is available in lang/py-mx-base and all the dependent ports seem to
work fine with py-mx-base.
The removal of py-mxProxy was approved by Johann Visagie
<johann@egenetics.com>.
According to my tests it is *amazingly* efficient - it gave me about 10%
memory saving (SIZE in the top(1)) for the large processes like X and jre
without any measureable performance saturation.
Moreover, due to not very clear for me reasons Python benchmark (pybench)
is about 60% (!!!) faster with this allocator comparing to the libc one.
Obviously we should investigate this further and if there is no error
then tune either Python or our own malloc.
Alternatively we may evaluate a possibility to make it (or part of it) our
default malloc(), because it is friendly licensed and actively maintained.
- Put kdesupport11 dependency back. I forgot that it is not implied by
kdelibs11 as kdesupport2 is by kdelibs2.
- HAS_CONFIGURE -> GNU_CONFIGURE in some cases were needed.
- Add X11 headers to CPPFLAGS for moonshine (extreme case).
- bsd.port.mk update to use bsd.kde.mk for USE_{QT,KDE}*
- Cleanup corresponding ports for bsd.kde.mk update.
- Fix bsd.kde.mk: use correct kdelibs dependency, put qt at the bottom,
introduce QT_NONSTANDARD variable for nonstandard configure setup.
- Update KDE2 to 2.1.1. Two patches included in x11/kdelibs2 to fix the
proxy authentication that was broken for 2.1.1. Remove old patches.
- Potentially fix kdelibs build for alpha.
- Fix qt-designer 2.3.0 build.
- Ruby stuff left alone since it looks like black magic to me. Should
still work w/ compat shims for older USE_QT[,2] style. Some others
were also left alone for the same reason.
Reviewed by: portmgr, ports (bsd.kde.mk+bsd.port.mk)
Submitted by: David Faure <faure@kde.org> (proxy auth patches)
Alex Zepeda <garbanzo@kde.org> (old patches removal)
Among four submissions, this commit based on the port sent by
Ernst de Haan <ernst@jollem.com> at ports/24291.
PR: 21435, 23368, 24291
Submitted by: Leo Kim <leo@florida.sarang.net>,
Dave Glowacki <dglo@ssec.wisc.edu>,
Sean Blakey <sean@beastie.bellevue.virtualtek.com>,
Ernst de Haan <ernst@jollem.com>
Also thanks to: Palle Girgensohn <girgen@partitur.se>,
Richard Stockley <rws@procopia.demon.co.uk>,
Kees Jan Koster <kjkoster@kjkoster.org>
It features an user friendly IDE, project wizard, generation of GNU
makefiles and automake/autoconf skeletons, fast run-time source parsing
with syntax coloring and autoformating of C, Fortran, and Eiffel sources.
CvsGui features
WinCvs is written in C++ using the Microsoft MFC.
MacCvs is written in C++ using Metrowerks PowerPlant.
gCvs is written in C++ using GNU gtk+.
They are using the latest cvs source code.
They are making cvs easier for the novice.
They are increasing the power of cvs by providing an high-end interface.
The project is growing because it is supported and developed
by several cvs users.
WWW: http://cvsgui.sourceforge.net/
o don't apply bitwise shift to components when setting palette - vgl don't
need it unlike fbcon (after which libvgl driver was modelled). Bump
PORTREVISION as a result of bugfix.
o use internal freetype2 for consistency with x11/XFree86-4.
o added xthreads obtained from x11/XFree86-4.
o install "ws" type config sample for xdm.
o build DRI only if kernel source installed in /sys.
o fix Riva128/SGRAM driver(patch-riva_hw.c).
PR: 24338(4.0.2)
Submitted by: maintainer, keith
this involves is this: Cull GL from Qt by default, but still provide a
Qt+GL library that may or may not have threads. Then also provide a Qt
library that has threads but not GL. This allows us to make KDE2 depend
on a library that will *not* have threads, ever. Threads will be
revisited at a later date. Ports that require GL support need to be
updated to use the hacked library, libqtgl.so.4. The net result is that
we bloat our qt2 package by 1.5-2.5MB for compatability. Also, static
qt will not have GL support.
Introduce bsd.kde.mk, which will be tested on bento before becoming
fully activated.
Replace qt22-static with qt2-static, since it's just a proxy. Update
qt-designer to depend on qt23. Also make the old hack to package the
correct lib obsolete by using PLIST_SUB instead.
