Since FreeBSD 8.4 and FreeBSD 9.1 make(1) do support :tu and :tl as a
replacement for :U and :L (which has been marked as deprecated)
bmake which is the default on FreeBSD 10+ only support by default
:tu/:tl a hack has been added at the time to support :U and :L to ease
migration. This hack is now not necessary anymore
Note that this makes the ports tree incompatible with make(1) from
FreeBSD 8.3 or earlier
With hat: portmgr
- Remove references to Tcl/Tk 8.3 from bsd.tcl.mk
- Update ports that required 8.3+ to require 8.4+ (since 8.5 is the
default, I don't expect this change to cause much trouble)
Chase the lang/tcl85's decision to install man1/tclsh8.5.1 instead of
man1/tclsh.1 -- now, when we can't find a man-page foo, we try looking
for foo$tcl_version.
PR: ports/140978
Submitted by: olli hauer
should now work properly.
Add patch to fix profiling on 8.5 AND whenever the system has been up
for more than 25 days -- the latter used to overflow the clock_t-sized
variables, because TclX uses these 31-bit values to store milliseconds.
Use the proper method to obtain CLK_TCK -- if only on FreeBSD.
in an unintended call to write ZERO bytes to any descriptor, that was
opened for writing, when closing it.
When the descriptor belongs to a pipe, with its other end closed -- such
as when the started process has exited, doing so results in a SIGPIPE.
Tcl's core ignores SIGPIPEs, but TclX' self-tests didn't and began to
fail.
The bug should now be fixed in Tcl, but this patch is needed for TclX to
pass its own tests even when using the uncorrected version of Tcl.
care for, because the line was never reachable on FreeBSD.
Our new and improved sh cares deeply about such problems, however, and
configure started to fail even on 7-STABLE at the beginning of October.
Notified by: miwi
- Switch to USE_TCL
- s/INSTALLS_SHLIB/USE_LDCONFIG/
- Support tcl84-thread
- Fix tclsh detection in configure
- Pet portlint.
Note: this port is still broken, because a test is failing, and tests
are part of the all target.
PR: ports/110805
Submitted by: /me
Approved by: mi (excepted the move of tests to a regression-test target)
years of idleness. The changes are fairly substantial -- all Tk bits,
however insignificant they were, are now removed completely, for
example. No TclX enabled executable is built/installed either -- a Tcl
script is expected to load the Tclx package via the "package require
Tclx" only.
The changes to port add the handling of SIGINFO (made possible by the
recent fix-up of lang/tcl84), running the authors' test-target as part
of the build, and skipping building/installing of help files in the
NOPORTDOCS case.
The devel/tcl-neo is the only dependant port of tclX and is updated to
depend on the new version.
eliminate all warnings (on i386) and const-ify the KeyedList-API and
others (patch-warnings grew up).
Fix a typo in the previous commit (s/KeyList/KeyedList/g in MLINKS).
Bump PORTREVISION again.
. when parsing man-files, do not choke on files
containing '.so man.macros';
. be ready for the .bz2 man-pages;
. do not use vendor's install-doc target at all -- install
the vendor's manual pages in post-install and declare
MLINKS to have bsd.port.mk do the Right Thing (TM) --
links instead of copies of the manual pages.
This may even work now :-)
It's worth pointing out that the Tcl project used to
*recommend* that application writers copy this kludge.
These ports work fine with Tcl 8.3, which provided a fake
matherr() even if the OS didn't, but here's what the
Tcl 8.4 changelog has to say:
2002-05-31 (dead code)[474335,555635] removed all use of matherr() (english)
*** POTENTIAL INCOMPATIBILITY ***
The following lines can be safely removed from any application that
has them, unless compatibility with SunOS 4.X is desired.
extern int matherr();
int *tclDummyMathPtr = (int *) matherr;
The build problems were pointed out by Kris, as usual. ;-)
of TclPro made TclX-8.3 available last year. There is not a whole
lot of changes, and most of them deal with the Windows parts of the
thing, but here is the upgrade 8.3 anyway. Please, test...
I think, it is time to repocopy tclX82 into tclX and get rid of
tclX75, tclX80, and tclX82...
some occasions, this prevent it from being loaded properly, even
though other libs already loaded already bring in -lm. I don't
understand the rtld :(
Only build the Tk related help files, if WITHOUT_X11 is not
specified. Previously they were built, but not installed. If you
did not have Tk installed, that build failed for you -- see PR
29722...
PR: ports/29722
myself (for the lack of children, whom I could've prohibited to
do it). Sorry.
Upgrade this port to:
. build against TCL-8.3 by default (controllable by TCL_VER)
. build with or without TK (controllable by NO_X)
. take over maintainership -- regretfully, Justin was
rather idle recently
. build the helpfiles once -- during the build stage --
not during the install stage
On a side note, I more and more resent the fact, that our TCL
8.3 is built with the -stubs. It just introduces more variables
without noticeable benefit. On FreeBSD shared libraries work
well...
I tested this with TCL-8.3 (with and without TK), and with
TCL-8.2 (without TK only). Please, test this more.
the road to obliterating tcl/tk 8.0 from the tree.
Note 1: the patches have been extensively re-worked from the
original submission, to make this port PREFIX-safe,
and also to deal with multiple patchfiles updating the
same source file.
Note 2: I've left the MAINTAINER as jseger. Justin, you may
want to cast your TCL-expert eye over this one, it seems
to do the right thing.
PR: 16970
Submitted by: Mikhail Teterin <mi@aldan.algebra.com>
