Re: UPDATE borgbackup-1.1.15
On Sat 26/12/2020 06:44, Bjorn Ketelaars wrote:
> A newer version of borgbackup is available - 1.1.15 -, which is a
> maintenance release. List of changes can be found at
> https://github.com/borgbackup/borg/blob/1.1.15/docs/changes.rst.
>
> Changes to the port:
> - Switch to xxhash from ports instead of using the bundled version
> - Install shell completions
> - No need to have ${MODPY_LIB_DEPENDS} in LIB_DEPENDS as borgbackup does
> not link against libpython. Same for ${MODPY_WANTLIB} in WANTLIB
>
> Testing:
> - Run tested on amd64: creating a snapshot and extracting some file from
> it
> - 'make test' runs successful
>
> Comments/OK?
Ping!
Diff enclosed for your convenience.
Index: Makefile
===
RCS file: /cvs/ports/sysutils/borgbackup/Makefile,v
retrieving revision 1.36
diff -u -p -r1.36 Makefile
--- Makefile7 Oct 2020 18:28:08 - 1.36
+++ Makefile25 Dec 2020 18:37:58 -
@@ -2,7 +2,7 @@
COMMENT = deduplicating backup program
-MODPY_EGG_VERSION =1.1.14
+MODPY_EGG_VERSION =1.1.15
DISTNAME = borgbackup-${MODPY_EGG_VERSION}
CATEGORIES = sysutils
@@ -21,14 +21,14 @@ MODPY_PYTEST = Yes
MODPY_PYTEST_ARGS =--ignore lib* --ignore src/borg/testsuite/benchmark.py
MODPY_VERSION =${MODPY_DEFAULT_VERSION_3}
-WANTLIB += ${COMPILER_LIBCXX} ${MODPY_WANTLIB} b2 crypto lz4 m zstd
+WANTLIB += ${COMPILER_LIBCXX} b2 crypto lz4 m xxhash zstd
COMPILER = base-clang ports-gcc base-gcc
-LIB_DEPENDS += ${MODPY_LIB_DEPENDS} \
- archivers/lz4 \
+LIB_DEPENDS += archivers/lz4 \
archivers/zstd \
- security/libb2
+ security/libb2 \
+ sysutils/xxhash
BUILD_DEPENDS += devel/py-setuptools_scm${MODPY_FLAVOR}
# XXX pytest-benchmark is currently not picked up as plugin by pytest, as
result
@@ -42,6 +42,18 @@ PORTHOME = ${WRKDIR}
post-install:
${INSTALL_MAN} ${WRKSRC}/docs/man/*.1 ${PREFIX}/man/man1/
+
+ ${INSTALL_DATA_DIR} ${PREFIX}/share/bash-completion/completions/
+ ${INSTALL_DATA} ${WRKSRC}/scripts/shell_completions/bash/borg \
+ ${PREFIX}/share/bash-completion/completions/
+
+ ${INSTALL_DATA_DIR} ${PREFIX}/share/fish/completions/
+ ${INSTALL_DATA} ${WRKSRC}/scripts/shell_completions/fish/borg.fish \
+ ${PREFIX}/share/fish/completions/
+
+ ${INSTALL_DATA_DIR} ${PREFIX}/share/zsh/site-functions/
+ ${INSTALL_DATA} ${WRKSRC}/scripts/shell_completions/zsh/_borg \
+ ${PREFIX}/share/zsh/site-functions/
pre-test:
${MODPY_CMD} build_ext --inplace
Index: distinfo
===
RCS file: /cvs/ports/sysutils/borgbackup/distinfo,v
retrieving revision 1.22
diff -u -p -r1.22 distinfo
--- distinfo7 Oct 2020 18:28:08 - 1.22
+++ distinfo25 Dec 2020 18:37:58 -
@@ -1,2 +1,2 @@
-SHA256 (borgbackup-1.1.14.tar.gz) =
fbsHR8yUhnP2lc1t4oSvIV+BD+0usqYV7ybdx8aR7bo=
-SIZE (borgbackup-1.1.14.tar.gz) = 3769460
+SHA256 (borgbackup-1.1.15.tar.gz) =
Scue7Zi44yrjuXvq7flM3/RveWRFBD8ZI/0Pzn7Twrw=
+SIZE (borgbackup-1.1.15.tar.gz) = 3777355
Index: pkg/PLIST
===
RCS file: /cvs/ports/sysutils/borgbackup/pkg/PLIST,v
retrieving revision 1.11
diff -u -p -r1.11 PLIST
--- pkg/PLIST 8 Mar 2020 14:52:29 - 1.11
+++ pkg/PLIST 25 Dec 2020 18:37:58 -
@@ -185,3 +185,12 @@ lib/python${MODPY_VERSION}/site-packages
@man man/man1/borg-with-lock.1
@man man/man1/borg.1
@man man/man1/borgfs.1
+share/bash-completion/
+share/bash-completion/completions/
+share/bash-completion/completions/borg
+share/fish/
+share/fish/completions/
+share/fish/completions/borg.fish
+share/zsh/
+share/zsh/site-functions/
+share/zsh/site-functions/_borg
[UPDATE] games/cataclysm-dda-0.E-2 -> 0.E-3
Small bugfix update to Cataclysm. Tim. Index: Makefile === RCS file: /cvs/ports/games/cataclysm-dda/Makefile,v retrieving revision 1.13 diff -u -p -r1.13 Makefile --- Makefile11 Jul 2020 17:52:19 - 1.13 +++ Makefile2 Jan 2021 16:55:10 - @@ -1,12 +1,12 @@ # $OpenBSD: Makefile,v 1.13 2020/07/11 17:52:19 kirby Exp $ COMMENT= rogue-like zombie survival game -DISTNAME= cataclysm-dda-0.E.2 +DISTNAME= cataclysm-dda-0.E.3 CATEGORIES=games GH_ACCOUNT=CleverRaven GH_PROJECT=Cataclysm-DDA -GH_TAGNAME=0.E-2 +GH_TAGNAME=0.E-3 EPOCH= 0 HOMEPAGE= https://cataclysmdda.org Index: distinfo === RCS file: /cvs/ports/games/cataclysm-dda/distinfo,v retrieving revision 1.10 diff -u -p -r1.10 distinfo --- distinfo11 Jul 2020 17:52:19 - 1.10 +++ distinfo2 Jan 2021 16:55:10 - @@ -1,2 +1,2 @@ -SHA256 (cataclysm-dda-0.E.2.tar.gz) = QVRuh34u7nnISSs+yAjvU6S5sgjXiNrCsaVw7xQ0Juk= -SIZE (cataclysm-dda-0.E.2.tar.gz) = 70144350 +SHA256 (cataclysm-dda-0.E.3.tar.gz) = IaxSJqmWrEZYQvGIyt6ogV6ucwn+OM+NlN4visl82CA= +SIZE (cataclysm-dda-0.E.3.tar.gz) = 70144192
Re: Using chirp
Hi Chris -- On Saturday, January 2, 2021 5:47 PM, Chris Bennett wrote: > I am going to buy a Ham radio, first a handheld, later something fixed. > > Are there any problems/hints that I need to knnow. > Any problems with a USB to serial cable? > Any problems with the radios listed on the chirp website? > I need to get one that is waterproof. > I will be getting my license up to the top USA level. > > Thanks, > Chris Bennett I have an Amateur Extra license. Chirp works just fine. The key is to pay the extra few dollars for a serial cable with an FTDI chip in it. This is a shortcoming of the radio hardware, not Chirp or OpenBSD. ~Brian
tinc: update to 1.1pre17 (second attempt)
https://marc.info/?l=openbsd-ports&m=156916051316517&w=2 was my first
attempt over a year ago to update.
naddy and sthen had provided feedback which made it to the final diff;
thread/diff and my interest stalled after that, now I have a use case
for tinc again, so hopefully we can get it in this time.
