all: drop support for FreeBSD 13.2 which reached EOL today

Thanks again to jbeich for noticing various things I overlooked.

Reviewed by:	acm, ashish, eduardo, ehaupt, eugen, jbeich, martymac, mat,
nobukata, sunpoet, tagattie, yuri, arrowd, kde
Differential Revision:	https://reviews.freebsd.org/D45141

*: Change the MAINTAINER field per request

PR:		278267

MANNPREFIX: eliminate the use of MANNPREFIX and move to share/man

devel/tcl-trf: increase an insufficiently large buffer

This ancient problem was discovered by gcc12's new -Warray-bounds
implementation. Many thanks.

Raise PORTREVISION :(

devel/tcl-trf: fix fatal compiler warning

Starting with version 15 clang includes the new -Wdeprecated-non-prototype
in -Wall. This warns, when a K&R-style function-definitions follow
ANSI function-declarations. And TCL code has LOTS of K&R definitions...

The warning is poorly worded -- and, in my opinion, outright bogus,
but the ports need to keep building, so I'm using a trick borrowed
from src/share/mk/sys.mk to conditionally disable it.

Remove WWW entries moved into port Makefiles

Commit b7f05445c00f has added WWW entries to port Makefiles based on
WWW: lines in pkg-descr files.

This commit removes the WWW: lines of moved-over URLs from these
pkg-descr files.

Approved by:		portmgr (tcberner)

Add WWW entries to port Makefiles

It has been common practice to have one or more URLs at the end of the
ports' pkg-descr files, one per line and prefixed with "WWW:". These
URLs should point at a project website or other relevant resources.

Access to these URLs required processing of the pkg-descr files, and
they have often become stale over time. If more than one such URL was
present in a pkg-descr file, only the first one was tarnsfered into
the port INDEX, but for many ports only the last line did contain the
port specific URL to further information.

There have been several proposals to make a project URL available as
a macro in the ports' Makefiles, over time.

devel: remove 'Created by' lines

A big Thank You to the original contributors of these ports:

  *  "Waitman Gobble" <uzimac@da3m0n8t3r.com>
  *  <jkoshy@FreeBSD.org>
  *  Aaron Dalton <aaron@FreeBSD.org>
  *  Aaron Dalton <aaron@daltons.ca>
  *  Aaron H. K. Diep <ahkdiep@gmail.com>
  *  Aaron Hurt <ahurt@anbcs.com>
  *  Abel Chow <abel_chow@yahoo.com>
  *  Adam McLaurin
  *  Adam Saponara <as@php.net>
  *  Adam Weinberger <adamw@FreeBSD.org>
  *  Ade Lovett <ade@FreeBSD.org>

Remove # $FreeBSD$ from Makefiles.

The below-mentioned PR identified a problem, which should've been
caught by the test-harness -- but was not, because the harness was
not used correctly (for 10 years).

Provide a new harness and fix the reported problem as well as another
one -- with MD2 digests -- identified by the now-functioning
test-suit.

Bump PORTREVISION...

PR:		237132
Reported by:	Vadim Zborovsky

Cosmetic: remove the redundant declaration of ALL_TARGET
and an extraneous blank line from the Makefile.

Stop chasing the whims of OpenSSL and use our own -lmd
as the source of various digest-algorithms. Years ago,
when I was undertaking the replacement of the bundled
implementations with those found in the base system, it
seemed like OpenSSL will be more stable... Ha-ha.

Fix the WWW, while I'm here.

PR:		229026
Submitted by:	brnrd
Reported by:	Package Fallout

devel/tcl-trf: Add LICENSE

Strip libTrf2.1.4.so (Q/A warning)

PR:		228179
Submitted by:	Chris Hutchinson <portmaster@bsdforge.com>
Approved by:	portmgr (blanket)

security/openssl-devel: Mark ports broken cat devel

 - REASON from build log

devel/tcl-trf: Add SSL flags and allow building with LibreSSL

Approved by:	SSL blanket

Add "tea" argument to USES=tcl to facilitate porting of TEA-based extensions.

Quoting https://tcl.tk/doc/tea:

"The Tcl Extension Architecture, or TEA, by John Ousterhout and others, is a
set of guidelines and techniques for the distribution, configuration,
compilation, and installation of Tcl extensions. TEA also provides a set of
utilities that operate accordingly. Many Tcl extensions leverage the TEA
utilities, which are designed to be easily customizable."

To facilitate the porting of TEA-based extensions, I have added the new "tea"
argument to USES=tcl. This argument prepares the autoconf environment by
setting GNU_CONFIGURE and sets some commonly used CONFIGURE_ARGS. Also, it
tries to figure out the correct extension name, sets TCL_PKG accordingly, and
adds it to PLIST_SUB.

I have modified a few ports to take advantage of this. More will come.

Do not automatically run tests post-build.

Remove dependency on memchan -- the package is only needed for testing. This
will avoid the tcllib->critcil->tcl-trf->tcl-memchan->tcllib circular dependency
discussed in

PR:		200498

MASTER_SITES cleanup.

- Replace ${MASTER_SITE_FOO} with FOO.
- Merge MASTER_SITE_SUBDIR into MASTER_SITES when possible. (This means 99.9%
  of the time.)
- Remove occurrences of MASTER_SITE_LOCAL when no subdirectory was present and
  no hint of what it should be was present.
- Fix some logic.
- And generally, make things more simple and easy to understand.

While there, add magic values to the FESTIVAL, GENTOO, GIMP, GNUPG, QT and
SAMBA macros.

Also, replace some EXTRACT_SUFX occurences with USES=tar:*.

