Skip to content
GitLab
Explore
Sign in
Register
Primary navigation
Search or go to…
Project
cmucl-copy
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Package registry
Model registry
Operate
Environments
Terraform modules
Monitor
Incidents
Service Desk
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Terms and privacy
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
Richard M Kreuter
cmucl-copy
Commits
6f8b2114
Commit
6f8b2114
authored
19 years ago
by
emarsden
Browse files
Options
Downloads
Patches
Plain Diff
Minor reorganizations and cleanups to release notes.
parent
7ada75f7
No related branches found
No related tags found
No related merge requests found
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
general-info/release-19b.txt
+39
-37
39 additions, 37 deletions
general-info/release-19b.txt
with
39 additions
and
37 deletions
general-info/release-19b.txt
+
39
−
37
View file @
6f8b2114
...
@@ -15,21 +15,28 @@ Common Lisp. CMUCL is maintained by a team of volunteers collaborating
...
@@ -15,21 +15,28 @@ Common Lisp. CMUCL is maintained by a team of volunteers collaborating
over the Internet, and is mostly in the public domain.
over the Internet, and is mostly in the public domain.
New in this release:
New in this release:
* Newly supported platforms:
- CMUCL has been ported to Mac OS X. The port includes support for
callbacks from foreign code, and linkage-table support (which
makes it possible to dump images that contain references to
foreign code).
- NetBSD/x86 supported.
* Feature enhancements:
* Feature enhancements:
- CMUCL ported to Mac OS X.
- Added a much faster and much less consy bignum printer and a
- Added a much faster and much less consy bignum printer and a
faster bignum reader.
faster bignum reader.
- Extended REQUIRE to support *MODULE-PROVIDER-FUNCTIONS* from
- Extended REQUIRE to support *MODULE-PROVIDER-FUNCTIONS* from
SBCL.
SBCL.
- NetBSD supported.
- SIMPLE-STREAM fixes:
- SIMPLE-STREAM fixes:
o READ-SEQUENCE and WRITE-SEQUENCE for
o READ-SEQUENCE and WRITE-SEQUENCE for
signed-byte/unsigned-byte 8 streams.
signed-byte/unsigned-byte 8 streams.
o Memory-mapped file streams.
o Memory-mapped file streams.
- Large file support for Solaris, so CMUCL can handle files
- Large file support for Solaris, so CMUCL can handle files
larger than 2GB.
larger than 2GB.
- Support for Mac OS X.
- A much faster GCD implementation.
- A much faster GCD implementation.
- A faster Karatsuba multiplier when the number of words of the
- A faster Karatsuba multiplier when the number of words of the
bignums is not a power of 2. (Break-even is still around 1500
bignums is not a power of 2. (Break-even is still around 1500
...
@@ -41,21 +48,16 @@ New in this release:
...
@@ -41,21 +48,16 @@ New in this release:
- Lazy sxhash on symbols implemented for ppc.
- Lazy sxhash on symbols implemented for ppc.
- Pretty-printer for LOOP.
- Pretty-printer for LOOP.
- Added some more pprinters for various forms.
- Added some more pprinters for various forms.
- Callbacks supported on Mac OS X.
- Loop analysis code added to the compiler. It's incorporated
- Loop analysis code added to the compiler. It's incorporated
into the build and can be enabled by setting *loop-analyze* to
into the build and can be enabled by setting *loop-analyze* to
T, but the compiler doesn't use the loop information for
T, but the compiler doesn't use the loop information for
anything yet.
anything yet.
- Errors from socket functions signal a SOCKET-ERROR instead of
- Errors from socket functions signal a SOCKET-ERROR instead of
ERROR
.
a generic ERROR condition
.
- Efficiency notes are no longer printed if inhibit-warnings >=
- Efficiency notes are no longer printed if inhibit-warnings >=
2.5.
2.5.
- Simple support for datagram networking added: INET-RECVFROM,
- Simple support for datagram networking added: INET-RECVFROM,
INET-SENDTO, INET-SHUTDOWN.
INET-SENDTO, INET-SHUTDOWN.
- Linkage table support for PPC and Darwin.
- A socket reset signals a SOCKET-ERROR condition instead of a
general ERROR condition.
- :LINKAGE-TABLE feature is now supported on PPC.