This is the same diff I sent last time except for TEST_IS_INTERACTIVE
now being set along with a tiny patch to unbreak building tests.
OK?
Index: Makefile
===
RCS file: /cvs/ports/net/tinc/Makefile,v
retrieving revision 1.14
diff -u -p -r1.14 Makefile
--- Makefile4 Apr 2020 01:18:43 - 1.14
+++ Makefile3 Jan 2021 02:19:33 -
@@ -1,28 +1,30 @@
# $OpenBSD: Makefile,v 1.14 2020/04/04 01:18:43 kn Exp $
COMMENT = Virtual Private Network (VPN) daemon
-DISTNAME = tinc-1.0.36
+DISTNAME = tinc-1.1pre17
CATEGORIES = net security
EPOCH =0
-REVISION = 0
HOMEPAGE = https://www.tinc-vpn.org/
# GPLv2+
PERMIT_PACKAGE = Yes
-WANTLIB += c crypto lzo2 z
+WANTLIB += c crypto curses lzo2 m readline z
MASTER_SITES = https://www.tinc-vpn.org/packages/
LIB_DEPENDS += archivers/lzo2
+SEPARATE_BUILD = Yes
CONFIGURE_STYLE = gnu
-
CONFIGURE_ENV =CPPFLAGS="-I${LOCALBASE}/include" \
- LDFLAGS="-L${LOCALBASE}/lib"
+ LDFLAGS="-L${LOCALBASE}/lib" \
+ tinc_cv_attribute___nonnull__=yes
USE_GMAKE =Yes
+
+TEST_IS_INTERACTIVE = XXX spawns tincd processes
post-install:
${INSTALL_DATA_DIR} ${PREFIX}/share/examples/tinc
Index: distinfo
===
RCS file: /cvs/ports/net/tinc/distinfo,v
retrieving revision 1.10
diff -u -p -r1.10 distinfo
--- distinfo2 Sep 2019 20:55:08 - 1.10
+++ distinfo3 Jan 2021 02:03:03 -
@@ -1,2 +1,2 @@
-SHA256 (tinc-1.0.36.tar.gz) = QPc7s/rMSA7/4OdxRCpwb/BIjt6npfJQXUzLKqgWMQg=
-SIZE (tinc-1.0.36.tar.gz) = 500320
+SHA256 (tinc-1.1pre17.tar.gz) = YbnJ+fOWdoVR85IW7cxBkYxlkJ/9mvBx/rO1+fmsHCc=
+SIZE (tinc-1.1pre17.tar.gz) = 927313
Index: patches/patch-test_splice_c
===
RCS file: patches/patch-test_splice_c
diff -N patches/patch-test_splice_c
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-test_splice_c 3 Jan 2021 02:13:55 -
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+Include to fix
+ error: use of undeclared identifier 'fd_set'
+
+Index: test/splice.c
+--- test/splice.c.orig
test/splice.c
+@@ -23,6 +23,7 @@
+ #include
+ #include
+ #include
++#include
+ #include
+
+ #ifdef HAVE_MINGW
Index: pkg/PLIST
===
RCS file: /cvs/ports/net/tinc/pkg/PLIST,v
retrieving revision 1.4
diff -u -p -r1.4 PLIST
--- pkg/PLIST 9 Oct 2018 20:32:45 - 1.4
+++ pkg/PLIST 3 Jan 2021 02:03:04 -
@@ -4,7 +4,10 @@
@rcscript ${RCDIR}/tincd
@info info/tinc.info
@man man/man5/tinc.conf.5
+@comment @man man/man8/tinc-gui.8
+@man man/man8/tinc.8
@man man/man8/tincd.8
+@bin sbin/tinc
@bin sbin/tincd
share/examples/tinc/
share/examples/tinc/hosts/
UPDATE: devel/git-lfs to 2.13.1
Hi, The attached diff updates Git LFS to 2.13.1. I have tested it on amd64. It works. Please update. Thanks, Asher Index: Makefile === RCS file: /cvs/ports/devel/git-lfs/Makefile,v retrieving revision 1.17 diff -u -p -u -p -r1.17 Makefile --- Makefile 5 Nov 2020 15:56:39 - 1.17 +++ Makefile 2 Jan 2021 20:55:42 - @@ -4,7 +4,7 @@ COMMENT = Git extension for versioning GH_PROJECT = git-lfs GH_ACCOUNT = git-lfs -GH_TAGNAME = v2.12.1 +GH_TAGNAME = v2.13.1 CATEGORIES = devel Index: distinfo === RCS file: /cvs/ports/devel/git-lfs/distinfo,v retrieving revision 1.15 diff -u -p -u -p -r1.15 distinfo --- distinfo 5 Nov 2020 15:56:39 - 1.15 +++ distinfo 2 Jan 2021 20:55:42 - @@ -1,2 +1,2 @@ -SHA256 (git-lfs-2.12.1.tar.gz) = Ky5w8SM/fv6aAQdxUQORoHUn7HwK9yHs+O2rrF1g9is= -SIZE (git-lfs-2.12.1.tar.gz) = 2848245 +SHA256 (git-lfs-2.13.1.tar.gz) = W6fZRdlq1JSS4p7b/RzOUosqA0/d2/blQk51SkoBICk= +SIZE (git-lfs-2.13.1.tar.gz) = 2526962
NEW: games/inform6/{compiler,library}
Hello ports -- Attached is a new pair of ports: games/inform6/compiler and games/inform6/library. Combined, these two ports provide the dev tools for compiling interactive fiction games for the Z-Machine interpreters. Since they are so tightly coupled, I figured it made sense to put them under one directory. --- pkg/DESCR for games/inform6/compiler: Inform 6 is an object-oriented language for writing interactive fiction, also known as "text adventures" or "adventure games". As an interactive fiction developer, you create Inform 6 source code defining the story. Then compile from source code to a "story file". The resulting story files are highly portable, with interpreters available for many different platforms. This package contains the Inform 6 compiler. The Inform 6 compiler can produce a story file in these formats: * Z-code, executed by implementations of the Infocom Z-machine. * Glulx, a modern platform for interactive fiction. --- I took the DESCR for both from the Debian package description. You can find some sample games here: https://inform-fiction.org/examples/index.html You can compile them by running: inform +/usr/local/share/inform Game.inf Which you can then run using games/frotz. OK? ~Brian inform6.tgz Description: application/compressed-tar
Re: Ping - [New] devel/flatbuffers
Jeremie Courreges-Anglas writes: > On Mon, Dec 07 2020, Ashton Fagg wrote: >> Ping again. > > Just like devel/fmt, looking for another ok to import this. > Tarball attached. OK gnezdo@ > >> Ashton Fagg writes: >> >>> >>> Hello folks, >>> >>> Just wondering if anyone is still interesting in importing this port? >>> Reviewer-agreed version attached from the previous discussions. >>> >>> Thanks, >>> Ash >>
UPDATE: games/frotz 2.44 => 2.52
Hi ports --
Attached is an update to games/frotz.
