel/int.fs, kernel/require.fs, kernel/vars.fs, source.fs, wf.fs:
           replaced loadfilename# with loadfilename (bugfix)
           * kernel/int.fs, kernel/io.fs, kernel/main.fs, kernel/nio.fs, kernel/paths.fs, kernel/require.fs, kernel/toolsext.fs, kernel/vars.fs, engine/signals.c, engine/support.c, kernel/args.fs, kernel/basics.fs, kernel/comp.fs, kernel/cond.fs, kernel/files.fs, engine/Makefile.in, engine/engine.c, engine/forth.h, engine/main.c, arch/mips/machine.h, arch/power/machine.h, arch/sparc/machine.h, doc/gforth.ds, arch/386/machine.h, arch/alpha/machine.h, arch/m68k/machine.h, prims2x.fs, search.fs, startup.fs, struct.fs, vt100key.fs, wordinfo.fs, configure.in, cross.fs, debug.fs, doskey.fs, environ.fs, extend.fs, glocals.fs, hash.fs, look.fs, machpc.fs.in, Makedist.in, Makefile.in, blocks.fs:
           undid changes to copyright notices
   2003-01-19  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * kernel/int.fs, kernel/io.fs, kernel/main.fs, kernel/nio.fs, kernel/paths.fs, kernel/require.fs, kernel/toolsext.fs, kernel/vars.fs, engine/forth.h, engine/main.c, engine/signals.c, engine/support.c, kernel/args.fs, kernel/basics.fs, kernel/comp.fs, kernel/cond.fs, kernel/files.fs, kernel/input.fs, engine/engine.c, doc/gforth.ds, engine/Makefile.in, arch/386/machine.h, arch/alpha/machine.h, arch/m68k/machine.h, arch/mips/machine.h, arch/power/machine.h, arch/sparc/machine.h, look.fs, machpc.fs, machpc.fs.in, prims2x.fs, search.fs, startup.fs, struct.fs, vt100key.fs, wordinfo.fs, blocks.fs, configure.in, cross.fs, debug.fs, doskey.fs, environ.fs, extend.fs, glocals.fs, hash.fs, Makedist.in, Makefile.in:
           Updated copyright notices
           Added stack effects to kernel/input.fs
   2003-01-18  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * configure.in, kernel/int.fs:
           Added intptr_t and int128_t to types to be checked
           Changed boot message copyright date
   2003-01-14  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/main.c: flush_to_here now not called on superblock ends
           * engine/forth.h, engine/main.c, kernel/comp.fs, code.fs, configure.in:
           Icache-flushing is now not performed on generating a NEXT, just on the
              end of a definition through finish-code.
           Eliminated MIPS linker testing (not necessary with hybrid D/ITC).
           --no-dynamic is default on systems that don't define FLUSH_ICACHE.
   2003-01-10  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * Makefile.in, engine/Makefile.in, kernel/comp.fs, kernel/int.fs, prim:
           eliminated all greedy static superinstruction stuff (we will use something
             else for static superinstructions eventually)
           * engine/forth.h, engine/main.c, kernel/int.fs, prim:
           moved DOES-HANDLER! from prim to kernel/int.fs
           eliminated MAKE_DOES_HANDLER
           * ans-report.fs, engine/main.c, prim:
           deleted unused prims (HASHKEY) and COMPILE-PRIM
   2003-01-08  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/int.fs, prim:
           moved CODE-ADDRESS! DOES-CODE! /DOES-HANDLER from prim to kernel/int.fs
           * engine/forth.h, kernel/getdoers.fs, kernel/int.fs, prim:
           moved >code-address and >does-code from prim to kernel/int.fs
           * kernel/int.fs, prim: moved >BODY from prim to kernel/int.fs
           * prim: kludged NEWLINE portability bug
   2003-01-07  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/forth.h, engine/main.c, engine/threaded.h, prim:
           changes to eliminate gcc warnings
           * arch/mips/machine.h, arch/power/machine.h, arch/sparc/machine.h, engine/signals.c, arch/386/machine.h, arch/alpha/machine.h, arch/hppa/machine.h, arch/m68k/machine.h, configure.in, Makefile.in:
           Some changes for portability to AIX and Darwin
           Eliminated the (non-hybrid) direct threading stuff from arch/*/machine.h
   2003-01-06  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * backtrac.fs, look.fs, see.fs, test/other.fs:
           Changed interfaces of PRIM>NAME THREADED>NAME to be like >NAME.
           * Makefile.in, configure.in:
           added dependence on engine/config.h (for automatic autoheader call)
           removed most feature test macro definitions (problems with Darwin)
           dynamic is now default for all gcc versions
           various changes to configure.in (for Darwin, ia64, m68k, generic)
   2003-01-04  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * test/other.fs, backtrac.fs, look.fs, see.fs:
           bugfix: LOOK works again for xts of primitives
           new words: threaded>xt, prim>name, threaded>name (stack effect not final)
   2003-01-03  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/forth.h, engine/main.c, machpc.fs, machpc.fs.in:
           undid the unnecessary/broken stuff Bernd has done
           moved a change from machpc.fs to machpc.fs.in
           removed machpc.fs (generated file)
   2003-01-03  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * engine/forth.h, kernel/basics.fs, kernel/main.fs, Makefile.in, arch/4stack/mach.fs, arch/4stack/prim.fs, arch/4stack/relocate.fs, arch/misc/mach.fs, arch/misc/prim.fs, cross.fs, machpc.fs:
           Made 4stack port of Gforth EC work again
   2003-01-03  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * arch/generic/machine.h, arch/power/machine.h, engine/threaded.h:
           PPC fixes and speedups
   2003-01-02  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * Makefile.in, engine/main.c, engine/threaded.h, kernel/accept.fs, look.fs:
           Fixed look problem
           Added benchmarking of all engines
   2003-01-02  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/main.c, engine/support.c, prim:
           Alpha bug workaround (disabled ALIGN_CODE)
           fixed bugs in F>D and D>F
           * engine/forth.h:
           support for small (cell-sized or smaller) off_t on machines without proper
             double-cell type; typically needed on 64-bit machines.
           * configure.in, engine/forth.h:
           feature test macros now defined in configure.in (so the tests see them) instead
             of engine/forth.h
           Testing for size of off_t
   2003-01-01  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * kernel/cond.fs, kernel/input.fs, cross.fs, prim, see-ext.fs, see.fs:
           Cleaned up prim file (only one soft of branches, grouping)
           Added auto-expanding tib for file reading - the terminal line length remains
   2002-12-31  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/Makefile.in:
           dependence for engine/prim_grp.i in engine/Makefile.in
           * Makefile.in, engine/engine.c, engine/main.c, prims2x.fs:
           restored the old engine/prim_lab.i format (for vmgen compatibility)
           introduced engine/prim_grp.i for the new use
           * prims2x.fs:
           vmgen: introduced STACK-ACCESS-TRANSFORM to support upward-growing stacks.
   2002-12-30  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * prim: Forth replacement for absolute branch
   2002-12-28  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * engine/engine.c, engine/forth.h, engine/main.c, engine/threaded.h, filedump.fs, prim, prims2x.fs, Makefile.in, comp-i.fs, cross.fs:
           First inclusion of group-based primitive tokens (no reordering happend yet)
   2002-12-27  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/main.c, comp-i.fs, cross.fs: increased image version
           * prim, prims2x.fs:
           prims2x: INST_TAIL now produces NEXT_P2 again by default
           * Makefile.in, engine/Makefile.in, prims2x.fs, vmgen-ex/mini.h, vmgen-ex2/mini.h:
           some fixes to make make dist work
           bugfixes to make building outside srcdir work
           minor prims2x.fs bugfix (for vmgen-ex)
           vmgen-ex[2] bugfix
           * configure.in: modernized configure.in (with the help of autoupdate)
           * doc/version.texi.in, engine/Makefile.in, engine/main.c, kernel/args.fs, doc/gforth.ds, mkos2mf.sed, vmgen.in, Makefile.in, README, acinclude.m4, configure.in, extend.fs, gforth.lsm, gforthmi.in, mkdosmf.sed:
           replaced some automake macros with autoconf macros; related changes in
             autoconf variables (e.g., @VERSION@ -> @PACKAGE_VERSION@)
           gforth-prof and gforth-native are now not built by default
           modernized "Automatic Remaking" section of Makefile.in
           Updated bug reporting instructions
           Eliminated use of old bug reporting email address
   2002-12-26  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * code.fs, kernel/basics.fs, kernel/comp.fs, kernel/int.fs, stuff.fs, Makefile.in:
           made CODE and ;CODE work again
           * see-ext.fs, see.fs, startup.fs:
           SEE now works again for control structures and strings
   2002-12-25  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/comp.fs, see.fs: worked on SEE
           COMPILE, now produces LIT@ and LIT+ again
           * blocks.fs, kernel/paths.fs:
           Limit blocks files to 2GB (to avoid catastrphic interaction with gforth<0.6.0)
           blocks bugfixes: OPEN-BLOCKS tries to avoid deleting existing blocks files
             SAVE-BUFFERS stack underflow fixed
   2002-12-24  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/forth.h, engine/io.c, engine/main.c, engine/memcmpc.c, engine/profile.c, engine/signals.c, configure.in, engine/engine.c, prim:
           Gforth now supports large files (>2GB) on small machines (32-bits/cell).
           forth.h now asks for all kinds of POSIX, X/Open, and GNU support.
           rearranged include files such that forth.h precedes the system files.
           * elisp-comp, engine/main.c, install-sh, missing, mkinstalldirs, configure.in:
           some configure.in cleanups.  In particular, I replaced all echos with
             AC_MSG_... macros.
           The gcc version (for disabling dynamic code) is now checked in configure (not
             when compiling main.c)
           updated elisp-comp install-sh missing mkinstalldirs from autoconf-2.54
   2002-12-23  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, configure.in:
           the skipping code for engine2 is now checked with configure
   2002-12-22  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/threaded.h:
           cfa is now killed in NEXT_P2 in some direct threading schemes (for better
             register allocation)
   2002-12-21  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * extend.fs, gforth.el, look.fs: gforth.el: LEAVE ?LEAVE
           rest: made compilable with gforth 0.5.0
   2002-12-21  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * arch/386/machine.h, engine/engine.c:
           minimized differences between engine and engine2 to avoid register
             allocation differences
           disabled USE_TOS on 386 without FORCE_REG (otherwise sp is spilled)
           * BUILD-FROM-SCRATCH: some more BUILD-FROM-SCRATCH refinements
           * BUILD-FROM-SCRATCH, Makefile.in: refined BUILD-FROM-SCRATCH
   2002-12-19  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prim: compile-prim is now obsolete
           * kernel/comp.fs, engine/main.c:
           "compile," now uses compile-prim1 instead of compile-prim
           disablesd compile-prim
           * engine/Makefile.in, engine/engine.c:
           each engine is now compiled separately; unfortunately, the problem with
             lots of differences between engine and engine2 in gforth-fast on i386 without
             force-reg persists
           * engine/engine.c, engine/forth.h, engine/main.c:
           moved global/static vars from engine.c into main.c (to avoid duplicating them)
           * engine/support.c, doc/gforth.ds, engine/Makefile.in, engine/engine.c:
           moved all functions except engine() out of engine.c into (new file) support.c
           minor documentation changes (suggested by John A. Peters <japeters@pacbell.net>)
   2002-12-16  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * prim: Made the hash shift array static
   2002-12-15  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * Makefile.in, engine/main.c:
           "make check" now checks all engines and some variations
           bugfix in forget-dyncode
           * engine/engine.c, engine/rint.c, test/float.fs, configure.in, float.fs, prim:
           fround is a primitive again
           rint replacement function written
   2002-12-14  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * cross.fs, kernel/quotes.fs, machpc.fs, machpc.fs.in:
           Changed string compilation to not use the return stack to skipp the string
           added control-rack property to specify if the return stack does point to
           control-relevant information
   2002-12-13  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/main.c: bugfixes in DECOMPILE-PRIM (no crash on non-address,
              find @ instead of >code-address etc.)
           * engine/main.c, kernel/int.fs, look.fs, see.fs, simp-see.fs:
           implemented DECOMPILE-PRIM and use it in (LOOK); SIMPLE-SEE works again
           minor bugfix in SIMPLE-SEE
           HEAD? now delivers 1 on likely-false positives
           * engine/forth.h, engine/main.c, extend.fs, prim:
           dynamically generated code is now (sort of) released on executing a MARKER
   2002-12-08  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/main.c: code_area now allocated on demand (unlimited size)
           some refactoring in dynamic code generation
           * engine/Makefile.in, engine/main.c, Makefile.in, prims2x.fs:
           now engine/prim_num.i is generated, and used in main.c (for main-native.o)
   2002-12-07  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * hash.fs:
           the limit for hash table doubling is now an average 2 entries/bucket
           * engine/main.c: bugfix ("gforth" crashed, "./gforth" didn't)
   2002-12-05  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * ToDo-native, except.fs, float.fs, see-ext.fs:
           TRY should now be partially ready forth gforth-native
           FLITERAL should be ready for gforth-native
   2002-12-04  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * assert.fs, debugs.fs, doc/gforth.ds, source.fs:
           source positions are now passed through literals instead of using
             return stack tricks; appropriate changes in the users of source.fs.
           * kernel/quotes.fs, quotes.fs, see.fs, extend.fs:
           eliminated all uses of "lit (abort") (s") (.") outside cross.fs, except one
           eliminated (c"); cliteral is now in the kernel.