Miscellaneous changes: remove LIBQTFILE from CONFIGURE_ENV, it's not
used anymore. Solve namespace pollution problems with the devel/pth and
devel/libgnugetopt ports. Hopefully.
Suggested by: ade, asami, sobomax (bsd.kde.mk)
Repocopied by: asami (qt22-static --> qt2-static)
port (see below).
- Submitted this to current maintainer on Wednesday 14 March; no response yet
- The Python modules were not correctly installed to the Python site library
directory before. They are now. (Thanks to Michael Ewe <m.ewe@t-online.de>
for pointing out the issue.)
- Used PORTDOCS pragma in $PLIST.
PR: 25934
Submitted by: Johann Visagie <johann@egenetics.com>
the version to 2.3.2.
- Add checks for empty files, empty directories, core files, more
possible backup files, dotfiles, symlinks and CVS directories.
- Do not assume PATCHDIR always includes "/files/". Use the best
method to check whether a file is added to @checker as a patch file.
- Some trivial message style fixes.
The State Threads is a small application library which provides a
foundation for writing fast and highly scalable Internet applications
(such as web servers, proxy servers, mail transfer agents, and so on) on
UNIX-like platforms. It combines the simplicity of the multithreaded
programming paradigm, in which one thread supports each simultaneous
connection, with the performance and scalability of an event-driven
state machine architecture. In other words, this library offers a
threading API for structuring an Internet application as a state
machine.
The State Threads library is a derivative of the Netscape Portable
Runtime library (NSPR).
WWW: http://oss.sgi.com/projects/state-threads/
PR: 25189
Submitted by: tobez@tobez.org (Anton Berezin)
by forcing the CFLAGS to -O -pipe. Somehow, the alpha build always
tries to enforce a particular -mcpu=ev4 flag which of course cannot be
understood by the (AVR) xgcc later on. This looks to me like a bug in
the cross-compilation environment of gcc, but i'm tired of actually
finding the bug.
The compiled result of avr-gcc MD5 compares equal to something build
from an IA32 host platform.
next version. In the meantime, apply with the port.
FYI, the bug is demonstrated by this program:
#include <stdio.h>
#include <stdlib.h>
#include <pth.h>
static int check_something(void *arg)
{
return (0);
}
static void *thread(void *arg)
{
printf("thread sleeping five seconds...\n");
pth_nap(pth_time(5, 0));
printf("thread is done sleeping\n"); /* bug: we never get here */
return (NULL);
}
int main(int ac, char **av)
{
pth_event_t ev;
pth_init();
pth_spawn(NULL, thread, NULL);
ev = pth_event(PTH_EVENT_FUNC, check_something, NULL, pth_time(2, 0));
while (1)
pth_wait(ev);
}
Since gcc (in the assumption of generating a native compiler) doesn't
want to cbe configured for an alpha*-*-freebsd* system, we hack the
configure script to allow this (similarly to netbsd). In the end, all
this will be ignored anyway since it's getting to become a
cross-compiler.
redirect to a new server, but they don't carry the old files anymore.
Unfortunately, in order to build the new stuff, we'd need a newer
avr-gcc first -- and that one's by now only available directly from
gcc's CVS repo, which is a little too `green' to me. Let's wait
another couple of months until they rolled a newer release of gcc and
binutils, and switch then.
By now, store the old distfile elsewhere.
Submitted by: fenner's ports build survey (very valuable service!)
<support@scitools.com> Tech Support from Understand C/C++ manufacturer,
we are allowed to remove the RESTRICTED tag as well as the LEGAL
warning.
Update to version 1.4 Build 104 from March 8, 2001.
- Correct a port versioning typo:
o Instead of using PORTVERSION=14b104 which both reflects the
distfile naming convention and follows the naming convention used
in this port; we will switch to PORTVERSION=1.4.104 which reflects
the Changelog information
o To avoid "who came before" problems since last port was
PORTVERSION=14b103, BUMP PORTEPOCH.
- Fix a PLIST unnecessary warning
- Version Changelog: http://www.scitools.com/ucpp_build_log.html
- textproc added in category (like astyle)
Ccdoc is a tool for extracting comments from C++ source code and presenting it
in HTML format, very similar to Java's JavaDoc tool. The tagging used in ccdoc
is very similar to that of Javadoc, with adaptations for the C++ specifics, of
course. Ccdoc supports extracting comments from both header and implementation
files.
In contrast to most other C++ doc'ing applications, ccdoc analyses the code
before it has been run through the pre-processor, so things such as macros can
actually be included in the documentation.
It's usage is not quite as straight forward as JavaDoc's, but considering the
quality of the output, it is well worth the effort.
WWW: http://www.joelinoff.com/ccdoc/
PR: 22794
Submitted by: lonewolf@flame.org
I was planning to rename pyncurses into more appropriate py-ncurses
before committing it out, but forgot to do so. Therefore delete
pyncurses and readd it back as py-ncurses.
Py-ncurses is a Ncurses binding for Python.
PR: 18633
Submitted by: adsharma@sharams.dhs.org
- fix installation path of this perl module because this command failed:
perl -M'ConfigReader::DirectiveStyle' -e 1
ConfigReader files have to be put under
lib/perl5/site_perl/%%PERL_VER%%/
instead of
lib/perl5/site_perl/
- now my new port FlowScan finds the ConfigReader::DirectiveStyle PM.
"configure:1271: checking for ConfigReader::DirectiveStyle"
has strange problems downloading it from the author's site. I really puzzled
because I can't reproduce those problems even from freefall.
Submitted by: bento (sorta)
Use EXTRACT_SUFX instead of single DISTFILES (by kuriyama).
Submitted by: Dave Chapeskie <dchapes@ddm.crosswinds.net>
PR: ports/25711
Not reach for: maintainer (mail service unavailable)
AutoGen is a tool designed for generating program files that contain
repetitive text with varied substitutions. Its goal is to simplify the
maintenance of programs that contain large amounts of repetitious text.
This is especially valuable if there are several blocks of such text that
must be kept synchronized.
manually add the dependency for autoheader(1), but don't have the ports
infrastructure run `autoconf' (which clobbered the top-level configure
script).
should fix the port build on bento.
Still doesn't want to be built on the alpha arch, i'm not sure whether
i'll be able to fix that or whether i'll have to exclude it from the
alpha build. In theory, since it's a cross-compiler already anyway, it
should be possible to build it on non-i386 platforms as well.
Understand can parse a C/C++ project helping reverse engineer it
[begin snip from the www site]
Understand parses any sized C or C++ project to help you reverse
engineer, document and understand it and thus maintain it better.
It supports K&R C, ANSI C, or C++ source code. Projects can contain
mixed C/C++ code as well. The parser is efficient, fast, and can
handle very large projects. Very little is needed to get started
- just aim it at your source tree. Optionally (and for more parsing
accuracy) add any externally defined macro definitions and include
paths. All of this is done from the GUI - you should be able to do
an initial analysis of your project in a few minutes.
Understand for C++ quickly documents Class inheritance hierarchies
(Base Class and Derived Classes), Call and CallBy Trees, Include
and Include By Trees, as well as where and how everything in your
source code is used (cross reference). Understand for C++ creates
detailed automatic documentation about your source code in HTML
and text reports Using the PERL and C API you can write your own
documentation generators.
[end snip from the www site]
You need a limited time period license to run the product in
evaluation mode. One can be obtained in the www site. A permanent
license can be purchased there too.
- Instalation issues:
This port has a hardcoded directory structure which is required to
run it. Therefore, some measures are necessary to install it under
FreeBSD following handbook rules'
1) All files but binaries are installed according to hier(7)
2) A fake structure is created under ${PREFIX}/lib/understand_c to
please the program and soft links are made there to the correct
locations under hier(7)
3) The binaries are installed under ${PREFIX}/lib/understand_c/bin
and a wrapper is installed under ${PREFIX}/bin pointing to the fake
structure
4) Since a license is required to run the program, warnings are
issued when either it is installed or a package is added
5) Since the legal status is still being debated with the developers,
it will be marked as RESTRICTED for the time being and LEGAL
will be updated accordingly
* Fix the build problem that was created when bsd.port.mk started setting
CXXFLAGS in MAKE_ENV. The problem is if a port uses a BSD-style Makefile
and C++, sys.mk will not craft a proper CXXFLAGS because it uses "=?".
KDE Studio is an IDE (integrated development environment) for the
K Desktop Environment (KDE).
PR: 24576
Submitted by: Trenton Schulz <twschulz@cord.edu>
Allegro is a cross-platform library intended for use in computer games and
other types of multimedia programming.
The WITHOUT_DEVEL option will prevent installation of some development
utilities, while WITHOUT_MAN won't install the (many) manual pages.
More about allegro at http://www.talula.demon.co.uk/allegro/
PR: 25331
