• https://ftpmirror.gnu.org/libtool/ (Download)
• https://ftp.gnu.org/gnu/libtool/ (Download)
• https://mirrors.dotsrc.org/gnu/libtool/ (Download)
• https://ftp.nluug.nl/pub/gnu/libtool/ (Download)
• ftp://ftp.gnu.org/gnu/libtool/ (Download)
• http://ftp.funet.fi/pub/gnu/prep/libtool/ (Download)
• http://ftp.kddilabs.jp/GNU/libtool/ (Download)

• (2011-11-02) Updated to version: libtool-2.4.2
• (2011-09-27) Updated to version: libtool-2.4
• (2008-11-17) Updated to version: libtool-1.5.26
• (2006-10-24) Updated to version: libtool-1.5.22p6
• (2006-10-22) Updated to version: libtool-1.5.22p5
• (2006-10-21) Updated to version: libtool-1.5.22p1
• (2006-08-23) Updated to version: libtool-1.5.22p0
• (2006-07-21) Package added to openports.se, version libtool-1.5.22 (created)

Log message:
default for PKGNAME-sub where it makes sense, no pkgnames change

Log message:
bump REVISION where my screwed-up commit changed @comments in plists

Log message:
revert accidentally committed files

Log message:
update to py-ptyprocess-0.7.0, make py3-only

Log message:
replace simple PERMIT_PACKAGE_CDROM=Yes with PERMIT_PACKAGE=Yes

Log message:
garbage collect CONFIGURE_SHARED

Log message:
fold lib-only PFRAG.shared into PLIST, and some related cleanups

Log message:
Drop remaining MD5/RMD160/SHA1 checksums.

Log message:
move depends to new style

Log message:
Remove g77 dependency.  The generated libtool is the same.  (Except
for now also allowing static linking for Fortran, which was previously
misdetected.)  ok maintainer

Log message:
Ouch, if libtool is built without g77 around, then it can't build fortran.
BUILD_DEPENDS only (on gcc4 systems), should be possible to CONFIGURE_ENV
things out, to revisit later.

Log message:
backport important bug-fix from newest ltdl.
There are probably other bad bugs in there, this code is a textbook example
of how NOT to write code if you want it to work.
Between the gnu style that obfuscate the stuff, the various thingies added
"just in case", the misuse of autoconf results, and the really high
abstraction level together with long functions and internal concepts, how
can anyone make sense of this code ? (short answer: it doesn't make actual
sense).
Also, no cookies for the gnu guys who, along with an important bug-fix,
manage *again* to push out 10000+ lines of useless diff thanks to a change
in automake/autoconf itself.
We don't need to make fun of Windows  and its security, the GNU project
manages to produce as many problems on its own...
When will they learn secure practices ? I would bet "not this century",
but then I probably won't be around to collect the bet...

Log message:
Override detection of GNU sed otherwise, if it's installed at build
time, the generated libtool script tries to use 'gsed' instead of 'sed'.
ok Brad.

Log message:
Do not revert the list of libraries in the generated .la file.
This used to be needed but with the current libtool it breaks
the library order in some cases (including xenocara) for static
only archictectures.
ok naddy@.

Log message:
fix autoconf version.
from brad@ (maintainer)

Log message:
update to libtool 1.5.26
from brad@ (maintainer)
ports build testing by naddy, ajacoutot, and me

Log message:
tweak FAKE_FLAGS semantics to saner defaults.

Log message:
Change Brads email address for now, so people can reach him instead of
getting bounces or 550s. Bump pkgnames. While here, remove quotes from
comments.
prompted by espie@

Log message:
- bump PKGNAME to force upgrade as former powerpc snap package did not
contain the correct revision of patch-ltmain_in (thanks to CVS tag and
yours truly's stupidity)
gnome-applets,gstreamer-0.8 and gdm should be able to build correctly now
thanks to kurt@ for discovering my brain damage...

Log message:
Propagate -pthread in dependency_libs so linking with -pthread makes
its way up the dependency chain. no objections usual suspects, okay
maintainer Brad <brad at comstyle.com>

Log message:
Force to use /usr/bin/grep at build time, even if GNU grep is
installed.
ok espie@, steven@

Log message:
modify libtool to run install with *no* executable permissions on files that
do not need it.
After discussion with matthieu, theo, naddy, and a lot of tests, that
caught up the fact that sometimes, libtool uses cp, so we let it use cp
(for now).

Log message:
base64 distinfo with SHA256

Log message:
Add an explicit check for shared libraries support in autoconf bits.
ok brad@, todd@.

Log message:
Bump pkgname for 4.1; add SHA256.
ok bernd@

Log message:
new MULTI_PACKAGES

Log message:
cosmetic change in shared_libs.log

Log message:
fix, again, sigh.

Log message:
The comments make it not work, have to use
SHARED_LIBS +=

Log message:
oops, forgot to strip the `lib' part.

Log message:
much cleaner version of the shared libs log

Log message:
let build produce a SHARED_LIBS_LOG file which contain lines like
SHARED_LIBS+=	<openbsd_version>	# <original libtool version>
Those lines you can put into a port's Makefile... and so notice more
easily what the software people upstream change.

Log message:
bump PKGNAME for 4.0
noticed by Deanna Phillips <deanna at sdf.lonestar.org>
ok espie@

Log message:
rev 1.174.2.21
(LT_DLMUTEX_GETERROR): check if `lt_dlmutex_geterror_func' is set
instead of `lt_dlmutex_seterror_func'.
>From libtool CVS

Log message:
upgrade to libtool 1.5.22
Thanks to naddy@ for testing this
with a full ports tree build.

Log message:
override -version-info through env variables, as discussed with
bernd, steven, nikolay, naddy, pvalchev...

Log message:
Redundant patch as pointed out by jakemsr@. The libltdl copy of ltmain.sh
will be regenerated from ltmain.in.

Log message:
Fix a FAKE_FLAGS pasto. No PKGNAME bump required.
ok espie@

Log message:
- Use new SHARED_LIBS framework
- USE_LIBTOOL and bump PKGNAME where appropriate
libtool okay by brad@, espie@

Log message:
- remove non-existent dirs from libsearch path.
- if DESTDIR is defined, avoid relinking (relinking doesn't make sense
with the fake framework anyways).
okay brad@, `as long as it doesn't break anything'.

Log message:
Minor fixes:
- do not link -lresolv, even though this will be gone from -current.
- ld.so path is just /usr/lib, no /lib on OpenBSD.

Log message:
Fix two annoying issues (fixed upstream as well):
--tag CXX works for relink now.
--tag=disable-static/disable-shared is now independent from language tags,
thus libtool --tag=disable-static --tag CXX works.
Almost allows KDE from using libtool. Still one issue to hunt down and fix.
Okay brad@, who reminded me about updating both ltmain.in and ltmain.sh...

Log message:
rev 1.334.2.92
With `-static', only link statically against uninstalled libtool
libraries.  Fixes 1.5.x regression to match documented (and actual
1.4.x) behavior.
>From libtool CVS

Log message:
upgrade to libtool 1.5.20

Log message:
rev 1.334.2.78
When a library is installed, dependent on a convenience library, and it
involves relinking, the object files extracted from the convenience
library are not removed after the relink. This is a problem if you build
as non-root, install as root, then try to remove the build directory as
non-root; Clean up properly if relink fails; Change"$realname"U to
${realname}T to be consistent.
Missing backport of 2004-03-24 patch reported against MirLibtool
by Marc Matteo <marcm@lectroid.net>.

Log message:
rev 1.334.2.72
Add fully-qualified paths to temp_rpath rather than unqualified paths in order
to avoid possible errors when computing the fully-qualified path later.
rev 1.334.2.70
Do not add installed static litool libraries to convenience, they are not
convenience libraries.
>From libtool CVS

Log message:
upgrade to libtool 1.5.18

Log message:
upgrade to libtool 1.5.16

Log message:
upgrade to libtool 1.5.14

Log message:
Now that @info transparently picks up all .info-* files, don't go through
extra steps to rebuild with --no-split.

Log message:
SIZE

Log message:
rev 1.28
Propagate a convenience library's dependency libs correctly when
it is being linked into a libtool library.  Deplibs are now
propagated whether libdir is set or not.
rev 1.27
The dlpreopen pass over libraries reverses the elements in the
$deplibs list.  This causes problems when the link pass tries to
find libraries when they are located in non-standard places
denoted by -L options.  Due to the reversed order these -L options
occur after the libraries that need them, and they are not found:
(Un)Reverse $deplibs list at the start of the link pass in lib mode.
>From libtool CVS

Log message:
When linking an installed libtool library on the command line
using -lfoo we need to find the library in the search paths
and add it's dependency_libs to the link in the conv pass so
that any libtool libs listed are correctly expanded in the
link pass.
>From libtool CVS

Log message:
upgrade to libtool 1.5.10

Log message:
switch to new plist, killing a few install scripts along the way.

Log message:
upgrade to libtool 1.5.8

Log message:
bump PKGNAME

Log message:
I'm not sure if this is the end of all the relink issues
but this is definitely an improvement on the workaround
currently in place at the moment.
From: Jacob Meuser <jakemsr at jakemsr dot com>

Log message:
set hardcode_direct and hardcode_shlibpath_var in OpenBSD' C++ section too.

Log message:
Use sysctl to get the max cmd length on *BSD
From: libtool CVS

Log message:
ltmain.in: Fix problem with .libs/.libs/libfoo.so appearing on the link line
From: libtool CVS

Log message:
until I can figure out this total gayness with libtool relinking
improperly then I'll just put the usual workaround in place.

Log message:
upgrade to libtool 1.5.6, libltdl is now a sub-package.

Log message:
remove WWW lines

Log message:
and more..