   2002-12-03  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/comp.fs, kernel/quotes.fs, prim: New word: "MEM,"
           new implementations of CLITERAL (independent of threaded code size) and
              SLITERAL (also can deal with long strings now)
   2002-11-24  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * acconfig.h:
           Deleted acconfig.h after cleaning up configure (no longer needed)
           * configure.in: Cleaned up configuring
           * acconfig.h, configure.in, engine/main.c:
           Added check for directory separation
           * hash.fs:
           Added tracking of the number of hash table entries and doubling the hash
           size when this number exceeds the number of buckets
   2002-11-24  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/threaded.h, kernel/basics.fs, kernel/comp.fs, kernel/quotes.fs, arch/386/machine.h, engine/Makefile.in, engine/engine.c, engine/forth.h, engine/main.c, prims2x.fs, Makefile.in, prim:
           new engine gforth-native (works on 386 arch only for now).
             appropriate changes in control-flow instructions in prim
             new primitives SET-NEXT-CODE and CALL2 (not necessary for the other engines)
             new primitives COMPILE-PRIM1 and FINISH-CODE
             prims2x.fs now produces IMMARG(...) macros for initializing immediate args
             prims2x.fs: changes in some of the output-c-tail words (goes with the
                changes in the control-flow words).
             appropriate changes in engine.c
             engine.c: rewrite of check_prims, support for gforth-native (NO_IP)
             threaded.c: support for NO_IP
             various kernel files: started to eliminate return stack manipulations for
               embedding data (e.g. string literals); incomplete.
           dynamic superinstructions now use LABEL2 instead of IS_NEXT_JUMP
           FORCE_REG has no effect if DOUBLY_INDIRECT (gcc-2.95.1 crashes otherwise;
              it's unclear which change provoked this).
   2002-11-10  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, engine/forth.h, engine/main.c:
           No engine2 and other dynamic stuff is compiled for gforth-itc, -iditc, -prof
   2002-11-02  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * test/float.fs, Makefile.in, engine/signals.c, float.fs:
           New file test/float.fs
           FROUND-OFFSET now defined without f**
           bugfix in signals (disabled SA_ONSTACK in most cases to allow graceful_exit).
   2002-10-27  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/signals.c, prim, float.fs:
           FROUND is now implemented in high-level Forth.
             The primitive is now called "(FROUND)".
   2002-10-26  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/signals.c: work around missing MacOS X FPE_... si_codes
   2002-10-25  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/main.c, engine/signals.c, engine/forth.h:
           Gforth now survives destruction of the C stack pointer (it now uses
             sigaltstack() to set up signal handling on another stack).
   2002-10-12  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * peeprules.vmg: emptied peeprules.vmg again
           * peeprules.vmg, prims2x.fs:
           bugfix (use of inst-pointer-update left stack items)
           bugfix (now no debugging output at superinstruction end)
           refactored output-c-tail...
           * prims2x.fs:
           In superinstructions, move stores into the last part writing to the stack item.
           * prims2x.fs:
           Stack loads in superinstructions are now delayed until the part that needs them
   2002-10-07  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * config.guess, config.sub:
           config.guess and config.sub from autoconf-2.54
   2002-10-04  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * engine/main.c, configure.in, prim, prims2x.fs: Fixes for GCC 3.2
   2002-10-02  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * arch/misc/asm.fs: warn if end-label / end-code is missing
           * arch/misc/prim.fs: fix: missing end-label statements
           * arch/misc/optcmove.fs: added
           * arch/misc/mach.fs, arch/misc/sim.fs: romable feature
           * arch/misc/asm.fs: added txd io stuff
   2002-09-26  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * cross.fs:
           If abranch option is on we compiler abranch, a?branch, a(do), ...
           (instead of branch, ...) and use absolute addresses for branch
           * machpc.fs.in, machpc.fs: added abranch option
   2002-09-24  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/cond.fs, except.fs:
           replace all uses of relative branches with absolute branches
             Exception: cross still produces relative branches in the kernel.
           * prim: added absolute versions of all relative branches
             I am not sure if I got conditional compilation right.
             I left the Forth code for acondbranch alone because I don't know
               what may be used there.
           * prim: changed lit+ and lit@ into simple primitives
   2002-09-22  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs: added include-skipped-insts to prims2x.fs
           * Makefile.in, prims2x.fs:
           Changed creation of engine/super.i into engine/costs.i, which contains
              all primitives, not just superinstructions; costs now work (mostly)
           * doc/vmgen.texi, Makefile.in, prims2x.fs:
           Added generation of engine/supers.i
           Deleted generation of kernel/peephole.fs
   2002-09-19  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/Makefile.in, engine/main.c:
           Disabled dynamic native code generation for gcc >=3.1
           added flag --dynamic
   2002-09-15  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * test/other.fs, engine/main.c:
           Bugfixes: compile_prim now flushes the I-cache; fixed test case
   2002-09-14  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * look.fs, see.fs:
           bugfix: (complex) SEE now works (output quality could be improved, though).
           * look.fs: bugfix: simple decompilation of static hybrid threaded code
           * blocks.fs, compat/strcomp.fs, engine/forth.h, startup.fs:
           fixed bugs in compat/strcomp.fs and its usage
           * kernel/comp.fs, kernel/paths.fs, kernel/require.fs, compat/strcomp.fs, kernel/args.fs, kernel/basics.fs, savesys.fs, startup.fs, stuff.fs, wf.fs, fi2c.fs, glocals.fs, history.fs, httpd.fs, locate.fs, oof.fs, prims2x.fs, proxy.fs, quotes.fs, Makefile.in, cross.fs, ds2texi.fs:
           Added compat/strcomp.fs, introducing STR=, STRING-PREFIX?, and STR<
           replaced most occurences of COMPARE with STR= and STRING-PREFIX?