The complete changelog is here:
https://gitlab.com/DavidGriffith/frotz/-/releases
The two big items for users is UTF-8 and audio support.
Update HOMEPAGE/MASTER_SITES while here; it moved to GitLab.
Also enable the SDL2 client. I don't think it hurts to have it.
Slightly tweak COMMENT: I'm not sure it's fair to call it
"curses-based" when only 1 of the 3 clients actually uses ncurses.
While here, take MAINTAINER.
Discovered when porting the Inform compiler, to be sent separately.
OK?
~Brian
Index: Makefile
===
RCS file: /cvs/ports/games/frotz/Makefile,v
retrieving revision 1.20
diff -u -p -r1.20 Makefile
--- Makefile 12 Jul 2019 20:46:18 - 1.20
+++ Makefile 2 Jan 2021 20:30:41 -
@@ -1,28 +1,40 @@
# $OpenBSD: Makefile,v 1.20 2019/07/12 20:46:18 sthen Exp $
-COMMENT= curses-based interpreter for Infocom-compatible games
+V = 2.52
+COMMENT= interpreter for Infocom-compatible games
-DISTNAME= frotz-2.44
-REVISION= 0
+DISTNAME= frotz-${V}
CATEGORIES= games
-HOMEPAGE= http://frotz.sourceforge.net/
+HOMEPAGE= https://davidgriffith.gitlab.io/frotz/
+MAINTAINER= Brian Callahan
# GPLv2+
PERMIT_PACKAGE= Yes
-WANTLIB= c curses
-
-MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=frotz/}
+WANTLIB += SDL2 SDL2_mixer ao c curses execinfo freetype jpeg
+WANTLIB += m modplug png pthread samplerate sndfile vorbisfile
+WANTLIB += z
+
+MASTER_SITES= https://gitlab.com/DavidGriffith/frotz/-/archive/${V}/
+
+LIB_DEPENDS= audio/libao \
+ audio/libmodplug \
+ audio/libsamplerate \
+ devel/libexecinfo \
+ devel/sdl2-mixer \
+ graphics/jpeg \
+ graphics/png
USE_GMAKE= Yes
MAKE_FLAGS= CC="${CC}" OPTS="${CFLAGS}" CONFIG_DIR="${SYSCONFDIR}"
FAKE_FLAGS= PREFIX=${WRKINST}${TRUEPREFIX}
-ALL_TARGET= frotz dfrotz
-INSTALL_TARGET= install install_dumb
+INSTALL_TARGET= install install_dumb install_sdl
CFLAGS += -DUSE_UNISTD_H -DCOLOR_SUPPORT -DEMACS_EDITING
+
+FAKE_FLAGS = DESTDIR= MAN_PREFIX="${DESTDIR}${PREFIX}"
NO_TEST= Yes
Index: distinfo
===
RCS file: /cvs/ports/games/frotz/distinfo,v
retrieving revision 1.6
diff -u -p -r1.6 distinfo
--- distinfo 4 Jun 2015 06:17:38 - 1.6
+++ distinfo 2 Jan 2021 20:30:41 -
@@ -1,2 +1,2 @@
-SHA256 (frotz-2.44.tar.gz) = c+wg8DTS6RV4w4fP868knVK4gz5CcesQU/bTNXIv4+w=
-SIZE (frotz-2.44.tar.gz) = 275534
+SHA256 (frotz-2.52.tar.gz) = Prmvuecm84ijNz/s7ZuCB3ufUpxtkEE+N79/cg6404U=
+SIZE (frotz-2.52.tar.gz) = 353485
Index: patches/patch-Makefile
===
RCS file: patches/patch-Makefile
diff -N patches/patch-Makefile
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-Makefile 2 Jan 2021 20:30:41 -
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+Remove hardcoded optimizations
+
+Index: Makefile
+--- Makefile.orig
Makefile
+@@ -6,7 +6,7 @@
+ #CC ?= clang
+
+ # Enable compiler warnings. This is an absolute minimum.
+-CFLAGS += -Wall -std=c99 -O3 #-Wextra
++CFLAGS += -Wall -std=c99 #-Wextra
+
+ # Define your optimization flags.
