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 initializeinstance 
34 

and sharedinitialize. 
35 


36 
* Bugfixes: 
* Bugfixes: 
37 
 Floatingpoint traps are now handled on ppc. Previously, no 
 Floatingpoint traps are now handled on ppc. Previously, no 
38 
traps were signalled and SETFLOATINGPOINTMODES did nothing. 
traps were signalled and SETFLOATINGPOINTMODES did nothing. 
39 
 FILEPOSITION no longer returns negative values for mapped 
 FILEPOSITION no longer returns negative values for mapped 
40 
filesimplestream's. 
filesimplestream'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 doubledoublefloat'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 (doublefloat (0d0) 1d0) and y of type 
for x of type (doublefloat (0d0) 1d0) and y of type 
47 
(doublefloat (0d0)) now returns (or (member 0d0) (doublefloat 
(doublefloat (0d0)) now returns (or (member 0d0) (doublefloat 
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 doubledoublefloats 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 DOUBLEDOUBLEFLOAT and (COMPLEX 


DOUBLEDOUBLEFLOAT) 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. 


 (INTEGERDECODEFLOAT <doubledoublefloat>) was sometimes 


returning the wrong integer value because the two components had 


the wrong sign. 


 Some issues with debugger printing out DOUBLEDOUBLEFLOATs and 


(COMPLEX DOUBLEDOUBLEFLOAT)'s have been fixed. 

54 
 Fix issue with UNIX:UNIXMMAP handling of "large" addresses that 
 Fix issue with UNIX:UNIXMMAP handling of "large" addresses that 
55 
appeared to be negative numbers. 
appeared to be negative numbers. 
56 

 DOUBLEDOUBLEFLOAT fixes: 
57 

o Converting negative rationals to doubledoublefloat's doesn't 
58 

produce wrong answers anymore. 
59 

o Conversion of bignums to doubledoublefloats 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 DOUBLEDOUBLEFLOAT and (COMPLEX 
65 

DOUBLEDOUBLEFLOAT) 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 (INTEGERDECODEFLOAT <doubledoublefloat>) 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 DOUBLEDOUBLEFLOATs and 
72 

(COMPLEX DOUBLEDOUBLEFLOAT)'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 