- MAKE-INDENTING-STREAM and INDENTING-STREAM-P are exported from
- MAKE-INDENTING-STREAM and INDENTING-STREAM-P are exported from
the EXTENSIONS package instead of being internal.
the EXTENSIONS package instead of being internal.
- TRACE will work now for the known-return convention, which
- TRACE will work now for the known-return convention, which
...
@@ -64,7 +66,7 @@ New in this release:
...
@@ -64,7 +66,7 @@ New in this release:
- Bivalent streams are supported. Thus, streams with
- Bivalent streams are supported. Thus, streams with
element-type (unsigned-byte 8) can be used for character
element-type (unsigned-byte 8) can be used for character
input/output as well as binary input/output. Supply the
input/output as well as binary input/output. Supply the
option :CLASS 'EXT:BINARY-TEXT-STREAM for OPEN to get bivalen
option :CLASS 'EXT:BINARY-TEXT-STREAM for OPEN to get bivalen
t
streams. EXT:READ-VECTOR and EXT:WRITE-VECTOR also
streams. EXT:READ-VECTOR and EXT:WRITE-VECTOR also
implemented. This is somewhat experimental right now, and may
implemented. This is somewhat experimental right now, and may
change slightly in a future version.
change slightly in a future version.
...
@@ -81,7 +83,7 @@ New in this release:
...
@@ -81,7 +83,7 @@ New in this release:
fixes an issue with ASDF and logical pathnames.
fixes an issue with ASDF and logical pathnames.
- Numerous fixes to the float printer, including a new float
- Numerous fixes to the float printer, including a new float
printer.
printer.
- Justification (~< ~>) works better.
- Justification (~< ~>
FORMAT directives
) works better.
- ~V with the value of NIL in format strings is treated to mean
- ~V with the value of NIL in format strings is treated to mean
as if it weren't given at all, as specified by ANSI CL.
as if it weren't given at all, as specified by ANSI CL.
- Characters print consistently with ~:C and ~S.
- Characters print consistently with ~:C and ~S.
...
@@ -91,7 +93,7 @@ New in this release:
...
@@ -91,7 +93,7 @@ New in this release:
- UPGRADED-COMPLEX-PART-TYPE returns RATIONAL for integer types
- UPGRADED-COMPLEX-PART-TYPE returns RATIONAL for integer types
now. (CMUCL doesn't really have a (COMPLEX INTEGER) type).
now. (CMUCL doesn't really have a (COMPLEX INTEGER) type).
- When a pathname can't be printed readably, signal a
- When a pathname can't be printed readably, signal a
PRINT-NOT-READABLE condition
PRINT-NOT-READABLE condition
.
- Printing arrays readably was sometimes wrong.
- Printing arrays readably was sometimes wrong.
- FOR-AS-ON-LIST works with dotted lists.
- FOR-AS-ON-LIST works with dotted lists.
- DEFSTRUCT predicates for list structures don't produce errors
- DEFSTRUCT predicates for list structures don't produce errors
...
@@ -99,7 +101,7 @@ New in this release:
...
@@ -99,7 +101,7 @@ New in this release:
- UNBOUND-SLOT condition no longer takes the :slot keyword. It
- UNBOUND-SLOT condition no longer takes the :slot keyword. It
uses the ANSI-specified :name keyword. This is not backward
uses the ANSI-specified :name keyword. This is not backward
compatible!
compatible!
- MAKE-SYMBOL takes STRING
'
s, not just SIMPLE-STRING
'
s.
- MAKE-SYMBOL takes STRINGs, not just SIMPLE-STRINGs.
- PPRINT-POP did not return NIL when the list arg to
- PPRINT-POP did not return NIL when the list arg to
PPRINT-LOGICAL-BLOCK was NIL.
PPRINT-LOGICAL-BLOCK was NIL.
- Signal appropriate error conditions instead of a plain error
- Signal appropriate error conditions instead of a plain error
...
@@ -116,49 +118,49 @@ New in this release:
...
@@ -116,49 +118,49 @@ New in this release:
- Format bugs in ~{ and ~@[X~], as noted by ansi-tests, have
- Format bugs in ~{ and ~@[X~], as noted by ansi-tests, have
been fixed.
been fixed.
- Evaluation order of REMF, INCF, and DECF.
- Evaluation order of REMF, INCF, and DECF.