+ #
Index: pkg/PLIST
===
RCS file: /cvs/ports/games/frotz/pkg/PLIST,v
retrieving revision 1.5
diff -u -p -r1.5 PLIST
--- pkg/PLIST 18 Oct 2016 19:19:29 - 1.5
+++ pkg/PLIST 2 Jan 2021 20:30:41 -
@@ -1,5 +1,7 @@
@comment $OpenBSD: PLIST,v 1.5 2016/10/18 19:19:29 sthen Exp $
@bin bin/dfrotz
@bin bin/frotz
+@bin bin/sfrotz
@man man/man6/dfrotz.6
@man man/man6/frotz.6
+@man man/man6/sfrotz.6
Using chirp
I am going to buy a Ham radio, first a handheld, later something fixed. Are there any problems/hints that I need to knnow. Any problems with a USB to serial cable? Any problems with the radios listed on the chirp website? I need to get one that is waterproof. I will be getting my license up to the top USA level. Thanks, Chris Bennett
Re: [New] devel/fmt
On Sat, Jan 02 2021, Rafael Sadowski wrote: > On Sat Jan 02, 2021 at 03:44:22PM +0100, Jeremie Courreges-Anglas wrote: >> On Tue, Dec 15 2020, Ashton Fagg wrote: >> > Ping. Updated version attached. >> >> BTW upstream provides proper releases upstream, see the zip files at >> https://github.com/fmtlib/fmt/releases >> Usually I would suggest using said tarballs, but here it doesn't seem to >> bring a significant improvement (generating the documentation would >> still require extra bdeps). Such possible improvements can be done >> in-tree. >> >> We need another ok to import this. Same tarball attached for convenience. > > OK rsadowski@ Thanks Rafael & Marc. :) While we're here: using codesearch.debian.net I looked for possible ports that might pick this up, and I only found mkvtoolnix (newer osm2pgsql could, but our version doesn't use fmt). So Rafael, regarding mkvtoolnix, either we use the bundled fmt with CONFIGURE_ENV= ac_cv_fmt=no or we reuse the system library as in the diff below. While here, I noticed that cmark is only used when building qt. ok? Index: Makefile === RCS file: /cvs/ports/multimedia/mkvtoolnix/Makefile,v retrieving revision 1.96 diff -u -p -r1.96 Makefile --- Makefile2 Dec 2020 06:44:11 - 1.96 +++ Makefile2 Jan 2021 16:14:15 - @@ -3,6 +3,7 @@ COMMENT= create, alter and inspect Matroska files DISTNAME= mkvtoolnix-51.0.0 +REVISION= 0 CATEGORIES=multimedia x11 @@ -14,7 +15,7 @@ MAINTAINER = Rafael Sadowski https://www.bunkus.org/videotools/mkvtoolnix/sources/ @@ -38,12 +39,12 @@ LIB_DEPENDS=audio/flac \ audio/libogg \ audio/libvorbis \ devel/boost \ + devel/fmt \ devel/gettext,-runtime \ devel/libdvdread \ devel/libmagic \ devel/pcre2 \ multimedia/libmatroska>=1.6.2 \ - textproc/cmark \ textproc/libebml>=1.4.0 \ textproc/pugixml @@ -71,8 +72,9 @@ CONFIGURE_ARGS+= --disable-gui .else MODULES+= x11/qt5 WANTLIB += Qt5Concurrent Qt5Core Qt5DBus Qt5Gui Qt5Multimedia Qt5Network -WANTLIB += Qt5Widgets -LIB_DEPENDS += x11/qt5/qtmultimedia +WANTLIB += Qt5Widgets cmark +LIB_DEPENDS += x11/qt5/qtmultimedia \ + textproc/cmark RUN_DEPENDS+= devel/desktop-file-utils \ misc/shared-mime-info \ x11/gtk+3,-guic -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
Re: lang/ruby: mention ruby-shims in MESSAGE
On Sat, Jan 02, 2021 at 08:59:31AM -0600, joshua stein wrote: > Installing ruby-shims gives a more user-friendly default experience > since it will automatically make /usr/local/bin/ruby and friends > work with whichever ruby package is installed, without having to use > symlinks. OK kn wit REVISION bumped as this changes the packages.
Re: lang/ruby: mention ruby-shims in MESSAGE
On Sat, Jan 2, 2021 at 7:00 AM joshua stein wrote: > Installing ruby-shims gives a more user-friendly default experience > since it will automatically make /usr/local/bin/ruby and friends > work with whichever ruby package is installed, without having to use > symlinks. > I've not used ruby-shims, so I don't feel comfortable providing an OK for this. I don't have objections to it, so if other developers have used ruby-shims and think this is worth mentioning in MESSAGE, it can be committed with their OK. Thanks, Jeremy
Re: [New] devel/fmt
On Sat, Jan 02, 2021 at 03:44:22PM +0100, Jeremie Courreges-Anglas wrote: > On Tue, Dec 15 2020, Ashton Fagg wrote: > > Ping. Updated version attached. > > BTW upstream provides proper releases upstream, see the zip files at > https://github.com/fmtlib/fmt/releases > Usually I would suggest using said tarballs, but here it doesn't seem to > bring a significant improvement (generating the documentation would > still require extra bdeps). Such possible improvements can be done > in-tree. > > We need another ok to import this. Same tarball attached for convenience. Okay
Re: www/castor: avoid DIST_SUBDIR (and put cargo crates with others)
On Sat, Jan 02, 2021 at 03:04:56PM +0100, Sebastien Marie wrote:
> Hi,
>
> While working on latest lang/rust update, I saw that www/castor is
> using a special layout for distfiles: all crates are put in castor/
> directory, with castor tarball named "0.8.16.tar.gz".
>
> Regarding cargo crates, having the different directories that others
> crates means all commun files are duplicated.
>
> The following diff should put www/castor in line with others cargo
> using ports:
>
> - fetch upstream tarball with "0.8.16.tar.gz" name and save it as
> "castor-0.8.16.tar.gz"
> - use default option for cargo module to put all downloaded crates to
> "cargo/" directory
>
> It makes the 25 specific crates be saved in "cargo/", and avoid
> downloading the 75 others crates already present in "cargo/".
>
> DISTFILES variable follows the magic as shown in bsd.port.mk man page
> for renaming upstream tarball.
>
> Comments or OK ?
I never understood how that thing worked :-)
ok for me
> --
> Sebastien Marie
>
> diff 5bb352a5b727fa07be529004f9bc16847be7cd0f
> /data/semarie/repos/openbsd/ports