   2002-09-01  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/vmgen.texi, engine/main.c, configure.in: changed version
           disabled dynamic code generation for now (problems with gcc-3.2).
           documentation changes
   2002-08-28  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/vmgen.texi, prim, prims2x.fs:
           Made store optimization optional (default off), and documented it
           * engine/main.c, engine/Makefile.in, engine/engine.c, engine/forth.h, arch/386/machine.h, arch/generic/machine.h, doc/gforth.ds, testdist:
           gforth now stores a local ip to saved_ip instead of using a global ip
             associated changes in register allocation, threading, documentation
           testdist now also tests vmgen and vmgen-ex
   2002-08-22  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * README, README.vmgen, configure.in, cross.fs, doc/vmgen.texi, Makefile.in:
           Getting ready for the Vmgen release
           New snapshot dates, various documentation changes, Makefile and configure fixes
   2002-08-21  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/vmgen.texi: documentation changes
           * vmgen-ex2/Makefile, vmgen-ex2/engine.c, vmgen-ex/Makefile, vmgen-ex/engine.c:
           made all threading schemes run in vmgen-ex*
   2002-08-20  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * Makefile.in, doc/vmgen.texi, prims2x.fs, vmgen.in:
           prims2x.fs now outputs #line directives at the end of the user C code
           documentation changes
           * vmgen-ex/engine.c, vmgen-ex/mini-inst.vmg, vmgen-ex2/engine.c, vmgen-ex2/mini-inst.vmg, doc/vmgen.texi, prims2x.fs, vmgen.in:
           prims2x.fs changes: now supports C code delimited by braces;
                               Forth or C names for primitives depending on [ifdef] vmgen
           Documentation changes
           adapted vmgen-ex* to earlier prims2x.fs changes
   2002-08-19  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/vmgen.texi, engine/engine.c, prim, prims2x.fs:
           Vmgen documentation changes
           Small changes in Vmgen and related changes in prim and engine/engine.c
   2002-08-16  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/version.texi.in, doc/vmgen.texi, Makefile.in, configure.in:
           Documentation changes
   2002-08-14  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * Makefile.in, doc/texinfo.tex, doc/vmgen.texi:
           vmgen documentation changes
   2002-08-13  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * COPYING.DOC, Makefile.in, doc/fdl.texi, doc/vmgen.texi:
           documentation changes
   2002-08-12  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * stuff.fs: added str= and string-prefix?
   2002-08-09  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, Benchres, doc/vmgen.texi, prims2x.fs:
           added LABEL2 generation to prims2x.fs for future portable superinstructions
   2002-08-08  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/vmgen.texi: documentation changes
           * doc/vmgen.texi: more documentation
   2002-08-07  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/vmgen.texi: documentation changes
           * vmgen-ex/engine.c, vmgen-ex/mini.h: more disassembly stuff
           * vmgen-ex2/disasm.c, vmgen-ex2/engine.c, vmgen-ex2/mini.h, prims2x.fs, vmgen-ex/disasm.c, vmgen-ex/engine.c, vmgen-ex/mini.h:
           VM disassembler now prints immediate arguments nicely
   2002-08-01  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/gforth.1, doc/vmgen.texi: documentation changes
   2002-07-30  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * vmgen-ex/README, vmgen-ex/engine.c, vmgen-ex/mini-super.vmg, vmgen-ex/mini.h, vmgen-ex/peephole.c, vmgen-ex/support.c, vmgen-ex2/README, vmgen-ex2/mini-super.vmg, vmgen-ex2/mini.h, vmgen-ex2/peephole.c, vmgen-ex2/support.c:
           profiling now turns off superinstructions in vmgen-ex*
   2002-07-26  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * wf.fs: Old version of parse-string for Gforth 0.5.0 added
   2002-07-12  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * tags.fs: fixes (thanks to Erik Rossen)
   2002-06-17  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * Benchres, prim: Added Athlon XP 1800+ to Benchres
           fixed type in wcall
   2002-06-03  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * ekey.fs: EKEY? now does not require complete ESC-sequences
   2002-06-02  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * httpd.fs, prim, tags.fs: Added tags.
           Security fix in httpd.fs
           Documentation for directory access in prim
   2002-06-02  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * vmgen-ex2/profile.c, vmgen-ex2/support.c, README.vmgen, vmgen-ex/engine.c, vmgen-ex/mini.h, vmgen-ex/mini.l, vmgen-ex/mini.y, vmgen-ex/peephole.c, vmgen-ex/profile.c, vmgen-ex/support.c, vmgen-ex2/engine.c, vmgen-ex2/mini.h, vmgen-ex2/mini.l, vmgen-ex2/peephole.c:
           vmgen* now works with lcc and "gcc -O3 -ansi --pedantic -U__GNUC__"
           * vmgen-ex2/mini.y, vmgen-ex2/peephole-blacklist, vmgen-ex2/peephole.c, vmgen-ex2/profile.c, vmgen-ex2/seq2rule.awk, vmgen-ex2/simple.mini, vmgen-ex2/stat.awk, vmgen-ex2/support.c, vmgen-ex2/test.mini, vmgen-ex2/test.out, vmgen-ex2/test.prof, vmgen-ex/mini.h, vmgen-ex/profile.c, vmgen-ex2/Makefile, vmgen-ex2/README, vmgen-ex2/disasm.c, vmgen-ex2/engine.c, vmgen-ex2/fib.mini, vmgen-ex2/fib.prof, vmgen-ex2/mini-inst.vmg, vmgen-ex2/mini-super.vmg, vmgen-ex2/mini.h, vmgen-ex2/mini.l, doc/vmgen.texi, engine/engine.c, engine/forth.h, prims2x.fs, vmgen-ex/engine.c:
           vmgen-related changes:
           in prims2x:
             Conversion macros for single items now take 2 arguments
             Converting from two items to a type has changed order
             argument printing for disassembler disabled (for now)
             disassembler now also uses VM_IS_INST
           in Gforth and vmgen-ex: adapted to work with changed prims2x
           new: vmgen-ex2: uses union for Cell instead of casting (lots of
              changes compared to vmgen-ex)
           * doc/vmgen.texi, ekey.fs, prim, prims2x.fs:
           EKEY now returns single characters instead of waiting for the rest of
              an escape sequence
           documentation changes
   2002-05-28  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * doc/vmgen.texi, doc/gforth.ds, glocals.fs: Documentation changes
   2002-05-16  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * README, doc/vmgen.texi: changed mailing list address in README
           started working on vmgen documentation
   2002-05-01  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * quotes.fs: bugfix
   2002-04-30  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * Makefile.in, quotes.fs: added quotes.fs in Makefile.in
           * quotes.fs: \n now means newline, not LF
   2002-04-27  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/int.fs, kernel/require.fs, doc/gforth.ds, startup.fs, gforth.el, quotes.fs:
           gforth.el: while now increases indentation
           new words s\" .\" \"-parse for dealing with non-printing characters
   2002-04-18  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el:
           Fixed bug that caused scrolling to slow down extremely since
           forth-mode's fontification function was by default installed to
           `fontification-functions' (in Emacs21).  Changed motion-hooking (for
           screen-number display in block files) to using timers instead of
           post-command-hook to prevent further speed lag.
   2002-04-01  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/int.fs, test/signals.fs, doc/gforth.ds, engine/main.c, kernel/comp.fs, Makefile.in, errors.fs, see.fs:
           added regression test for signal on broken execute (not working on Linux-PPC)
           new error message and code for ticking compile-only words
           bug workaround for gforth binary (dynamic superinsts produced wrong code for
   2002-03-25  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * arch/misc/prim.fs: deleted x! and x@
           * arch/misc/prim.fs: before deleting x! and x@
   2002-03-23  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el: Minor bug- und compatability fixes.
