sbcl Log


Commit Date  
[57328d] (sbcl_0_7_10) by William Harold Newman William Harold Newman

0.7.10:
release, tagged as sbcl_0_7_10

2002-11-27 14:47:09 Tree
[6e7ae2] by Christophe Rhodes Christophe Rhodes

0.7.9.68:
Text file frobbing (mostly new BUG)

2002-11-26 16:31:24 Tree
[435b7a] by Christophe Rhodes Christophe Rhodes

0.7.9.67:
MIPSel build patch
... write a noddy C program to determine target endianness
(KLUDGE: assumes int is 32 bits)
... use it in make-config.sh if we are building a MIPS target

2002-11-25 13:24:15 Tree
[24407d] by William Harold Newman William Harold Newman

0.7.9.66:
merged emu patch from sbcl-devel 2002-11-18, tweaking
.32 patch to cope with PEEK-CHAR/UNREAD-CHAR
on ECHO-STREAMs better

2002-11-24 22:40:52 Tree
[5bf4a6] by Christophe Rhodes Christophe Rhodes

0.7.9.65:
Firefighting on showstopper bugs:
... commit change as per CSR sbcl-devel 22-11-2002 for BUG 227
... Julian Fondren "leap before looking" patch for
sbcl --core (sbcl-devel 21-11-2002)
... DB "silly debugging bug" sbcl-devel 24-11-2002

2002-11-24 15:57:46 Tree
[43c1fa] by Alexey Dejneka Alexey Dejneka

0.7.9.64:
* Fix bug in SUBTYPEP of function types, reported by Valtteri
Vuorik, explained by CSR.
* Fix 3AND.

2002-11-24 06:08:39 Tree
[7e6782] by William Harold Newman William Harold Newman

0.7.9.63:
(Oops.)
restored bug 226: I had deleted it from BUGS in my last commit
thinking that it was just there because APD had
forgotten to delete it, when in fact it was put back
on purpose by APD in .60 when he unrolled his
changes from .54.

2002-11-21 21:12:08 Tree
[e16754] by William Harold Newman William Harold Newman

0.7.9.62:
noted bug 227
(bug 226 gone w/ recent APD fix)

2002-11-21 16:52:06 Tree
[731bc6] by Christophe Rhodes Christophe Rhodes

0.7.9.61:
Fix destructuring of LOOP WITH <x> where <x> is a tree with NIL
in it.
... define and use a somewhat KLUDGEy LOOP-DESTRUCTURING-BIND.

2002-11-21 15:55:43 Tree
[7f008d] by Alexey Dejneka Alexey Dejneka

0.7.9.60:
Undone patch from 0.7.9.54 for bugs 115 and 226.

2002-11-21 14:36:01 Tree
[89925c] by Christophe Rhodes Christophe Rhodes

0.7.9.59:
No longer put an extra BLOCK NIL around named LOOPs.
... what were they thinking? It was very deliberate. Any ideas?

2002-11-20 15:40:34 Tree
[ea652c] by Christophe Rhodes Christophe Rhodes

0.7.9.58:
Some more error-checking at DEFCLASS
... duplicate :METACLASS, :DEFAULT-INITARGS options
... :READER and :INITARG options to slots must be symbols
Fix up error messages
... add some spaces to previous commit
... CLASS is not what you think it is in ENSURE-CLASS-VALUES

2002-11-19 19:02:15 Tree
[a96eb7] by Christophe Rhodes Christophe Rhodes

0.7.9.57:
DEFINE-METHOD-COMBINATION now works with the :ARGUMENTS option
(more or less as per Gerd Moellmann cmucl-imp
2002-10-19)
... extra slot, extra logic;
... test from CLHS DEFINE-METHOD-COMBINATION page.

2002-11-19 16:00:19 Tree
[160e30] by Alexey Dejneka Alexey Dejneka

0.7.9.56:
Fix miscellaneous bugs in LOOP REPEAT:
* incorrect type inference
* REPEAT is a `main clause'
* semantics change during constant folding

2002-11-19 07:28:53 Tree
[8fa443] by Christophe Rhodes Christophe Rhodes

0.7.9.55:
Fix some more LOOP compliance issues
... signal a package error if the package doesn't exist
... the loop block should surround variable initialization
forms

2002-11-19 00:42:29 Tree
[532694] by Alexey Dejneka Alexey Dejneka

0.7.9.54:
* Remove bug entry 54.
* Do not propagate liveness of :DEBUG-ENVIRONMENT TNs into
another environment.

2002-11-18 05:52:18 Tree
[237ece] by William Harold Newman William Harold Newman

0.7.9.53:
tidying left over from failed hunt for bug #226
some more renaming to make LEXENV vs. PHYSENV distinction

2002-11-17 13:56:56 Tree
[c7dc5b] by Alexey Dejneka Alexey Dejneka

0.7.9.52:
* Fixed bug: loop-for-as-package does not require package to
be explicitely specified;
* PRINT-IR2-BLOCKS shows corresponding IR1 block numbers.

2002-11-16 10:26:13 Tree
[89f199] by Christophe Rhodes Christophe Rhodes

0.7.9.51:
Fix some LOOP bugs reported by Paul Dietz cmucl-imp
... NIL is an ignored variable name
Minor frobs
... loop.lisp now compiles without style-warnings
... signal a package error in duplicate package logic

2002-11-15 15:21:57 Tree
[7ffdb2] by Christophe Rhodes Christophe Rhodes

0.7.9.50:
Implement NO-PRIMARY-METHOD, following WHN/Gerd Moellman
... and use it when we can't use NO-APPLICABLE-METHOD

2002-11-15 14:47:05 Tree
[742e0b] by Christophe Rhodes Christophe Rhodes

0.7.9.49:
Fix bug 136 (CALL-NEXT-METHOD and lexical rebinding of
arguments):
entomotomy: call-next-method-lexical-args
... delete the MACROLET "optimization" path

2002-11-14 19:03:16 Tree
[3ecee4] by Christophe Rhodes Christophe Rhodes

0.7.9.48:
Fix COMPUTE-SLOTS :AROUND to do predictable things on
STANDARD-CLASS
... put a new slot (ALLOCATION-CLASS) into slotds, to be
initialized to the relevant class if the allocation of
the slot is :CLASS
... use SLOT-ALLOCATION-CLASS where SLOT-ALLOCATION was used in
the case of a :CLASS allocation
... divide the COMPUTE-SLOTS :AROUND method into two, and leave
the FUNCALLABLE-INSTANCE one alone

2002-11-14 11:31:52 Tree
[4ecf0a] by Alexey Dejneka Alexey Dejneka

0.7.9.47:
s/#.most-positive-fixnum/#.sb!xc:most-positive-fixnum/ in seq.lisp.
Thanks, Christophe!

2002-11-13 04:07:58 Tree
[8d021a] by Alexey Dejneka Alexey Dejneka

0.7.9.46:
:COUNT argument to sequence functions may be a BIGNUM.

2002-11-12 16:50:09 Tree
[a29fb6] by Christophe Rhodes Christophe Rhodes

0.7.9.45:
Merge patch from Gerd Moellmann cmucl-imp 2002-10-29
"COMPUTE-CLASS-PRECEDENCE-LIST and AMOP"
... make COMPUTE-CLASS-PRECEDENCE-LIST specialize on CLASS, not
SLOT-CLASS
... catch some forward-referenced-class cases that slipped
through the net
write a FINALIZE-INHERITANCE method for forward-referenced-class

2002-11-12 15:33:39 Tree
Older >