> blob - 447e197d6b990febe326014e44f53283b4d68550
> file + www/castor/Makefile
> --- www/castor/Makefile
> +++ www/castor/Makefile
> @@ -6,9 +6,7 @@ V=0.8.16
> DISTNAME=castor-${V}
> REVISION=0
>
> -DIST_SUBDIR= castor
> -DISTFILES= ${V}${EXTRACT_SUFX}
> -MODCARGO_DIST_SUBDIR=# empty
> +DISTFILES= castor-{castor/archive/}${V}${EXTRACT_SUFX}
>
> CATEGORIES= www
>
> @@ -22,7 +20,7 @@ PERMIT_PACKAGE= Yes
> WANTLIB += c c++abi crypto gdk-3 glib-2.0 gobject-2.0 gtk-3 pthread
> WANTLIB += ssl
>
> -MASTER_SITES=https://git.sr.ht/~julienxx/castor/archive/
> +MASTER_SITES=https://git.sr.ht/~julienxx/
>
> MODULES= devel/cargo
>
> blob - 6902987be18db426453da8ec6bb43d1e390eef17
> file + www/castor/distinfo
> --- www/castor/distinfo
> +++ www/castor/distinfo
> @@ -1,202 +1,202 @@
> -SHA256 (castor/0.8.16.tar.gz) = UEFPSYw7YjLvsA6IMEOFRRcpC4QuBQHB0ezPr3i+muM=
> -SHA256 (castor/aho-corasick-0.7.13.tar.gz) =
> BDFk2LpcTDA1/sm77oZHwCYdeI80dDBvk7tlkByuDoY=
> -SHA256 (castor/ansi-parser-0.6.5.tar.gz) =
> dhrGdfFjimpJ4m9qw6QGfKP++oApgWrk740/o9BqUZQ=
> -SHA256 (castor/arrayref-0.3.6.tar.gz) =
> pMUnFS43z3V6P3iq5aBvvu/bB8zFNcmAoyCO4wYN1UQ=
> -SHA256 (castor/arrayvec-0.5.1.tar.gz) =
> z/d9hoaGfs7/MQUynUaY2WwjkcF21dA63JDHOJFitbg=
> -SHA256 (castor/atk-0.8.0.tar.gz) =
> RE2u+lXyKa8UXqWNd+/SNyUCTuH28xAnQ3CaprGMZj4=
> -SHA256 (castor/atk-sys-0.9.1.tar.gz) =
> 5VLBd2c3pMgBENBrNtCZ9HxyczX5qqXZQqcraGOo7G8=
> -SHA256 (castor/autocfg-1.0.0.tar.gz) =
> +KrHcPGIX9fjh6zt12BlMCVRNkSW5Gs90Ahgsvg1m50=
> -SHA256 (castor/base64-0.11.0.tar.gz) =
> tBt+pUoMnZIZneieIOWNSfAvjmmYFO8/3yZvb3SNFcc=
> -SHA256 (castor/bitflags-1.2.1.tar.gz) =
> zx3i/ox1vBRaL1d63ZUfgTSIm0eV1HRmpUpchG1pFpM=
> -SHA256 (castor/blake2b_simd-0.5.10.tar.gz) =
> 2PstdCVKOgtcrDOsn47Q5EqlA3jZ27Ll2DvSHtHcLIo=
> -SHA256 (castor/cairo-rs-0.8.1.tar.gz) =
> FXBJupYYqjphw51deFECwE07H0BjKnBsYhqa7cIeYIQ=
> -SHA256 (castor/cairo-sys-rs-0.9.2.tar.gz) =
> /2W6AsrHFb6Db2NCmrAKdn1IM278VJfFY3r7U7TxTWM=
> -SHA256 (castor/cc-1.0.55.tar.gz) =
> sb40CflNe9zutfX6xVEDnZs/AOJdp6dPxNM0AKDZY2g=
> -SHA256 (castor/cfg-if-0.1.10.tar.gz) =
> R4W90clrKoRrK9fMAuhraz2/FOflNEbE9UySo2EECCI=
> -SHA256 (castor/constant_time_eq-0.1.5.tar.gz) =
> JFCX6aRTXuHj45Mfz81Vp5akTGQ+hZb/ZWbWjwm4e7w=
> -SHA256 (castor/core-foundation-0.7.0.tar.gz) =
> V9JMehPEPocON8FVa3RVVDeHCgRRT3aF9bNU4JBWcXE=
> -SHA256 (castor/core-foundation-sys-0.7.0.tar.gz) =
> s6catJTAtbhgvchAeuCJeAUkFwcMLO04VzqRV611uKw=
> -SHA256 (castor/crossbeam-utils-0.7.2.tar.gz) =
> w8fHOi0en8CIagi5PpjrZDRhIw1fGSXkA2IE1fLiYag=
> -SHA256 (castor/dirs-2.0.2.tar.gz) =
> E66omlyTNkqY6bN7L6I37/u2lNXP4BxbcJQffrCH1eM=
> -SHA256 (castor/dirs-sys-0.3.5.tar.gz) =
> jpPX9XBd4+SYlaK14LiFWhwn8IAZKunDKmQy1QdBpXo=
> -SHA256 (castor/foreign-types-0.3.2.tar.gz) =
> 9vM564rcBSzSyniRD9qGmu+jjSLVy2SOZIXk0/wG87E=
> -SHA256 (castor/foreign-types-shared-0.1.1.tar.gz) =
> ALAihBGQjKhoXbp/ws3XDsmZCm51Pom2rJGoTED7r0s=
> -SHA256 (castor/futures-channel-0.3.5.tar.gz) =
> 82atdMKMymukVtleZCKIPPtLJSqDvtkpyDq/278pZ9U=
> -SHA256 (castor/futures-core-0.3.5.tar.gz) =
> WfX/+Q/V2XH5Nq1nSAJIK6RBtvCbpeFf2LORRVgso5k=
> -SHA256 (castor/futures-executor-0.3.5.tar.gz) =
> ENa7iIvhFT06vrkAaxGwLPXpsgn9ooaTwxrh5OAS4xQ=
> -SHA256 (castor/futures-io-0.3.5.tar.gz) =
> 3icUKwE6joacFJV+bS7e74npfCieadBC7jpJrNi1F4k=
> -SHA256 (castor/futures-macro-0.3.5.tar.gz) =
> 0LWjCkMoq1Rzh4I3xEczPAkyl73tg6SYPRD03uokDTk=
> -SHA256 (castor/futures-task-0.3.5.tar.gz) =
> vbZrXwniIBmxqwgw93hbzqjnpCFIaD+ZIU9z+OwhpiY=
> -SHA256 (castor/futures-util-0.3.5.tar.gz) =
> h2RXT/CLcBoIRILDxwMTSRBLB6yJc5MBBJS+qhjOMsY=
> -SHA256 (castor/gdk-0.12.1.tar.gz) =
> ++Xody/AhlxSRgzdelnX1HcA9E2YCdHdAO7M63aadYk=
> -SHA256 (castor/gdk-pixbuf-0.8.0.tar.gz) =
> 4kgiDEazKbCX1LFY0nF
Re: [New] devel/fmt
On Sat Jan 02, 2021 at 03:44:22PM +0100, Jeremie Courreges-Anglas wrote: > On Tue, Dec 15 2020, Ashton Fagg wrote: > > Ping. Updated version attached. > > BTW upstream provides proper releases upstream, see the zip files at > https://github.com/fmtlib/fmt/releases > Usually I would suggest using said tarballs, but here it doesn't seem to > bring a significant improvement (generating the documentation would > still require extra bdeps). Such possible improvements can be done > in-tree. > > We need another ok to import this. Same tarball attached for convenience. OK rsadowski@ > > > Ashton Fagg writes: > > > >> Jeremie Courreges-Anglas writes: > >> > >>> [...] > >>> > >>> I think the port as-is is ripe and ready to commit, but I don't know yet > >>> if there are ports that might pick this up. > > A newer version of openimageio requires this, as noted by Dimitri Karamazov. > > >>> [...] > >> > >> Just following up on this - if anyone is still interested in importing > >> this new port, I have updated it to the latest version - 7.1.3 (vs > >> 7.1.1). Tarball is attached. > >> > >> Tested the same way - builds, tests, fakes, packages, installs and > >> deinstalls. > > > -- > jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
Re: Ping - [New] devel/flatbuffers
On Mon, Dec 07 2020, Ashton Fagg wrote: > Ping again. Just like devel/fmt, looking for another ok to import this. Tarball attached. > Ashton Fagg writes: > >> >> Hello folks, >> >> Just wondering if anyone is still interesting in importing this port? >> Reviewer-agreed version attached from the previous discussions. >> >> Thanks, >> Ash > flatbuffers.tgz Description: Binary data -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
lang/ruby: mention ruby-shims in MESSAGE
Installing ruby-shims gives a more user-friendly default experience
since it will automatically make /usr/local/bin/ruby and friends
work with whichever ruby package is installed, without having to use
symlinks.