Checked by:	make fetch-urlall-list
With hat:	portmgr
Sponsored by:	Absolight

Cleanup plist

Bump PORTREV due to changing resulting binary

- Unbreak by using Debian's reimplemented MD2
  https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=664713
- Stage
- Convert to USES=tar

- Still broken as it depends on MD2 and md2.h still.

./generic/md2.c:31:10: fatal error: 'md2.h' file not found

With hat:	portmgr

Automatically switch to using -lmd instead of -lcrypto on 10.x,
where OpenSSL, most tiringly, no longer provides <openssl/md2.h>.

Submitted by:	pkg-fallout@

- Convert to USES+=tcl
- USE_LDCONFIG
- Reorder variables to make portlint non-fatal

Add NO_STAGE all over the place in preparation for the staging support (cat:
devel part 4)

Fix the build in case, a stricter compiler and Tcl-8.6 are used. Keep
compatibility
with Tcl-8.4, however.

Notified by:	pkg-fallout@

Correct the typo in previous commit.

Notified by:	gahr

Seek memchan's pkgIndex.tcl in its current location.

Obtained from:	gahr

-remove MD5

Use SourceForge's mirrors, now that the distfile is available there AND
no longer available in the interim location.

Pushed by:      erwin (pointyhat)

Add ``regression-test'' and ``check'' aliases to the test-target. The
first one is, I'm told, invoked by our build-cluster while building
official packages. The second is common enough out there for some
people to attempt...

Add the plist, which I forgot to re-add yesterday, when restoring this
port from the Attic.

Notified by:    The Restless Daemon

Resurrect this port from the Attic based on the new version from the
author. As before, we modify the author's code to use either OpenSSL's
-lcrypto or BSD's -lmd /directly/ instead of compiling the bundled
definitions digests.

We also link with -lz and -lbz2 instead of dlopen-ing them at run-time.

Remove devel/tcl-trf as it has been marked BORKEN for over 5 months.

- Mark BROKEN: self-tests fail

Reported by:    pointyhat

- Move from versioned tcl/tk CATEGORIES to simple tcl and tk categories

With hat:       portmgr

While trying to upgrade devel/tcllib several test failures were detected
in the RIPEMD128 implementation.

Turns out, the problem was fixed a few months ago by the TRF author in CVS.
This commit adds a patch and bumps portrevision.

Although the same problem existed in the TRF's RIPEMD160 implementation,
it did not affect this port, because it uses the OpenSSL's (or -lmd's)
implementations of all digests -- except those, which are considered
obsolete and no longer shipped. Like rmd128.

Thanks to:      Denis Shaposhnikov

Remove USE_REINPLACE from all categories starting with D

SHA256ify

Approved by:    krion@

Don't compile haval.1996/haval.c separately, but as part of the top
haval.c (as in vendor's code). This works around weirdness in vendor's
endianness-determining pre-processor code and unbreaks sparc64. Much
rejoicing...

Add a patch ensuring uint32_t is used, where a 32-bit unsigned
integer is implied. This fixes RIPEMD128 on Sparc64 and AMD64, as
well as HAVAL on AMD64.  HAVAL on Sparc64 remains broken due,
apparently, to an endiannes issue, which, for now, escapes my
understanding.

Unbreak on amd64.

Retreat to using OpenSSL's message digest routines instead of -lmd. They
are sometimes faster and never slower. Using -lmd is still possible by
building with ``TRF_USE_MD=yes''. Bump PORTREVISION.

Use the RIPEMD160* routines from -lcrypto (or -lmd) instead of our own
-- just like for md[25] and sha*. Neither -lcrypto nor -lmd provide
RIPEMD128 (not strong enough?), so stay with our own implementation.

sparc64/amd64 are not expected to work yet...

Use ``$(CC) -shared'' insted of ld to create shared library. This
allows to use non-default compilers, such as icc.

Explicitly include <sys/types.h> and <unistd.h> to fix the breakage on
recent -current. Fix warnings -- on x86, at least, the package now
builds with ``-Wall -Werror''.

Also fails testsuite on amd64.

BROKEN on sparc64: Fails testsuite during build

Unbreak. Builds, installs fine on fresh current. Disable the parallel
build, as the only thing, that could've prevented it from properly
installing in the past.

SIZEify (maintainer timeout)

Per bento logs, mark as broken on 5.x, although not for the usual gcc
reasons.  Informed maintainer.

Parametrize the version number to unbreak packaging -- we switched
the default to Tcl-8.4, but the version in pkg-plist was hardcoded
as 8.3

Prodded by: kris

Change the default TCL_VER to 8.4 to match that of tcl-memchan.

Submitted by:   kris, the ports janitor

Clear moonlight beckons.
Requiem mors pacem pkg-comment,
And be calm ports tree.

E Nomini Patri, E Fili, E Spiritu Sancti.

- Use -E, since Ex. RE are used;
- Fix plist and therefore bump PORTREVISION

Noticed by:     bento

PERL -> REINPLACE_CMD.

Make work with the new bsd.port.mk. The benefit of the magick, that
used to work to figure out the name of the installed library automaticly
is, probably, not worth the disadvantages USE_SUBMAKE would bring. Just
hardcode the SHLIB_NAME and figure the SHLIB_LINK from it.

While I'm here, add tcl-memchan to the BUILD_DEPEND, so the automatic
self-testing is done every time this package is built

Submitted by:   bento

Resolve  a  few  more  symbols  at compile/link  time,  rather  than  at  
run-time. This  fixes the  bz2 commands.  We link  against -lmd,  so use  
<md2.h>  and <md5.h>,  instead  of <openssl/md[25].h>  (why  do we  even  
install those separately?). Bump up portrevision.    

From the software's self-description: