yep
--
Sent from a phone, apologies for poor formatting.
On 14 January 2026 15:12:39 Stuart Cassoff <exokoide@yahoo.ca> wrote:
Like this?Remove incorrect and unneededUPDATE_PLIST_ARGS= -I MODPY_WANTLIBthat was breaking update-plist.diff -Nurp /usr/ports/devel/subversion/Makefile ./Makefile--- /usr/ports/devel/subversion/Makefile Tue Jul 22 08:45:05 2025+++ ./Makefile Wed Jan 14 10:06:24 2026@@ -184,7 +184,6 @@ CONFIGURE_ARGS+=--without-gnome-keyringTEST_DEPENDS=${MODPY_RUN_DEPENDS}TEST_DEPENDS+= ${FULLPKGNAME-gnome-keyring}:${BUILD_PKGPATH},-gnome-keyringSUBST_VARS+= LOCALBASE SO_VERSION MODRUBY_BINREV-UPDATE_PLIST_ARGS= -I MODPY_WANTLIBpre-patch:@ln -s ${MODPY_BIN} ${WRKDIR}/bin/pythonReturn Ruby subpackage to the unflavored build.Removed in r1.198 due to build failure with clang 15.Ok now with clang 19.diff -Nurp /usr/ports/devel/subversion/Makefile ./Makefile--- /usr/ports/devel/subversion/Makefile Wed Jan 14 10:06:24 2026+++ ./Makefile Wed Jan 14 10:08:17 2026@@ -43,7 +43,7 @@ EXTRACT_SUFX= .tar.bz2PSEUDO_FLAVORS= no_bindings no_ap2 no_perl no_python no_ruby no_gnome-keyringFLAVORS= maintainer_mode-FLAVOR?= no_ruby+FLAVOR?=.if ${FLAVOR:Mno_bindings}FLAVOR += no_perl no_python no_ruby.endif@@ -55,7 +55,7 @@ WANTLIB= expat iconv intl lz4 m pthread zBUILD_DEPENDS= devel/gettext,-tools \devel/py3c-MULTI_PACKAGES = -main -ap2 -perl -python -gnome-keyring+MULTI_PACKAGES = -main -ap2 -perl -python -ruby -gnome-keyringMAINSPEC = ${PKGNAME}:devel/subversion,-mainLIB_DEPENDS-main = ${LIB_DEPENDS} \On Wed, 14 Jan 2026, Stuart Henderson wrote:On 2026/01/14 09:15, Stuart Cassoff wrote:Ruby bindings are a bit stale (lots of deprecation warnings)but otherwise build and test ok on i386 and amd64.If the port were left as-is, would -ruby need tobe conditionally added to MULTI_PACKAGES?StuReturn Ruby subpackage to the unflavored build.Removed in r1.198 due to build failure with clang 15.Ok now with clang 19.Remove incorrect and unneededUPDATE_PLIST_ARGS= -I MODPY_WANTLIBthat was breaking update-plist.if readding them:diff -Nurp /usr/ports/devel/subversion/Makefile ./Makefile--- /usr/ports/devel/subversion/Makefile Tue Jul 22 08:45:05 2025+++ ./Makefile Tue Jan 13 08:35:45 2026@@ -43,7 +43,7 @@ EXTRACT_SUFX= .tar.bz2PSEUDO_FLAVORS= no_bindings no_ap2 no_perl no_python no_ruby no_gnome-keyringFLAVORS= maintainer_mode-FLAVOR?= no_ruby+FLAVOR?=.if ${FLAVOR:Mno_bindings}FLAVOR += no_perl no_python no_ruby.endif@@ -55,7 +55,7 @@ WANTLIB= expat iconv intl lz4 m pthread zBUILD_DEPENDS= devel/gettext,-tools \devel/py3c-MULTI_PACKAGES = -main -ap2 -perl -python -gnome-keyring+MULTI_PACKAGES = -main -ap2 -perl -python -ruby -gnome-keyringthat would be one commitMAINSPEC = ${PKGNAME}:devel/subversion,-mainLIB_DEPENDS-main = ${LIB_DEPENDS} \@@ -184,7 +184,6 @@ CONFIGURE_ARGS+=--without-gnome-keyringTEST_DEPENDS=${MODPY_RUN_DEPENDS}TEST_DEPENDS+= ${FULLPKGNAME-gnome-keyring}:${BUILD_PKGPATH},-gnome-keyringSUBST_VARS+= LOCALBASE SO_VERSION MODRUBY_BINREV-UPDATE_PLIST_ARGS= -I MODPY_WANTLIBand that is unrelated to ruby so would be a separate commitpre-patch:@ln -s ${MODPY_BIN} ${WRKDIR}/bin/pythonOn Sat, 10 Jan 2026, Stuart Henderson wrote:that should have been -i not -I, but it's not doing anything useful anyway, so just drop it.update-plist(1)--ᅵᅵ Sent from a phone, apologies for poor formatting.On 10 January 2026 06:15:01 Stuart Cassoff <exokoide@yahoo.ca> wrote:Removing UPDATE_PLIST_ARGS unbreaks update-plist."UPDATE_PLIST_ARGS= -I MODPY_WANTLIB"I can't see why this is causing breakage.Ultimately unrelated to ruby.On Wed, 7 Jan 2026, Stuart Henderson wrote:On 2026/01/07 09:30, Stuart Cassoff wrote:$ make update-plist===> ᅵUpdating plist for subversion-1.14.5Installing /usr/ports/mystuff/devel/subversion/pkg/svnserve.rc as /usr/perts/pobj/subversion-1.14.5/fake-amd64/etc/rc.d/svnserveUsage: update-plist [-FfmnOrvx] [-C dir] [-c comment] [-E ext] [-e ext] [-H hints]ᅵ[-i var] [-I ignored] [-j jobs] [-L logfile] [-P pkgdir]ᅵ[-S var] [-s var] [-V var] [-w suffix] [-X path]ᅵ-- pkg_create_args ...I think it's due to how the ruby bindings were knocked out (afterfailing to build with newer clang).
No comments:
Post a Comment