diff --git lang/ruby/2.6/pkg/MESSAGE-main lang/ruby/2.6/pkg/MESSAGE-main
index 46eef2f38aa..673902dad71 100644
--- lang/ruby/2.6/pkg/MESSAGE-main
+++ lang/ruby/2.6/pkg/MESSAGE-main
@@ -1,5 +1,6 @@
If you want to use this package as your default system ruby, as root
create symbolic links like so (overwriting any previous default):
+
ln -sf ${PREFIX}/bin/ruby26 ${PREFIX}/bin/ruby
ln -sf ${PREFIX}/bin/erb26 ${PREFIX}/bin/erb
ln -sf ${PREFIX}/bin/irb26 ${PREFIX}/bin/irb
@@ -9,3 +10,6 @@ create symbolic links like so (overwriting any previous
default):
ln -sf ${PREFIX}/bin/gem26 ${PREFIX}/bin/gem
ln -sf ${PREFIX}/bin/bundle26 ${PREFIX}/bin/bundle
ln -sf ${PREFIX}/bin/bundler26 ${PREFIX}/bin/bundler
+
+The ruby-shims package is also available to automatically select an
+appropriate Ruby version per-project directory or system-wide.
diff --git lang/ruby/2.7/pkg/MESSAGE-main lang/ruby/2.7/pkg/MESSAGE-main
index d7687ea6b65..42897ddacc5 100644
--- lang/ruby/2.7/pkg/MESSAGE-main
+++ lang/ruby/2.7/pkg/MESSAGE-main
@@ -1,5 +1,6 @@
If you want to use this package as your default system ruby, as root
create symbolic links like so (overwriting any previous default):
+
ln -sf ${PREFIX}/bin/ruby27 ${PREFIX}/bin/ruby
ln -sf ${PREFIX}/bin/erb27 ${PREFIX}/bin/erb
ln -sf ${PREFIX}/bin/irb27 ${PREFIX}/bin/irb
@@ -12,3 +13,6 @@ create symbolic links like so (overwriting any previous
default):
ln -sf ${PREFIX}/bin/racc27 ${PREFIX}/bin/racc
ln -sf ${PREFIX}/bin/racc2y27 ${PREFIX}/bin/racc2y
ln -sf ${PREFIX}/bin/y2racc27 ${PREFIX}/bin/y2racc
+
+The ruby-shims package is also available to automatically select an
+appropriate Ruby version per-project directory or system-wide.
diff --git lang/ruby/3.0/pkg/MESSAGE-main lang/ruby/3.0/pkg/MESSAGE-main
index 8ec5c9814ce..69a3d56627c 100644
--- lang/ruby/3.0/pkg/MESSAGE-main
+++ lang/ruby/3.0/pkg/MESSAGE-main
@@ -1,5 +1,6 @@
If you want to use this package as your default system ruby, as root
create symbolic links like so (overwriting any previous default):
+
ln -sf ${PREFIX}/bin/ruby30 ${PREFIX}/bin/ruby
ln -sf ${PREFIX}/bin/erb30 ${PREFIX}/bin/erb
ln -sf ${PREFIX}/bin/irb30 ${PREFIX}/bin/irb
@@ -12,3 +13,6 @@ create symbolic links like so (overwriting any previous
default):
ln -sf ${PREFIX}/bin/bundle30 ${PREFIX}/bin/bundle
ln -sf ${PREFIX}/bin/bundler30 ${PREFIX}/bin/bundler
ln -sf ${PREFIX}/bin/typeprof30 ${PREFIX}/bin/typeprof
+
+The ruby-shims package is also available to automatically select an
+appropriate Ruby version per-project directory or system-wide.
Re: New port: net/qtox and databases/sqlcipher and an update to net/toxcore
On Fri, Jan 01, 2021 at 10:25:36PM +, Stuart Henderson wrote: > On 2021/01/01 14:30, Tom Murphy wrote: > > Hi all, > > > > I've been wanting to get qtox running (since utox is rather > > unstable). > > > > Attached to this email are two new ports: > > > > databases/sqlcipher - Allows you to encrypt sqlite databases > > net/qtox - qtox client for the tox framework > > > > I'm also including a diff for net/toxcore to upgrade it from 0.2.3 to > > 0.2.12. > > Here's a diff I had hanging around in my tree to update > toxcore/toxic/utox, the API changed a bit and IIRC they all needed > updating together. Builds OK but I don't use any of this, since > you have a setup already could you give it a spin please? At least > toxcore needs SHARED_LIBS bumping that wasn't in your diff. > Diff for that is sent as a gzipped attachment because it was > unfortunately imported with MSDOS line endings in the patches > which usually gets mangled in email. > > > I'd like to propose that net/utox might be removed from the tree. > > It's alpha and pretty unstable (sometimes it makes the cpu fan race and > > I've had to kill -9 it sometimes) > > 0.18.0 changelog for utox suggests it's likely to be better: > https://github.com/uTox/uTox/releases/tag/v0.18.0 > > > Note: net/qtox relies on databases/sqlcipher to compile and run. > > It also relies on the newer net/toxcore. I'm not sure about net/toxic > > client as I don't use it. OK I just ran patch from my home dir and manually put in the paths to the files and it applied. net/toxcore patch to 0.2.12 - works and installs here. net/utox patch to 0.18.0 - works and installs and seems more stable net/toxic patch to 0.10.0 - Looks like PLIST is broken. Was your diff missing an updated PLIST? Running make update-plist produced a package with no files in it. Thanks, Tom
Re: [New] devel/fmt
On Tue, Dec 15 2020, Ashton Fagg wrote: > Ping. Updated version attached. BTW upstream provides proper releases upstream, see the zip files at https://github.com/fmtlib/fmt/releases Usually I would suggest using said tarballs, but here it doesn't seem to bring a significant improvement (generating the documentation would still require extra bdeps). Such possible improvements can be done in-tree. We need another ok to import this. Same tarball attached for convenience. > Ashton Fagg writes: > >> Jeremie Courreges-Anglas writes: >> >>> [...] >>> >>> I think the port as-is is ripe and ready to commit, but I don't know yet >>> if there are ports that might pick this up. A newer version of openimageio requires this, as noted by Dimitri Karamazov. >>> [...] >> >> Just following up on this - if anyone is still interested in importing >> this new port, I have updated it to the latest version - 7.1.3 (vs >> 7.1.1). Tarball is attached. >> >> Tested the same way - builds, tests, fakes, packages, installs and >> deinstalls. fmt-7.1.3.tgz Description: Binary data -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
www/castor: avoid DIST_SUBDIR (and put cargo crates with others)
Hi,
While working on latest lang/rust update, I saw that www/castor is
using a special layout for distfiles: all crates are put in castor/
directory, with castor tarball named "0.8.16.tar.gz".
Regarding cargo crates, having the different directories that others
crates means all commun files are duplicated.
The following diff should put www/castor in line with others cargo
using ports:
- fetch upstream tarball with "0.8.16.tar.gz" name and save it as
"castor-0.8.16.tar.gz"
- use default option for cargo module to put all downloaded crates to "cargo/"
directory
It makes the 25 specific crates be saved in "cargo/", and avoid
downloading the 75 others crates already present in "cargo/".
DISTFILES variable follows the magic as shown in bsd.port.mk man page
for renaming upstream tarball.
Comments or OK ?
