OpenBSD Mail Box: December 2018

Monday, December 31, 2018

sparc64 bulk build report

On Sun, Dec 30, 2018 at 11:18:32PM +0200, Hannu Vuolasaho wrote:
> svn has fixed those too... and man page PREFIX expansion is on TODO.
> So next release should be patchless.
Great!

> There are also --symbols and --profile configure options. The
> debugging option is
> mainly used for portability check with std=c89.
Hm, CFLAGS is currently ignored completely and I missed this previously.
Passing it through CONFIGURE_ARGS will propagate DEBUG for us, so let's
use that.

However, I need to remove the `--debug' flag, otherwise it overwrites
stuff again. Since the configure script is not a script but a binary
being compiled right before execution without proper usage, I'm
reluctant to look any further into this.

> This is harder thing than I originally thought. SUBST_CMD doesn't have effect
> unless braces aren't patched to man page. The man page is installed with cp.
Correct, for using SUBST_CMD sources are usually patched. `cp' is fine
here as the manual ends up with proper permissions.

> There could be post-install target which would delete the installed man page and
> use and install it again with SUBST_MAN which isn't elegant solution.

> The PREFIX expansion wil happen by xschem and install method will change in
> next release, the search for most beautiful installation method is
> rather useless
> as long it is good enough.
Ok, thanks for your work with upstream again.

> First of all I don't know how to disable all the unsupported simulators and
> waveform viewers without causing side-effects.
>
> This release supports Icarus verilog, GHDL and GTKWave from open source
> tool chains and those are for digital only. For analog side there are only
> commercial spices available.
>
> The reason for getting xschem to ports is its capability export tEDAx netlists
> which pcb-rnd can use and I'm working on that. I have few show stoppers there.
>
> Next release of xschem will also support ngspice and maybe the waveform viewer
> problem can be also solved.
Sounds like stuff we can work on after this port got in.

> So I'll try get the upstream to give some configuation option to hide the
> unsupported simulators and viewers.
Cool.

> Attached today's work.
Looks much better. I've made a few minor tweaks:

* zap trailing tab in WANTLIB line, sort libs
* add some newlines and slightly reorder MOD* variables for readability
* pass CFLAGS, drop `--debug' in CONFIGURE_ARGS
* break LIB_DEPENDS into one port per line
* quote variables
* moved targets (pre-configure) to the end as usual
* zap useless NO_TEST comment

What do you think?

Feedback? OK to import?

Re: UPDATE: multimedia/mlt (enable frei0r-plugins)

On Sun, Dec 30, 2018 at 01:45:48PM +0100, Rafael Sadowski wrote:
> Comments, feedback, OK?
* melt.1 is in the sources but not installed
* can you split MAKE_ENV and TEST_ENV into one assignment per line?
* use ${MODQT_QMAKE} (Qt5) in do-test,
`qtmak4' hardcodes a missing TDEP on qt4

One test fails, test.log attached.

> BUILD_DEPENDS = audio/ladspa \
> - textproc/libxml
> + textproc/libxml \
> + multimedia/frei0r-plugins
>
> RUN_DEPENDS-main = audio/ladspa \
> - x11/gtk+3,-guic
> + x11/gtk+3,-guic \
> + multimedia/frei0r-plugins
Please sort these just like LIB_DEPENDS-gpl2.

Re: NEW: net/spoofer-1.4.2

* portcheck(1) reveals hardcoded paths
* please do not use HOMEPAGE in MASTER_SITES
* you can use SEPARATE_BUILD=yes
* COMPILER is required due to -std=gnu++11
* wrong/incomplete WANTLIB, see port-lib-depends-check target
* at least Qt5 missing as dependency
* update-plist picks up spoofer-{cli,gui,scheduler}

Did you miss this or do you want to package the prober only?

When running spoofer-cli or spoofer-gui, I get

Config: Missing "dataDir" in "/etc/xdg/CAIDA/Spoofer.conf"
(No such file or directory). Make sure the scheduler is running
and using the same configuration.

More information is required. At the very least you should improve both
COMMENT and DESCR, but maybe a short README or some docs installed
along the binaries is even better.

Personal preference: space before "=" makes the Makefile easier to read.

Re: UPDATE: sysutils/entr 4.0 -> 4.1

On Mon, Dec 31, 2018 at 04:59:57PM -0500, Eric Radman wrote:
> Yes, this update is good.
Cool.

New diff that

* drops default TEST_TARGET
* drops REV in favour of globbing, making updates simpler
* break MAKE_FLAGS into one assignment per line

OK?

Index: Makefile
===================================================================
RCS file: /cvs/ports/sysutils/entr/Makefile,v
retrieving revision 1.12
diff -u -p -r1.12 Makefile
--- Makefile 20 Feb 2018 19:23:34 -0000 1.12
+++ Makefile 1 Jan 2019 00:58:29 -0000
@@ -2,8 +2,7 @@

COMMENT = run arbitrary commands when files change

-DISTNAME = entr-4.0
-REV = d5110481f5b9
+DISTNAME = entr-4.1

CATEGORIES = sysutils

@@ -20,10 +19,9 @@ MASTER_SITES = http://entrproject.org/co
https://bitbucket.org/eradman/entr/get/

MAKE_FILE = Makefile.bsd
-MAKE_FLAGS = CC="${CC}" LDFLAGS="${LDFLAGS}"
-WRKDIST = ${WRKDIR}/eradman-entr-${REV}
+MAKE_FLAGS = CC="${CC}" \
+ LDFLAGS="${LDFLAGS}"
+WRKDIST = ${WRKDIR}/eradman-entr-*
FAKE_FLAGS = DESTDIR=""
-
-TEST_TARGET = test

.include <bsd.port.mk>
Index: distinfo
===================================================================
RCS file: /cvs/ports/sysutils/entr/distinfo,v
retrieving revision 1.9
diff -u -p -r1.9 distinfo
--- distinfo 20 Feb 2018 19:23:34 -0000 1.9
+++ distinfo 31 Dec 2018 22:02:09 -0000
@@ -1,2 +1,2 @@
-SHA256 (entr-4.0.tar.gz) = StT+kQixeRmZUc/HilgaimlgKwc9rlm8rkuBD24fbIs=
-SIZE (entr-4.0.tar.gz) = 24758
+SHA256 (entr-4.1.tar.gz) = xQPJPM/8HmstqXnNy4jFpobiYRA1Ae7ecHf6CJzf73g=
+SIZE (entr-4.1.tar.gz) = 25265

Re: Update to claws-mail-3.17.2

On Sun, 30 Dec 2018 11:41:08 +0100, Remi Locherer
<remi.locherer@relo.ch> wrote:

> On Mon, Dec 17, 2018 at 09:19:12PM -0500, Daniel Jakots wrote:
> > Hi,
> >
> > Here's a diff to update to its latest release. Changelog is
> > available there: https://claws-mail.org/NEWS
> >
> >
> > I've been using for a day or so and it works fine. As usual, feel
> > free to test and report back.
> >
> > Comments? OK?
> >
>
> builds and runs fine.
> OK remi@

Thanks, I didn't commit it because they released .3. I'm not sure I
understand why it's another point release.
I had difficulties with the WANTLIB stuff about after enough bruteforce
it says now "Scanning: ok"

Here's the diff:

Index: Makefile
===================================================================
RCS file: /cvs/ports/mail/claws-mail/Makefile,v
retrieving revision 1.117
diff -u -p -r1.117 Makefile
--- Makefile 24 Oct 2018 14:28:07 -0000 1.117
+++ Makefile 31 Dec 2018 23:18:33 -0000
@@ -6,18 +6,13 @@ COMMENT-spamassassin= spamassassin plugi
COMMENT-pdfviewer= pdfviewer plugin
COMMENT-gdata= gdata plugin

-V= 3.17.1
+V= 3.17.3
DISTNAME= claws-mail-${V}
PKGNAME-main= ${DISTNAME}
PKGNAME-bogofilter= claws-mail-bogofilter-${V}
PKGNAME-spamassassin= claws-mail-spamassassin-${V}
PKGNAME-pdfviewer= claws-mail-pdfviewer-${V}
PKGNAME-gdata= claws-mail-gdata-${V}
-REVISION-bogofilter= 0
-REVISION-gdata= 0
-REVISION-main= 0
-REVISION-pdfviewer= 0
-REVISION-spamassassin= 0

CATEGORIES= mail news

@@ -59,7 +54,7 @@ WANTLIB-main += croco-0.6 db dbus-1 dbus
WANTLIB-main += gpgme hogweed ical icalss icalvcal icudata icui18n icuuc idn2
WANTLIB-main += ltdl lz4 lzma nettle notify ogg p11-kit perl rsvg-2
WANTLIB-main += startup-notification-1 ${COMPILER_LIBCXX} tasn1 unistring vorbis
-WANTLIB-main += vorbisfile xcb-util xml2
+WANTLIB-main += vorbisfile xcb-util xml2 zstd

WANTLIB-gdata = ${WANTLIB}
WANTLIB-gdata += gck-1 gcr-base-3 gcrypt gdata goa-1.0 gpg-error json-glib-1.0
@@ -67,7 +62,8 @@ WANTLIB-gdata += lzma nspr4 nss3 oauth p
WANTLIB-gdata += soup-2.4 sqlite3 ssl3 xml2

WANTLIB-pdfviewer = ${WANTLIB}
-WANTLIB-pdfviewer += jpeg lcms2 openjp2 poppler poppler-glib tiff
+WANTLIB-pdfviewer += jpeg lcms2 openjp2 png16 poppler poppler-glib tiff
+WANTLIB-pdfviewer += ${COMPILER_LIBCXX}

