Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
(true for 1.5, but officially branded as such for 6), so this is called
"sun-j{dk,re}6" rather than "sun-j{dk,re}16".
amd64 support is not currently included, but initial provision exists in
the sun-jre6 package as it was cloned initially from sun-jre15.
|
|
|
|
|
|
|
|
|
|
maintainer as it obviously doesn't honour CFLAGS et al.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
solaris in 64 bit mode. Patch from Gilles Dauphin.
|
|
*sigh*
|
|
|
|
- Update lang/sun-jre15 to 5.0.11 from 5.0.10
Changes since 5.0.10:
hotspot terminates at assert during GC in java_g
Unexpected async exception in jvmti_GetLoadedClasses()
icache invalidation code can segv
Need a way to determine if java process was started from sun launchers
unmarshal error on CORBA alias type in CORBA any
Regression: UnicodeLittle broken on 1.4.2_12
PrinterJob.printDialog() crashes JVM when called from subthread
Few Hindi glyph formations are failing on Win-Vista only
StackOverflowError due to recursion at sun.font.TrueTypeFont.getTableBuffer
Need to make Java process High-DPI-aware on Windows Vista
sun/java2d/DirectX/D3dTranslucency.java test needs to be backported from Mustang to Tiger workspace
java/awt/font/StyledMetrics/StyledMetrics.java fails on Vista
deriveFont is not working on FontUIResource with a dynamically loaded TTF
Unable to retrieve printer list on system with unconnected printers
RI crashes on Windows when executing tests.api.java.awt.Toolkit.CursorTests
Pressing 'copy' on a TextField with echo char set copies the actual text to the clipboard, XToolkit
Regression in backport of flickering-related bug fix
Vista: The right mark of the CheckboxMenu item is broken when compared with winXP
Uncanonicalized absolute filepath with length 248-260 no longer works (win)
(process) subprocess environment sort order differs from Windows native sort order
(process) Process.destroy() can kill wrong process (Unix)
VISTA: api/java_net/NetworkInterface/index.html#static fails.
SocketTimeoutException : Timing out early
Socket creation on Windows takes a long time if web proxy does not have a DNS entry
HTTP Request-URI contains fragment when connecting through proxy
javax/sound/sampled/Clip/JavaSoundAudioClip/bug6269854.java test appeared as unstable on tiger
Regression: com/sun/java/swing/plaf/gtk/5048769/bug5048769.sh fails
OceanTheme causes swing components to serialize with sun.* classes
Vista:FileChooserDemo throws an exception when trying to open it in Windows L&F
JFileChooser throws exception when trying to access a CD drive on Vista.
REGRESSION: AWT-EventQueue Deadlock AWTTreeLock/MetalUtils$GradientPainter
REGRESSION: JCheckBox doesn't show on JToolBar under Windows L&F
(fmt) java/util/Formatter/Basic.java fails in hi_IN locale
(tz) Support tzdata2006n
(tz) Support tzdata2006p
(tz) Support tzdata2006o
"com.sun.jdi.InternalException: Inconsistent suspend policy" in internal event handler
Vista: JRE 5.0u8 installation failed with 1722 MSI error
Vista: Add manifest to installers to run with elevated privileges
Vista: Use registry key to pre-approve Active-X controls
1.5.0_u7 has "shielded" exe in Windows Vista
[l10n] l10n meeesage update after deployment bug 6458761 is fixed
rmic does not use manifest classpath
Vista: disable change dialog for cache location ?
Vista: Implement work around for Runtime.exec() problem in deployment
InstallStats2 not sending full version for offline installers with 'failure' status
IS2 : InstallComplete ping 'ic' with status 'success' is not posted in silent mode
Vista: Java auto-update is not enabled on Windows Vista
Vista: jusched.exe requires user approval with every login
jusched.exe needs to be signed
dead lock occurs by pressing "refresh" many times in 5.0u10b02
REGRESSION: JFrame goes to back after toFront() is called when Applet starts
vista: changes made by java control panel has no effect
browser become non-responsive if we switch page during applet JAR initial download
TCK-jnlp 1.5: jnlp_file/resources/j2se/index.html#vmArgs hang in some cases.
|
|
newer versions (11) of solaris. Verified that the 'bmake test' output
is unchanged on 32-bit solaris-2.9/sparc.
|
|
|
|
|
|
|
|
from the original, that it has become easier to maintain it as its own
file instead of a patch, so do that.