   2002-03-22  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * arch/m68k/machine.h, arch/mips/machine.h, arch/power/machine.h, arch/sparc/machine.h, engine/Makefile.in, engine/main.c, Makefile.in, arch/386/machine.h, arch/alpha/machine.h, arch/generic/machine.h, arch/hppa/machine.h, configure.in:
           cleaned up #defines of DIRECT_THREADED and INDIRECT_THREADED
           introduced engine gforth-itc (indirect threading, no replication)
           fixed bug in indirect threading
           * ans-report.fs:
           bugfix (replace-word adapted to primitive-centric code)
           * engine/main.c:
           support for --no-super (no dynamic superinstructions) and
                       --no-dynamic (use only static primitives)
   2002-03-21  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * cross.fs: fix?!
           * cross.fs: MAXU MINI MAXI work now with dlit,
           * cross.fs: added optional write protection for memory regions
           * cross.fs: added symentry and \?
           * arch/misc/prim.fs: cleanup
           * arch/misc/prim.fs:
           volatile vm data (registers and stack) go to seperate memory region
           (preparation for romable tests)
   2002-03-20  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * cross.fs: No forward references to native code definitions
           (primitives and doer) are allowed any more
           (didn't work in any case before, but now a warning is issued)
           * kernel/main.fs: sanity check for header address changed
           * kernel/main.fs:
           moved image-header down to avoid forward reference for doer
   2002-03-19  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * cross.fs: cleaned up bernds changes.
           tried to manage the peephole/call threading stuff with the
           existing plugin definitions.
           * kernel/main.fs: Oops. Sorry for that empty log-message.
           Moved forthstart definition to not have a forward reference on a primitive.
           * kernel/main.fs: *** empty log message ***
           * cross.fs: no functional change
   2002-03-13  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/io.c: portability change
           * INSTALL, Makefile.in, configure.in:
           Better discussion of cross-installation in INSTALL.
           Force direct threading everywhere (indirect threading currentldoes not work).
           make target "gforths" builds and copies all engines.
   2002-03-01  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * BUILD-FROM-SCRATCH, Makefile.in:
           fixed Makefile bugs (to get BUILD-FROM-SCRATCH running again)
   2002-02-10  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/comp.fs, engine/forth.h, engine/main.c, engine/peephole.c, engine/profile.c, prim, prims2x.fs, Makefile.in:
           lit@ and lit+ are now defined as superinstructions
           compile lit @ and lit + instead of lit@ and lit+
           extended prims2x to support superinstructions with non-C-names
             (syntax: forth-name /c-name = ...)
           support profiling of interpreters with superinstructions
             (with simple instructions in the output).
           profile output with prefixes only (enable by editing profile.c).
           optional reporting of static superinstruction lengths (compared to
             dynamic superinstructions); enable by compiling with -DPRINT_SUPER_LENGTHS
   2002-02-04  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, kernel/comp.fs, kernel/cond.fs, kernel/vars.fs, gforth.el, peeprules.vmg:
           Compiles static superinstructions combined with dynamic superinstructions
           added ENDIF to gforth.el
   2002-01-26  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/main.c:
           eliminated searching in compile_prim (faster startup and compilation)
           * engine/main.c, see.fs:
           dynamic superinstructions for the code in the image
           SEE works again for CODE words
   2002-01-26  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el:
           Fixed everything that prevented `gforth.el' from working with XEmacs
           21.1. Fixed info-lookup.
   2002-01-20  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * arch/386/machine.h, engine/engine.c, engine/forth.h, engine/main.c, engine/threaded.h, prim:
           hybrid direct/indirect threading (replaces direct threading)
   2002-01-19  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el: Added code for making `gforth.el' work in Emacs 19.34.
   2002-01-17  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el:
           added and fixed `require's, fixed byte-code-compilation relevant
           things (avoid free variables, top-level `require's etc), added
           `(provide 'forth-mode)'
   2002-01-17  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * environ.fs, float.fs, startup.fs: Fixed enviroment queries
   2002-01-16  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/forth.h, engine/main.c, engine/threaded.h: bugfixes
   2002-01-15  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/main.c:
           compile_prim now converts "xts" into "threaded-code" for DOUBLY_INDIRECT,
                and reports non-primitive non-xt input
   2002-01-14  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, engine/forth.h, engine/main.c, engine/signals.c, engine/threaded.h, Makefile.in, comp-i.fs:
           Differentiate between threaded code and xts in gforth-ditc and in gforthmi
   2002-01-05  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * engine/main.c, engine/signals.c, kernel/aliases0.fs, cross.fs, prim, search.fs:
           Thrown out static vocabulary stack
           Changed cross to make mixed threading workable
           * kernel/args.fs, kernel/comp.fs, kernel/int.fs, kernel/paths.fs, cross.fs, prim:
           Cross compiler changes for mixed threading
   2002-01-05  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * search.fs, gforth.el: Made gforth.el hilight see and dbg.
           Added dynamic search order stack allocation code to search.fs.
   2002-01-04  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * engine/engine.c, see.fs, cross.fs, prim:
           Modified cross to get closer to mixed threading
           Added a few debugging aids
   2002-01-02  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/main.c: preparations in loader for hybrid threading
   2002-01-01  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * Benchres, Makefile.in:
           Added benchmark result (superinstructions) for 600 MHz Athlon
           Made cross work again with superinstructions (requires larger dictionary)
   2001-12-29  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * arch/alpha/machine.h, arch/power/_sync_cache_range.c, engine/main.c, Benchres, Makefile.in, arch/386/machine.h, machpc.fs, machpc.fs.in:
           default dictionary size 1M cells
           machine-specific stuff for dynamic superinstructions is now in machine.h
           support for dynamic superinstructions for Alpha
   2001-12-28  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * engine/engine.c: undefined symbols to make engine.c work with GNU m4
   2001-12-27  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/forth.h, engine/main.c, doc/gforth.ds, machpc.fs:
           gforth no longer uses TOS caching (for improved error detection, suggested by
                                              Dennis Donnelly).
           The default FP stack size is 16K (for improved error detection).
           minor changes.
   2001-12-25  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, kernel/toolsext.fs, prim:
           dynamic superinstructions now continue across conditional branches
           added DEFINED back in (as alias of [DEFINED])
   2001-12-24  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/Makefile.in, engine/forth.h, engine/main.c, kernel/comp.fs, Benchres, Makefile.in, prim, prims2x.fs:
           added dynamic superinstructions (currently for direct threading on 386,
                                            without checking for other platforms)
           * engine/Makefile.in, engine/engine.c, engine/main.c, Benchres, doc/gforth.ds, prims2x.fs:
           determine primitive relocatability for dynamic superinstructions
           documentation changes (from bug reports by Robert Epprecht and others)
   2001-12-15  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el:
           Cleaned up motion-dectection, used for displaying screen-numbers in
           forth-block-mode. Added support for just-in-time hilighting feature of
           Emacs21 (using fontification-functions hook).