--
Sebastien Marie
diff 5bb352a5b727fa07be529004f9bc16847be7cd0f /data/semarie/repos/openbsd/ports
blob - 447e197d6b990febe326014e44f53283b4d68550
file + www/castor/Makefile
--- www/castor/Makefile
+++ www/castor/Makefile
@@ -6,9 +6,7 @@ V= 0.8.16
DISTNAME= castor-${V}
REVISION= 0
-DIST_SUBDIR= castor
-DISTFILES= ${V}${EXTRACT_SUFX}
-MODCARGO_DIST_SUBDIR= # empty
+DISTFILES= castor-{castor/archive/}${V}${EXTRACT_SUFX}
CATEGORIES=www
@@ -22,7 +20,7 @@ PERMIT_PACKAGE= Yes
WANTLIB += c c++abi crypto gdk-3 glib-2.0 gobject-2.0 gtk-3 pthread
WANTLIB += ssl
-MASTER_SITES= https://git.sr.ht/~julienxx/castor/archive/
+MASTER_SITES= https://git.sr.ht/~julienxx/
MODULES= devel/cargo
blob - 6902987be18db426453da8ec6bb43d1e390eef17
file + www/castor/distinfo
--- www/castor/distinfo
+++ www/castor/distinfo
@@ -1,202 +1,202 @@
-SHA256 (castor/0.8.16.tar.gz) = UEFPSYw7YjLvsA6IMEOFRRcpC4QuBQHB0ezPr3i+muM=
-SHA256 (castor/aho-corasick-0.7.13.tar.gz) =
BDFk2LpcTDA1/sm77oZHwCYdeI80dDBvk7tlkByuDoY=
-SHA256 (castor/ansi-parser-0.6.5.tar.gz) =
dhrGdfFjimpJ4m9qw6QGfKP++oApgWrk740/o9BqUZQ=
-SHA256 (castor/arrayref-0.3.6.tar.gz) =
pMUnFS43z3V6P3iq5aBvvu/bB8zFNcmAoyCO4wYN1UQ=
-SHA256 (castor/arrayvec-0.5.1.tar.gz) =
z/d9hoaGfs7/MQUynUaY2WwjkcF21dA63JDHOJFitbg=
-SHA256 (castor/atk-0.8.0.tar.gz) = RE2u+lXyKa8UXqWNd+/SNyUCTuH28xAnQ3CaprGMZj4=
-SHA256 (castor/atk-sys-0.9.1.tar.gz) =
5VLBd2c3pMgBENBrNtCZ9HxyczX5qqXZQqcraGOo7G8=
-SHA256 (castor/autocfg-1.0.0.tar.gz) =
+KrHcPGIX9fjh6zt12BlMCVRNkSW5Gs90Ahgsvg1m50=
-SHA256 (castor/base64-0.11.0.tar.gz) =
tBt+pUoMnZIZneieIOWNSfAvjmmYFO8/3yZvb3SNFcc=
-SHA256 (castor/bitflags-1.2.1.tar.gz) =
zx3i/ox1vBRaL1d63ZUfgTSIm0eV1HRmpUpchG1pFpM=
-SHA256 (castor/blake2b_simd-0.5.10.tar.gz) =
2PstdCVKOgtcrDOsn47Q5EqlA3jZ27Ll2DvSHtHcLIo=
-SHA256 (castor/cairo-rs-0.8.1.tar.gz) =
FXBJupYYqjphw51deFECwE07H0BjKnBsYhqa7cIeYIQ=
-SHA256 (castor/cairo-sys-rs-0.9.2.tar.gz) =
/2W6AsrHFb6Db2NCmrAKdn1IM278VJfFY3r7U7TxTWM=
-SHA256 (castor/cc-1.0.55.tar.gz) = sb40CflNe9zutfX6xVEDnZs/AOJdp6dPxNM0AKDZY2g=
-SHA256 (castor/cfg-if-0.1.10.tar.gz) =
R4W90clrKoRrK9fMAuhraz2/FOflNEbE9UySo2EECCI=
-SHA256 (castor/constant_time_eq-0.1.5.tar.gz) =
JFCX6aRTXuHj45Mfz81Vp5akTGQ+hZb/ZWbWjwm4e7w=
-SHA256 (castor/core-foundation-0.7.0.tar.gz) =
V9JMehPEPocON8FVa3RVVDeHCgRRT3aF9bNU4JBWcXE=
-SHA256 (castor/core-foundation-sys-0.7.0.tar.gz) =
s6catJTAtbhgvchAeuCJeAUkFwcMLO04VzqRV611uKw=
-SHA256 (castor/crossbeam-utils-0.7.2.tar.gz) =
w8fHOi0en8CIagi5PpjrZDRhIw1fGSXkA2IE1fLiYag=
-SHA256 (castor/dirs-2.0.2.tar.gz) =
E66omlyTNkqY6bN7L6I37/u2lNXP4BxbcJQffrCH1eM=
-SHA256 (castor/dirs-sys-0.3.5.tar.gz) =
jpPX9XBd4+SYlaK14LiFWhwn8IAZKunDKmQy1QdBpXo=
-SHA256 (castor/foreign-types-0.3.2.tar.gz) =
9vM564rcBSzSyniRD9qGmu+jjSLVy2SOZIXk0/wG87E=
-SHA256 (castor/foreign-types-shared-0.1.1.tar.gz) =
ALAihBGQjKhoXbp/ws3XDsmZCm51Pom2rJGoTED7r0s=
-SHA256 (castor/futures-channel-0.3.5.tar.gz) =
82atdMKMymukVtleZCKIPPtLJSqDvtkpyDq/278pZ9U=
-SHA256 (castor/futures-core-0.3.5.tar.gz) =
WfX/+Q/V2XH5Nq1nSAJIK6RBtvCbpeFf2LORRVgso5k=
-SHA256 (castor/futures-executor-0.3.5.tar.gz) =
ENa7iIvhFT06vrkAaxGwLPXpsgn9ooaTwxrh5OAS4xQ=
-SHA256 (castor/futures-io-0.3.5.tar.gz) =
3icUKwE6joacFJV+bS7e74npfCieadBC7jpJrNi1F4k=
-SHA256 (castor/futures-macro-0.3.5.tar.gz) =
0LWjCkMoq1Rzh4I3xEczPAkyl73tg6SYPRD03uokDTk=
-SHA256 (castor/futures-task-0.3.5.tar.gz) =
vbZrXwniIBmxqwgw93hbzqjnpCFIaD+ZIU9z+OwhpiY=
-SHA256 (castor/futures-util-0.3.5.tar.gz) =
h2RXT/CLcBoIRILDxwMTSRBLB6yJc5MBBJS+qhjOMsY=
-SHA256 (castor/gdk-0.12.1.tar.gz) =
++Xody/AhlxSRgzdelnX1HcA9E2YCdHdAO7M63aadYk=
-SHA256 (castor/gdk-pixbuf-0.8.0.tar.gz) =
4kgiDEazKbCX1LFY0nF/jGiPFt120DmazoKz6YBivdc=
-SHA256 (castor/gdk-pixbuf-sys-0.9.1.tar.gz) =
2JkbBgqekWG6/Qm/SiAub9QE9bTdGgjVOh6EJW+zSrA=
-SHA256 (castor/gdk-sys-0.9.1.tar.gz) =
at9nnpHRv/DAaGAof4BAPn21TC0kJNzgpHACO1bIj7s=
-SHA256 (castor/getrandom-0.1.14.tar.gz) =
eryN2EUZIWBtgJujLpW2ERklzSkGBg0tzCnAcCIFA+s=
-SHA256 (castor/gio-0.8.1.tar.gz) = DNEPlBXM45tT+AJL85oh+E+BV6+lLaU4N7EC5
Re: New port: net/qtox and databases/sqlcipher and an update to net/toxcore
On Fri, Jan 01, 2021 at 10:42:00PM +, Stuart Henderson wrote:
> On 2021/01/01 18:52, Tom Murphy wrote:
> > Hi,
> >
> >Last port tarball was missing the #define USE_V4L so I've updated
> > it so it works.
> >
> >However, I can't seem to test this. Anything that tries to open
> > /dev/video even with kern.video.record=1 seems to fail.
>
> re the problem opening video:
>
> did you try video(1) too or just things from ports? kern.video.record=1
> should not stop it from opening, it just zeroes the video buffers before
> passing the frames to userland, though I have not updated to a version
> with that yet.
>
> re v4l2 in this port:
>
> it might be better to figure out a bit more what's going on and either
> prove that it works or doesn't work and enable or not based on that.
> v4l applications on OpenBSD often need some other changes than just the
> headers, I've had to workaround a problem with video ioctls failing on a
> few other ports.
I tried video(1) and it works on my system, so it's probably an ioctl.
I'll try to find the cause, but aside from the video device not working,
the rest of the qtox port works. I tested an audio call with a friend
and it worked fine. Also the messaging and other features work. It's
just the video that isn't working yet.
>
> >I'm on:
> >
> >OpenBSD 6.8-current (GENERIC.MP) #248: Tue Dec 29 12:07:27 MST 2020
> >
> >Anyway, it should properly try to open the video device now.
> >
> >Updated port attached as qtox3.tgz here.
> >
> >Thanks,
> >Tom
>
> couple of port tweaks below, I haven't looked closer yet.
>
>
> diff 4a421e6a1869cd82df717c62298316607805167d /usr/ports/mystuff
> blob - 5c7e9fee75b3ea2e473fb1d28a0e3274821b5ed5
> file + net/qtox/Makefile
> --- net/qtox/Makefile
> +++ net/qtox/Makefile
> @@ -1,20 +1,18 @@
> # $OpenBSD: Makefile,v 1.7 2020/04/05 09:45:30 ajacoutot Exp $
>
> -N = qtox
> -V = 1.17.3
> COMMENT =Tox client
>
> -PKGNAME =${N}-${V}
> GH_ACCOUNT = qTox
> GH_PROJECT = qTox
> -GH_TAGNAME = v${V}
> +GH_TAGNAME = v1.17.3
> +PKGNAME =${DISTNAME:L}
>
> CATEGORIES = net
>
> HOMEPAGE = https://qtox.github.io/
> MAINTAINER = Tom Murphy
>
> -# GPLv3
> +# GPLv3+
> PERMIT_PACKAGE = Yes
>
> WANTLIB += ${COMPILER_LIBCXX} KF5SonnetUi Qt5Core Qt5Gui Qt5Network
>
Tweaks look good. I guess all the ${V} and ${N} stuff isn't needed
any more -- looks a lot cleaner, thanks!
Tom
Re: New port: net/qtox and databases/sqlcipher and an update to net/toxcore
On Fri, Jan 01, 2021 at 10:25:36PM +, Stuart Henderson wrote:
> On 2021/01/01 14:30, Tom Murphy wrote:
> > Hi all,
> >
> > I've been wanting to get qtox running (since utox is rather
> > unstable).
> >
> > Attached to this email are two new ports:
> >
> > databases/sqlcipher - Allows you to encrypt sqlite databases
> > net/qtox - qtox client for the tox framework
> >
> > I'm also including a diff for net/toxcore to upgrade it from 0.2.3 to
> > 0.2.12.
>
> Here's a diff I had hanging around in my tree to update
> toxcore/toxic/utox, the API changed a bit and IIRC they all needed
> updating together. Builds OK but I don't use any of this, since
> you have a setup already could you give it a spin please? At least
> toxcore needs SHARED_LIBS bumping that wasn't in your diff.
> Diff for that is sent as a gzipped attachment because it was
> unfortunately imported with MSDOS line endings in the patches
> which usually gets mangled in email.
>
> > I'd like to propose that net/utox might be removed from the tree.
> > It's alpha and pretty unstable (sometimes it makes the cpu fan race and
> > I've had to kill -9 it sometimes)
>
> 0.18.0 changelog for utox suggests it's likely to be better:
> https://github.com/uTox/uTox/releases/tag/v0.18.0
I wasn't able to apply that multiple port patch cleanly. It fails on
toxcore's Makefile and then refuses to run elsewhere. Do I need to
split it out? What command line for patch should I use for it?
> > Note: net/qtox relies on databases/sqlcipher to compile and run.
> > It also relies on the newer net/toxcore. I'm not sure about net/toxic
> > client as I don't use it.
>
> here are a few tweaks for sqlcipher.
>
> diff 4a421e6a1869cd82df717c62298316607805167d /usr/ports/mystuff
> blob - 3daac481dad3c5c05bc9d0fabb11ec97653347cb
> file + databases/sqlcipher/Makefile
> --- databases/sqlcipher/Makefile
> +++ databases/sqlcipher/Makefile
> @@ -1,13 +1,10 @@
> # $OpenBSD: Makefile,v 1.10 2020/10/03 14:58:48 solene Exp $
>
> -N = sqlcipher
> -V = 4.4.2
> COMMENT =encrypted SQLite database
>
> -PKGNAME =${N}-${V}
> GH_ACCOUNT = sqlcipher
> GH_PROJECT = sqlcipher
> -GH_TAGNAME = v${V}
> +GH_TAGNAME = v4.4.2
>
> SHARED_LIBS += sqlcipher 0.0 # 8.6
>
> @@ -27,11 +24,14 @@ CONFIGURE_ARGS += --enable-tempstore=yes \
> --disable-tcl \
> --disable-editline \
> --disable-readline
> +CONFIGURE_ENV += TCLSH_CMD=${MODTCL_BIN}
>
> NO_TEST =Yes
>
> CFLAGS +=-DSQLITE_HAS_CODEC
>
> -BUILD_DEPENDS = lang/tcl/8.6
> +MODULES =lang/tcl
> +MODTCL_VERSION = 8.6
> +BUILD_DEPENDS = ${MODTCL_BUILD_DEPENDS}
>
> .include
>
The tweaks for databases/sqlcipher work fine here. I tested and it
works OK for me.
Thanks,
Tom