- When *
read-suppress
* is T, READ, READ-PRESERVING-WHITESPACE,
- When *
READ-SUPPRESS
* is T, READ, READ-PRESERVING-WHITESPACE,
READ-DELIMITED-LIST, and READ-FROM-STRING now return NIL, as
READ-DELIMITED-LIST, and READ-FROM-STRING now return NIL, as
specified by ANSI CL. (Previously, CMUCL would return a list
specified by ANSI CL. (Previously, CMUCL would return a list
of NILs.)
of NILs.)
- Pathname components of type :UNSPECIFIC are no
t
printed in
- Pathname components of type :UNSPECIFIC are no
longer
printed in
namestrings
anymore
.
namestrings.
- Dumping of symbols in fasls work according to ANSI
now
.
- Dumping of symbols in fasls
now
work
s
according to ANSI.
- READ-FROM-STRING was sometimes not returning the correct
- READ-FROM-STRING was sometimes not returning the correct
second value when the string was a displaced string.
second value when the string was a displaced string.
- The invalid trait for constituent characters is handled
- The invalid trait for constituent characters is handled
better.
better.
- (formatter "~V[...~]" now returns the correct result for
- (formatter "~V[...~]") now returns a correct second value.
second value.
- MAKE-DISPATCH-MACRO-CHARACTER is supposed to return T.
- MAKE-DISPATCH-MACRO-CHARACTER is supposed to return T.
- APROPOS and APROPOS-LIST
don't
take 3 arg
s anymore
. ANSI CL
- APROPOS and APROPOS-LIST
no longer
take 3 arg
uments
. ANSI CL
says they only take 2.
says they only take 2.
- A reader error is signaled when bad ratios (like 1/0) are read
- A reader error is signaled when bad ratios (like 1/0) are read
or when a float can't fit into the specified float type.
or when a float can't fit into the specified float type.
- A form like #(1 . 2) signals a reader error
now
.
- A form like #(1 . 2)
now
signals a reader error.
- NIL as a keyword arg is no longer silently accepted, unless
- NIL as a keyword arg is no longer silently accepted, unless
NIL was actually a specified keyword arg.
NIL was actually a specified keyword arg.
- Some cases of reader errors being signaled when
- Some cases of reader errors being signaled when
*READ-SUPPRESS* was T no longer occur.
*READ-SUPPRESS* was T no longer occur.
- The datum slot for check-type contains the place-value now
- The datum slot for check-type contains the place-value now
instead of the place.
instead of the place.
- The expected-type and datum should be consistent in
- The expected-type and datum should
now
be consistent in
type-errors
now.
(datum is not of the expected-type
.
)
type-errors (datum is not of the expected-type)
.
Similar situations in other errors are also corrected.
Similar situations in other errors are also corrected.
* Numerous bugfixes:
* Numerous bugfixes:
- FFLOOR and friends preserve the sign of the arg, so -0.0 is
- FFLOOR and friends preserve the sign of the arg, so -0.0 is
returned for negative args.
returned for negative args.
- EXPT sometimes returne NaN instead of 1 when the power was 0.
- EXPT sometimes returne
d
NaN instead of 1 when the power was 0.
- Weak pointers
don't
have O(n^2) GC behavior
anymore
.
- Weak pointers
no longer
have O(n^2) GC behavior.
- X11 authorization changes to solve problems seen with servers
- X11 authorization changes to solve problems seen with servers
that don't listen to TCP connection requests.
that don't listen to TCP connection requests.
-
unix-read
fix
-
UNIX-READ
fix
- The compiler would give an error on deriving the result type
- The compiler would give an error on deriving the result type
of RANDOM when the arg to RANDOM was a union type.
of RANDOM when the arg to RANDOM was a union type.
- "~*~" was not being parsed to mean a wildcard version.
- "~*~" was not being parsed to mean a wildcard version.
- UNIX-GETPWUID implemented for Darwin. Needed for FILE-AUTHOR.
- UNIX-GETPWUID implemented for Darwin. Needed for FILE-AUTHOR.
- The compiler no longer gives errors when deriving the result type of
- The compiler no longer gives errors when deriving the result type of
coerce
when converting a integer to a float that would not fit
COERCE
when converting a integer to a float that would not fit
in a float.
in a float.