   2001-12-09  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, engine/threaded.h, doc/gforth.ds, prim, prims2x.fs:
           use LABEL(name) instead of CASE I_name at the start of a primitive
           fixed read-line tutorial documentation bug
   2001-12-02  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * Benchres, arch/386/machine.h:
           more register allocation fine-tuning; USE_TOS is again used with gcc-3.*
           * Benchres, arch/386/machine.h, engine/Makefile.in:
           explicit register allocation for gcc-3.0.x (not as good as gcc-2.95, but
             better than before); don't use TOS with gcc-3.x, so rp can be allocated.
           Compile engine.c with -fno-gcse (to improve gcc-3.x code quality).
   2001-12-01  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * engine/engine.c, prim, proxy.fs, script.fs:
           Fixed popen access type (*not binary*)
           Fixed some bugs in proxy.fs
           Added scripting support
   2001-11-21  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * fi2c.fs: fixed fi2c.fs address size bug
   2001-11-11  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * arch/386/ncex/ncexcompiler.fs, arch/386/ncex/ncexcontrol.fs, arch/386/ncex/ncexcpu1.fs, arch/386/ncex/ncexcpu2.fs, arch/386/ncex/ncexregalloc.fs, Makefile.in, cross.fs, httpd.fs, prims2x.fs, proxy.fs, wf.fs:
           Added tags generation for vi
           small changes on httpd for EuroForth paper
           Added navigation button generation via Gimp to wf
           Made nccomp from Lars Krueger work at least somehow
   2001-10-26  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/threaded.h, vmgen-ex/engine.c, prims2x.fs:
           prims2x now generates CASE before the label
           * kernel/int.fs, BUGS, ekey.fs, machpc.fs:
           added some function key names and sequences (from Linux console)
           renamed NAME into PARSE-WORD
   2001-10-07  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * locate.fs: Added locate for vi fans
   2001-09-21  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * dosekey.fs, startup.fs: Added DOS support for ekey
   2001-09-17  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * kernel/toolsext.fs: Provided [defined] and [undefined]
           Changed stack comments in toolsext.fs
           * history.fs:
           Added ignoring command completion if there's another key in the queue
   2001-09-16  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * kernel/main.fs:
           Put full featured startup.fs into main.fs, so that both BUILD-FROM-SCRATCH
           and a build based on kernl*.fi can expect the same extensions.
   2001-09-16  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * arch/c165/prim.fs: fix
           * cross.fs: fix: cross now works again on top of kernlXYZ.fi
           * debug.fs, environ.fs, glocals.fs, look.fs: added requires
           * engine/.cvsignore: Added profile.i, peephole.i and gforth-prof
           * engine/main.c: Non-relocatable images with fixed base work again.
           Base address of $100 and 0 are now special cases that indicate
           a relocatable image.
   2001-09-15  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * arch/386/ncex/ncexcompiler.fs, engine/signals.c, history.fs:
           Fix for terminals where TIOCGWINSZ doesn't work
           Added bindkey to history.fs
   2001-09-12  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * cross.fs: added comment
           * kernel/main.fs: cleanup
           * machpc.fs, cross.fs, engine/main.c, kernel/basics.fs, kernel/main.fs:
           Image starts at $100 for cross-compilation now. A NULL-reference
           is really a reference to address 0 and NIL is 0.
           In the kernel image-header to get the headers address.
           Image loading is only changed minimal: The stuff with fixed
           addressed images is thrown out, because every image is relocatable.
           No header change. Old images should work.
           * cross.fs: renamed aprim
   2001-09-06  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * wf.fs: Improved table handling
   2001-09-06  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * cross.fs: added fixme comment
           * kernel/files.fs:
           changed warnings to redefinitions-start / end because
           we have to switch more than only the warnings
   2001-09-05  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * cross.fs: cleaned up
           EC stuff currently broken
           * cross.fs:
           fix: this version produces exacly the same images as version 1.101
           * cross.fs: - added changes for peephole optimization
           Test version!
           * cross.fs: - some cleanup
           - added dlit,
           * cross.fs: - oops, left some debugging output
           - added comment for ghost fields
           * cross.fs:
           merged in the sepearation from colon, and prim, through the >comp field in the ghost
           still TODOs:
           - peephole stuff
           - unification of new >exec-compile and >comp fields
           a running kernl32l.fi was created successfully with this cross.fs
   2001-09-04  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * cross.fs: char optimization
           * cross.fs: lots of changes from the G&D project
           - relocatable assembler output
           - vocabularies support prepared
           - lots of cleanup
           - hooks for instant (target execution while cross-compilation)
           - peephole changes from bernd missing
           - genereation of kernl32l.fi works correctly
           * cross.fs: this is the orignal cross.fs from 0.5.0 updated
           with input-method and long headers
           (used to compare generated images, does not last long)
           * hash.fs: relocation fixes for cross-compilation
           * kernel/comp.fs, kernel/int.fs:
           relocation fixes, use [compile] for immediate words
           added AValue in comp.fs
           * cross.fs: interims version to check in the kernel changes
           * kernel/cloop.fs: changed compile to postpone
           * kernel/cbr.fs: changed to postpone
           * kernel/basics.fs: fix for ec: AConstant relocatable
   2001-09-03  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el:
           Bugfix, Improved hilighting/indentation to recoginize structured
           data-types, that end on %. Indentation now also ignores non-immediate
           words inside definitions. Tested gforth.el on Emacs21. Works fine!!
   2001-09-03  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * arch/misc/tt.fs: fix for relocating
   2001-08-25  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * arch/386/asm.fs, wf.fs: Fixed 386 assembler jump table
           Some improvements to Wiki Forth
   2001-08-08  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * BUILD-FROM-SCRATCH: fix
   2001-08-07  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * BUGS, doc/gforth.ds, float.fs, stuff.fs: New bug reports.
           added slurp-fid; documented slurp-fid and slurp-file.
           bugfix in f. (thanks to Doug Bagley).
   2001-08-06  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * wf.fs: Another set of fixes for wf.fs
   2001-08-05  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * arch/386/ncex/ncexcompiler.fs, arch/386/ncex/ncexcpu1.fs, wf.fs:
           Added toc handling to wf
           Wrote wrapper code for ncex
   2001-07-29  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * wf.fs: Some further fixes
   2001-07-24  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * wf.fs: Fixed jpeg handling
           Improved data base
   2001-07-23  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * wf.fs: Added image size handling
   2001-07-22  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * wf.fs: Some further modification to wiki for Forth
   2001-07-17  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * wf.fs: fixed table end problem
           * wf.fs: Added embedded images (no size yet)
           Improved table parsing
           * wf.fs: Added HTML table, easier ways to add options to a tag
           * wf.fs: Improvements of wiki for Forth
           Added small text data base
   2001-07-16  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * wf.fs: Added first version of a Wiki in Forth
   2001-07-10  crook  <crook@TK147097.tuwien.teleweb.at>
           * cross.fs, arch/c165/mach.fs:
           added constant for bits/char -- seems that this is now obligatory.