- Improve the shell quoting quite a bit in the f2c-f77 script. Now you
can pass things like -DFOO=this\ is\ a\ test and have the escapes
actually work when the arguments get passed down to the underlying
commands. Many thanks to Alan Barrett for help with the shell quoting.
- bump pkgrev
|
|
|
|
NetBSD's GCC 3.3.3.
|
|
ftp://asim.lip6.fr/outgoing/packages/i386/3.1/20070114.1132/broken.html
(latest 3.1/i386 bulk build of 2006Q4).
Feel free to fix them...
|
|
- Include options.mk before rubyversion.mk, so PLIST for ri database
should be created suitably.
- make RUBY_RIDIR and its friends relative path to ${PREFIX}.
- Fix and improve handling of ${RUBY_RIDIR} handling and should
be fixed remaining ${RUBY_RIDIR} after pkg_delete ruby18-base.
(Noted by private mail from wiz@.)
Bump PKGREVISION of ruby18-base package.
|
|
Pointed out in PR 35504 by Ryo HAYASAKA.
|
|
|
|
|
|
not in pkgsrc and therefore not generally available.
PKGREVISION++
|
|
|
|
|
|
|
|
in extconf.rb to check the availability of DBM. This makes the package
build on Darwin.
|
|
|
|
|
|
|
|
|
|
|
|
Important changes excerpted from web pages:
2007/1/17
Gauche 0.8.9: Major maintenance release
+ Bug fixes
+ Miscellaneous improvements:
2006/11/18
Gauche 0.8.8 important patch: There is a bug in main.c that makes gosh
exits silently without reporting errors when a Scheme script raised an
unhandled error. Please apply the patch shown in the following message:
http://sourceforge.net/mailarchive/forum.php?thread_id=30949517&forum_id=
2043
2006/11/11
Gauche 0.8.8: Major maintenance release
+ Important Changes:
o Exact rational number is supported; now you get 1/3 from (/ 1 3).
To obtain inexact number from division of two exact numbers, you
have to use exact->inexact explicitly. With this change you can get
more exact result, but if your code has relied on the old Gauche
behavior that automatically converts rationals to inexact reals,
your code may run very slowly in this release of Gauche (since
exact rational arithmetic is much slower than flonum arithmetic).
For the smooth transition, a compatibility module compat.norational
is provided, which makes the / operator behaves like before. See
the manual entry for the details.
o The reader is more strict about utf-8 encoding. Consequently, some
source files in other encoding that happened to be accepted by
previous versions of Gauche may no longer work. If you get an
error, either convert the encoding of the source, or use "coding:"
magic comment (See "Multibyte scripts" section of the reference
manual).
o The test-module routine in gauche.test is fixed so that it detects
more references to undefined global variables that have been
overlooked. You may get an error something like "symbols referenced
but not defined: ...". In most cases, they are from typos. See the
manual entry of gauche.test - Unit testing for the details.
o New modules:
# sxml.serializer: Generic routine to convert SXML to other
formats like XML or HTML. Written by Dmitry Lizorkin and ported
to Gauche by Leonardo Boiko.
# util.trie: Implementation of Trie. Originally by OOHASHI
Daichi, and hacked by numerous Gauche hackers.
# util.rbtree: Implementation of Red-Black Tree. Written by Rui
Ueyama.
o A bug in port locking routine, that caused a race condition on
multiprocessor machine, is fixed. As a side effect, port lock
operation became a bit faster.
o C API prospected change: Scm_Eval, Scm_EvalCString, and Scm_Apply
will have different API in the next release. The current API is
kept under a different name, Scm_EvalRec, Scm_EvalCStringRec, and
Scm_ApplyRec. If you are using those functions, please make changes
until the next release.
+ Miscellaneous fixes and improvements:
2006/4/12
Gauche 0.8.7: Major maintenance release
+ Bug fixes:
+ Improvements:
|
|
|
|
It's disabled by default, to avoid too much dependancies
|
|
clear that these variables are completely unrelated to
BUILDLINK_TRANSFORM.
Added a legacy check that catches appearances of BUILDLINK_TRANSFORM.*.
XXX: Where should incompatible changes in pkgsrc be documented?
|
|
|
|
the change at all.
|
|
packlist is used to generate the pkgsrc PLIST. This is the case for most
of the Perl modules, so the default value is "yes".
Currently, there is no change in functionality. After some further
testing, the default will be that a Perl module provides a packlist and
needs to say if it doesn't.
|
|
|
|
|
|
|
|
I would list the changes to the package here, but SBCL makes a new
release every month, and there are usually a dozen or more minor
changes per release. The previous version of this package is about
18 months old...
|