- Fix bug in reading from streams with element-type SIGNED-BYTE
- Fix bug in reading from streams with element-type SIGNED-BYTE
for sizes greater than 32: Positive numbers were incorrectly
for sizes greater than 32: Positive numbers were incorrectly
...
@@ -168,7 +170,7 @@ New in this release:
...
@@ -168,7 +170,7 @@ New in this release:
because PPRINT-EXIT-IF-LIST-EXHAUSTED caused an early exit.
because PPRINT-EXIT-IF-LIST-EXHAUSTED caused an early exit.
- Macro expansion was not looking in the local environment
- Macro expansion was not looking in the local environment
properly for local definitions.
properly for local definitions.
- Circular
W
eak pointers can now be printed when *
print-circle
*
- Circular
w
eak pointers can now be printed when *
PRINT-CIRCLE
*
is T.
is T.
- "foo.lisp.*" is interpreted to have version :wild, just like
- "foo.lisp.*" is interpreted to have version :wild, just like
"foo.lisp.~*~".
"foo.lisp.~*~".
...
@@ -217,7 +219,7 @@ New in this release:
...
@@ -217,7 +219,7 @@ New in this release:
- Output to (make-two-way-stream (make-concatenated-stream)
- Output to (make-two-way-stream (make-concatenated-stream)
(make-broadcast-stream)) silently discards output instead of
(make-broadcast-stream)) silently discards output instead of
producing an error.
producing an error.
- STRING<, STRING<=, STRING>, STRING>= no longer produce
s
- STRING<, STRING<=, STRING>, STRING>= no longer produce
incorrect results.
incorrect results.
- COPY-SEQ sometimes produced erroneous results because the
- COPY-SEQ sometimes produced erroneous results because the
compiler was confused by the derived result type of COPY-SEQ.
compiler was confused by the derived result type of COPY-SEQ.
...
@@ -233,15 +235,15 @@ New in this release:
...
@@ -233,15 +235,15 @@ New in this release:
- The disassembler handles closures better.
- The disassembler handles closures better.
- The result array for bit-* operations was sometimes too short
- The result array for bit-* operations was sometimes too short
if the args had fill pointers.
if the args had fill pointers.
- Weak hash tables don't cause GC to go into infinite loops
- Weak hash tables no longer cause GC to go into an infinite loop.
anymore.
- (PEEK-CHAR nil s nil foo) doesn't remove the character
- (PEEK-CHAR nil s nil foo) doesn't remove the character
that was read if FOO happens to match the character that
that was read if FOO happens to match the character that
PEEK-CHAR returns.
PEEK-CHAR returns.
- PROFILE sometimes crashes when an internal data structure was
- PROFILE sometimes crashes when an internal data structure was
set to a non-boolean value.
set to a non-boolean value.
* Other changes:
* Other changes:
- Infinite loops in map and friends caused by specifying
- Infinite loops in map and friends caused by specifying
sequence types that are too hairy for CMUCL should no longer
sequence types that are too hairy for CMUCL should no longer
occur. But CMUCL will now signal an error, which may not be
occur. But CMUCL will now signal an error, which may not be
...
@@ -266,20 +268,20 @@ New in this release:
...
@@ -266,20 +268,20 @@ New in this release:
would just use it.
would just use it.
* Improvements to the PCL implementation of CLOS:
* Improvements to the PCL implementation of CLOS:
- AMOP-style custom slot allocation supported.
- AMOP-style custom slot allocation supported.
- The value of a slot that is specified as shared in the old
- The value of a slot that is specified as shared in the old
class and as local in the new class is now retained.
class and as local in the new class is now retained.
- CMUCL supports using the class name for the :metaclass
- CMUCL supports using the class name for the :metaclass
argument.
argument.
* Improvements to Hemlock, the Emacs-like editor:
* Changes to rebuilding procedure:
* Changes to rebuilding procedure:
- Added a new script to run all of the subscripts needed to
- Added a new script to run all of the subscripts needed to
build CMUCL. Won't handle all cases, but will handle the most
build CMUCL. Won't handle all cases, but will handle the most
common builds.
common builds.
* Deprecated features:
This release is not binary compatible with code compiled using CMUCL
This release is not binary compatible with code compiled using CMUCL
19a; you will need to recompile FASL files.
19a; you will need to recompile FASL files.
...
...
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment