/[cmucl]/src/general-info/release-19e.txt
ViewVC logotype

Diff of /src/general-info/release-19e.txt

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.8 by rtoy, Mon Apr 30 14:24:16 2007 UTC revision 1.9 by rtoy, Wed May 2 13:41:51 2007 UTC
# Line 30  New in this release: Line 30  New in this release:
30    * ANSI compliance fixes:    * ANSI compliance fixes:
31      - BOA constructors with &AUX variables are handled better now.      - BOA constructors with &AUX variables are handled better now.
32      - SHADOW accepts characters now.      - SHADOW accepts characters now.
33        - Default initargs are now passed correctly to initialize-instance
34          and shared-initialize.
35    
36    * Bugfixes:    * Bugfixes:
37      - Floating-point traps are now handled on ppc.  Previously, no      - Floating-point traps are now handled on ppc.  Previously, no
38        traps were signalled and SET-FLOATING-POINT-MODES did nothing.        traps were signalled and SET-FLOATING-POINT-MODES did nothing.
39      - FILE-POSITION no longer returns negative values for mapped      - FILE-POSITION no longer returns negative values for mapped
40        file-simple-stream's.        file-simple-stream's.
41      - Potential Version numbers that start with a leading 0 are not      - Potential Version numbers that start with a leading 0 are no
42        treated as version numbers.  Hence, "foo.~1~" has name "foo",        longer treated as version numbers.  Hence, "foo.~1~" has name
43        type nil, and version 1, but "foo.~01~" has type "~01~" and        "foo", type nil, and version 1, but "foo.~01~" has type "~01~"
44        version :NEWEST.        and version :NEWEST.
     - Converting negative rationals to double-double-float's doesn't  
       produce wrong answers anymore.  
45      - A bug in type derivation for EXPT has been fixed.  (expt x y)      - A bug in type derivation for EXPT has been fixed.  (expt x y)
46        for x of type (double-float (0d0) 1d0) and y of type        for x of type (double-float (0d0) 1d0) and y of type
47        (double-float (0d0)) now returns (or (member 0d0) (double-float        (double-float (0d0)) now returns (or (member 0d0) (double-float
# Line 51  New in this release: Line 51  New in this release:
51        now.  Previously the wrong instruction was decoded, which        now.  Previously the wrong instruction was decoded, which
52        produced the wrong operation and operands in the arithmetic        produced the wrong operation and operands in the arithmetic
53        error handler.        error handler.
     - Conversion of bignums to double-double-floats preserves more  
       bits of the bignum now.  This allows for read/print consistency  
       for this particular case.  (The general case is still broken.)  
     - (float <negative bignum> 1w0) no longer returns a positive  
       result.  
     - Some issues with creation of DOUBLE-DOUBLE-FLOAT and (COMPLEX  
       DOUBLE-DOUBLE-FLOAT) have been fixed on sparc, ppc, and x86.  
       These seem to work, except there appears to be some bugs on x86  
       when compiling at speed 3 and safety 0.  
     - (INTEGER-DECODE-FLOAT <double-double-float>) was sometimes  
       returning the wrong integer value because the two components had  
       the wrong sign.  
     - Some issues with debugger printing out DOUBLE-DOUBLE-FLOATs and  
       (COMPLEX DOUBLE-DOUBLE-FLOAT)'s have been fixed.  
54      - Fix issue with UNIX:UNIX-MMAP handling of "large" addresses that      - Fix issue with UNIX:UNIX-MMAP handling of "large" addresses that
55        appeared to be negative numbers.        appeared to be negative numbers.
56        - DOUBLE-DOUBLE-FLOAT fixes:
57          o Converting negative rationals to double-double-float's doesn't
58            produce wrong answers anymore.
59          o Conversion of bignums to double-double-floats preserves more
60            bits of the bignum now.  This allows for read/print consistency
61            for this particular case.  (The general case is still broken.)
62          o (float <negative bignum> 1w0) no longer returns a positive
63            result.
64          o Some issues with creation of DOUBLE-DOUBLE-FLOAT and (COMPLEX
65            DOUBLE-DOUBLE-FLOAT) have been fixed on sparc, ppc, and x86.
66            These seem to work, except there appears to be some bugs on x86
67            when compiling at speed 3 and safety 0.
68          o (INTEGER-DECODE-FLOAT <double-double-float>) was sometimes
69            returning the wrong integer value because the two components had
70            the wrong sign.
71          o Some issues with debugger printing out DOUBLE-DOUBLE-FLOATs and
72            (COMPLEX DOUBLE-DOUBLE-FLOAT)'s have been fixed.
73    
74    * Trac Tickets:    * Trac Tickets:
75      - #8 fixed so logs of bignums and ratios that won't fit into a      - #8 fixed so logs of bignums and ratios that won't fit into a
76        float can be computed, instead of signalling an error.        float can be computed, instead of signalling an error.
77        - #9 fixed.  An error is generated if the number of days doesn't
78          match the number of days in the given month.
79    
80    * Other changes:    * Other changes:
81    

Legend:
Removed from v.1.8  
changed lines
  Added in v.1.9

  ViewVC Help
Powered by ViewVC 1.1.5