           * arch/6502/asm.fs, arch/6502/mach.fs:
           fixed comments and formatting in asm.fs, and set a default processor variant
           (one of the reasons why kernl-6502.fi doesn't build). Added null >boot in
           * arch/misc/README:
           fixed up comments to reflect current file organisation.
           * arch/4stack/README, arch/4stack/mach.sh:
           kernl-4stack.fi now builds. Fixed paths in mach.sh, and changed reference
           therein from gforth to ./gforth to ensure that the engine is consistent
           with the source tree in which the .fi file is being built. Revised README
           to reflect the current file organisation.
           * asm/numref.fs: fix typos and comments.
           * kernel/kernel.fs, kernel/main.fs, kernel/pass.fs:
           minor formatting fixes. Fix typos and comments.
   2001-06-26  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el, etags.fs:
           Fixed some old bug in etags.fs and refined tag search in gforth.el. See my
           not in etags.fs for details.
   2001-06-19  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el: Improved imenu support.
   2001-06-17  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el: Improved interface for customizing hilighting/indentation.
           Added imenu and speedbar support (needs further refinement, though).
   2001-06-16  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * arch/misc/mach.fs, arch/shboom/mach.fs:
           Some further steps to make gforth EC work
           * kernel/quotes.fs, kernel/vars.fs, arch/386/ncex/ncexregalloc.fs, arch/4stack/mach.fs, arch/4stack/relocate.fs, arch/6502/mach.fs, arch/8086/mach.fs, engine/forth.h, engine/main.c, arch/386/ncex/ncexcpu1.fs, cross.fs, prim, prims2x.fs:
           Made some Gforth-EC targets work again
           Started grouping (no renumbering yet, groups are just dummies)
   2001-06-15  ak042  <ak042@TK147097.tuwien.teleweb.at>
           * engine/peephole.c: revert all 2001/05/28 changes.
   2001-06-14  crook  <crook@TK147097.tuwien.teleweb.at>
           * colorize.fs, ansi.fs:
           changed >f to >FG and >b to >BG (and <f <b to <FG <BG) to remove name
           clash between ansi.fs ">f" and gforth floating comparison word.
   2001-06-12  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * tasker.fs: Fixed kill-task
           Added glossary entries
           * tasker.fs: Fixed tasker (to handle new input method)
   2001-05-28  ak042  <ak042@TK147097.tuwien.teleweb.at>
           * engine/peephole.c: there is still a problem with iburg <-> burg.
   2001-05-18  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el: Refined forth-block-mode.
   2001-05-13  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * Makefile.in, prims2x.fs:
           simpy inlclude startup.fs in prims2x if needed
           * BUILD-FROM-SCRATCH, Makefile.in, prims2x.fs, startup.fs:
           cleanup, so BUILD-FROM-SCRATCH (should) works
           * exboot.fs: comment fix
   2001-05-11  crook  <crook@TK147097.tuwien.teleweb.at>
           * engine/forth.h:
           removed spurious ; from end of #endif to get rid of compile warning.
   2001-05-10  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * kernel/accept.fs: corrected control keys vector
   2001-05-09  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * Makefile.in: BUILDFORTH more comment
           * cross.fs: corrected problem with messed up word names in crossdoc.fd
   2001-05-07  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * ToDo: new topics
   2001-05-06  dvdkhlng  <dvdkhlng@TK147097.tuwien.teleweb.at>
           * gforth.el: *** empty log message ***
   2001-05-04  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/quotes.fs, README.vmgen:
           interpreted s" now ALLOCATEs the string
   2001-05-01  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * vmgen.in: vmgen base file
           * README, configure.in, testdist, vmgen-ex/README, Makefile.in:
           minor changes
           * engine/Makefile.in, prim, testdist, Makefile.in, README.vmgen, configure.in, peeprules.vmg:
           vmgen-related changes in Makefile
           fixed FORTHB/prims2x breakage
           removed lit_plus
           added peeprules.vmg as a place to put superinstructions
           testdist now also tests vmgen-ex
           * vmgen-ex/mini-super.vmg, vmgen-ex/mini.h, vmgen-ex/mini.vmg, vmgen-ex/peephole-blacklist, vmgen-ex/profile.c, vmgen-ex/seq2rule.awk, vmgen-ex/stat.awk, vmgen-ex/support.c, vmgen-ex/Makefile, vmgen-ex/README, vmgen-ex/engine.c, vmgen-ex/mini-inst.vmg:
           fixed profiling
           completed README
   2001-04-30  jwilke  <jwilke@TK147097.tuwien.teleweb.at>
           * arch/avr/asm.fs: fixes
           * BUILD-FROM-SCRATCH:
           with an installed gorth 0.5.0 just type ". BUILD-FROM-SCRATCH" after
           * BUILD-FROM-SCRATCH, Makefile.in, prims2x.fs:
           set of changes to use gforth0.5.0 to build right out of the cvs sources
           * arch/generic/machine.h: define HAS_PEEPHOLE only if not defined
           * arch/generic/machine.h:
           HAS_PEEPHOLE must be defined by default as in machpc.fs
   2001-04-29  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * vmgen-ex/Makefile, vmgen-ex/engine.c, vmgen-ex/mini.vmg:
           added more threading schemes and comments
           added USE_spTOS and commented out superflouos cache flushes and refills
           * vmgen-ex/Makefile, vmgen-ex/mini.y, vmgen-ex/test.out: minor changes
           * vmgen-ex/mini.h, vmgen-ex/mini.l, vmgen-ex/mini.vmg, vmgen-ex/mini.y, vmgen-ex/peephole.c, vmgen-ex/profile.c, vmgen-ex/simple.mini, vmgen-ex/support.c, vmgen-ex/test.mini, vmgen-ex/Makefile, vmgen-ex/README, vmgen-ex/disasm.c, vmgen-ex/engine.c, vmgen-ex/fib.mini:
           added vmgen-ex
   2001-04-24  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * vmgen-Attic/Makefile, vmgen-Attic/disasm.c:
           removing vmgen directory (will be replaced by vmgen-ex)
   2001-04-08  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * machpc.fs.in: source for machpc.fs (for peephole)
           * arch/generic/machine.h, engine/engine.c, prim, see.fs, Makefile.in, acconfig.h, configure.in, cross.fs, gforth.el:
           Made peephole a configurable option
   2001-03-28  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, engine/peephole.c:
           peephole optimization now uses a hash table
           primtable() moved to engine.c (threading dependent)
           * kernel/int.fs, see.fs, simp-see.fs, startup.fs: added SIMPLE-SEE
   2001-03-18  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * arch/generic/machine.h, kernel/comp.fs, kernel/int.fs, kernel/vars.fs, machpc.fs, prim, cross.fs:
           Made peephole a configurable option
           call compilation in cross works now
           Not done yet:
           - peephole itself
           - Create as alit,
           - DOES> compilation
   2001-03-18  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs:
           prefixes are now automatically defined in wordlist PREFIXES
           * engine/engine.c, prim, prims2x.fs:
           moved stack definitions from prims2x.fs to prim (and support for that)
           * prims2x.fs:
           cleaned up prims2x.fs, prepare for pulling stack definition out
           * engine/profile.c, prims2x.fs:
           VM_DEBUGGING now also prints stack-out values.
           output-profile now uses goto to avoid gcc parser overflow.
   2001-03-17  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * cross.fs, see.fs:
           Further peephole optimization preparation in cross.fs:
           - compile calls now works
           - compile Create/DOES> better prepaired (doesn't work yet)
           - compile Create/Variable doesn't work yet
           - no peephole optimization yet
           See calls works
   2001-03-11  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * cross.fs, gforth.el:
           Checked in cross with some part of peephole optimization (doers - not all of
           Checked in fontlock-mode coloring version of gforth.el
           * Makefile.in, cross.fs, prim, prims2x.fs:
           First steps to get peephole optimizing into cross
   2001-03-04  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/profile.c, prims2x.fs: minor bugfixes
   2001-03-01  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * Makefile.in, engine/profile.c, prims2x.fs:
           profiling now outputs subsequences
   2001-02-28  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * Makefile.in, engine/forth.h, engine/main.c, engine/profile.c, prim, prims2x.fs:
           continued block profiling
   2001-02-27  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, engine/forth.h, engine/main.c, engine/peephole.c, engine/profile.c, Makefile.in, backtrac.fs, engine/Makefile.in, prim, prims2x.fs:
           good start at profiling for peephole optimization
           backtrace now also works for calls done with CALL
           * prims2x.fs: "TAIL;" now works properly in superinstructions
   2001-02-26  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/comp.fs, test/string.fs, gforthmi.in, prim:
           gforthmi now deals properly with arguments containing spaces
           added CALL and USERADDR primitives
           all references to colon defs, constants etc. are now compiled to primitives
              with inline arguments
           improved COMPARE test case
   2001-02-24  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/comp.fs, kernel/cond.fs, kernel/int.fs, kernel/vars.fs, comp-i.fs, prims2x.fs:
           added peephole optimization to kernel
           * engine/Makefile.in, engine/engine.c, engine/forth.h, engine/peephole.c, prim, prims2x.fs, Makefile.in:
           added C and primitive support for peeophole optimization
           * Makefile.in, prims2x.fs, test/string.fs: added test for COMPARE
           more peephole optimization stuff
   2001-02-23  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs: first take on generating combined instructions
   2001-02-09  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs:
           fixed inst-stream access in parts of combined instructions
   2001-02-08  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * BUGS, prims2x.fs: more combined word changes
   2001-02-07  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/forth.h, prims2x.fs:
           type-specific conversion functions for fetch/store-double
   2001-02-06  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, BUGS, configure.in, prim, prims2x.fs:
           type conversion macros instead of casts for fetch/store-single (prims2x.fs)
   2001-02-04  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * arch/8086/mach.fs, cross.fs, kernel/int.fs, prim:
           Fixed problems with new word header format for Gforth EC (now only tested
           for gf8086.com)
   2001-01-30  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * kernel/int.fs: Fixed masks for character sizes > 8 bit
   2001-01-29  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * kernel/int.fs: Fixed [ alias-mask lcount-mask or ] Literal problem
           * cross.fs: Fixed input method problem
   2001-01-28  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * cross.fs, kernel/comp.fs, kernel/int.fs, prims2x.fs:
           Fixed cell-size dependent masks
   2001-01-28  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * kernel/comp.fs, kernel/int.fs, Makefile.in, wordinfo.fs:
           fixed bug (?; might also be due to an Emacs change) in prim.TAGS generation.
           bugfix in alias? (length related)
           NEXTNAME can now also handle long names
           added FREE-MEM-VAR
           * kernel/comp.fs, see.fs: SEE bugfix (coming from long names)
           NEXTNAME bugfix (but still length restrictions)
           * kernel/int.fs, comp-i.fs, cross.fs, hash.fs, kernel/comp.fs, table.fs:
           long names in dictionary
   2001-01-27  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/engine.c, prim:
           added primitives (listlfind) (hashlfind) (tablelfind) for dealing with
              long names
   2001-01-24  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs: replaced items with field in prim
           * kernel/tools.fs, prims2x.fs: factored out WORDLIST-WORDS from WORDS
           more work on combined primitives
   2001-01-23  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs: minor changes
           * kernel/int.fs: made bracktrace-rp0 restoration THROW-proof
           * Makefile.in, kernel/int.fs, prims2cl.fs, prims2x.fs:
           bugfix (complete?): backtrace-rp0 now is restored in interpret
           process-file in prims2x.fs now takes two xts
           various other chnages in prims2x.fs
   2001-01-21  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs: first parts for combined instructions
   2001-01-20  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs: introduced make-prim
   2001-01-19  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * comp-i.fs, prims2x.fs, stuff.fs:
           moved SLURP-FILE from comp-i.fs to stuff.fs
           various changes in prims2x.fs
   2001-01-18  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs: more refactoring in prims2x.fs
           * prims2x.fs: moved parser to the end
           * prims2x.fs: factored output out of DECLARATIONS
           * doc/gforth.ds, prims2x.fs: Better error reporting in prims2x.fs
           Bug fixed by changing the docs: RECOVER is not optional in TRY...ENDTRY.
   2001-01-17  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs: some refactoring
   2001-01-14  pazsan  <pazsan@TK147097.tuwien.teleweb.at>
           * prim: Fixed error report for read-dir
   2001-01-09  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * prims2x.fs, ChangeLog, prim: changes for better debugging output
           made a ChangeLog for real
           NEWLINE for Darwin (should be LF, but if does not define unix)
   2001-01-06  Anton Ertl  <anton@TK147097.tuwien.teleweb.at>
           * engine/threaded.h, prims2x.fs: bugfix
 2001-01-06  Anton Ertl  <anton@mips.complang.tuwien.ac.at>  2001-01-06  Anton Ertl  <anton@mips.complang.tuwien.ac.at>
         * engine/threaded.h, prims2x.fs: bugfix          * engine/threaded.h, prims2x.fs: bugfix

Removed from v.1.3  
changed lines
  Added in v.1.4

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>