LIB_DEPENDS= devel/libexecinfo \
mail/libetpan \
@@ -81,6 +77,7 @@ RUN_DEPENDS-main= ${RUN_DEPENDS} \
LIB_DEPENDS-main= ${LIB_DEPENDS} \
archivers/libarchive \
archivers/xz \
+ archivers/zstd \
audio/libcanberra,-gtk \
devel/libnotify \
devel/startup-notification \
@@ -95,7 +92,8 @@ LIB_DEPENDS-pdfviewer= ${LIB_DEPENDS} \
print/poppler

LIB_DEPENDS-gdata= ${LIB_DEPENDS} \
- devel/libgdata
+ devel/libgdata \
+ security/nss

#geolocation looks for champlain <= 0.10
#python tries to dlopen libpython2.7.so
Index: distinfo
===================================================================
RCS file: /cvs/ports/mail/claws-mail/distinfo,v
retrieving revision 1.35
diff -u -p -r1.35 distinfo
--- distinfo 7 Sep 2018 14:08:38 -0000 1.35
+++ distinfo 31 Dec 2018 23:18:33 -0000
@@ -1,2 +1,2 @@
-SHA256 (claws/claws-mail-3.17.1.tar.gz) = WNjMZ6v4+t7mFJP0vuVZmk14oj5Jd5CNMtJIjLx0En8=
-SIZE (claws/claws-mail-3.17.1.tar.gz) = 10171803
+SHA256 (claws/claws-mail-3.17.3.tar.gz) = lQtrOFI1x0F5oHzgnFq5TeP5hpPHBkV8mLRP2QoeYKA=
+SIZE (claws/claws-mail-3.17.3.tar.gz) = 11975996
Index: patches/patch-configure_ac
===================================================================
RCS file: /cvs/ports/mail/claws-mail/patches/patch-configure_ac,v
retrieving revision 1.19
diff -u -p -r1.19 patch-configure_ac
--- patches/patch-configure_ac 7 Sep 2018 14:08:38 -0000 1.19
+++ patches/patch-configure_ac 31 Dec 2018 23:18:33 -0000
@@ -2,7 +2,7 @@ $OpenBSD: patch-configure_ac,v 1.19 2018
Index: configure.ac
--- configure.ac.orig
+++ configure.ac
-@@ -156,7 +156,7 @@ AM_CONDITIONAL(CYGWIN, test x"$env_cygwin" = x"yes")
+@@ -166,7 +166,7 @@ AM_CONDITIONAL(CYGWIN, test x"$env_cygwin" = x"yes")

if test "$GCC" = "yes"
then
@@ -11,7 +11,7 @@ Index: configure.ac
#CFLAGS="-g -Wall -Wno-unused-function"
fi

-@@ -783,6 +783,7 @@ if test x"$enable_alternate_addressbook" = xno; then
+@@ -785,6 +785,7 @@ if test x"$enable_alternate_addressbook" = xno; then
AC_CHECK_LIB(resolv, res_query, LDAP_LIBS="$LDAP_LIBS -lresolv")
AC_CHECK_LIB(socket, bind, LDAP_LIBS="$LDAP_LIBS -lsocket")
AC_CHECK_LIB(nsl, gethostbyaddr, LDAP_LIBS="$LDAP_LIBS -lnsl")
@@ -19,7 +19,7 @@ Index: configure.ac
AC_CHECK_LIB(lber, ber_get_tag, LDAP_LIBS="$LDAP_LIBS -llber",,
$LDAP_LIBS)

-@@ -855,7 +856,7 @@ if test x"$enable_alternate_addressbook" = xno; then
+@@ -857,7 +858,7 @@ if test x"$enable_alternate_addressbook" = xno; then
AC_DEFINE(USE_JPILOT, 1, Define if you want JPilot support in addressbook.) ])
fi

Index: pkg/PLIST-main
===================================================================
RCS file: /cvs/ports/mail/claws-mail/pkg/PLIST-main,v
retrieving revision 1.40
diff -u -p -r1.40 PLIST-main
--- pkg/PLIST-main 7 Sep 2018 14:08:38 -0000 1.40
+++ pkg/PLIST-main 31 Dec 2018 23:18:33 -0000
@@ -41,10 +41,11 @@ include/claws-mail/autofaces.h
include/claws-mail/avatars.h
include/claws-mail/browseldap.h
include/claws-mail/claws-features.h
-include/claws-mail/codeconv.h
include/claws-mail/common/
include/claws-mail/common/claws.h
+include/claws-mail/common/codeconv.h
include/claws-mail/common/defs.h
+include/claws-mail/common/file-utils.h
include/claws-mail/common/hooks.h
include/claws-mail/common/log.h
include/claws-mail/common/md5.h
@@ -67,6 +68,7 @@ include/claws-mail/common/tags.h
include/claws-mail/common/template.h
include/claws-mail/common/timing.h
include/claws-mail/common/tlds.h
+include/claws-mail/common/unmime.h
include/claws-mail/common/utils.h
include/claws-mail/common/uuencode.h
include/claws-mail/common/version.h
@@ -117,7 +119,6 @@ include/claws-mail/gtk/combobox.h
include/claws-mail/gtk/description_window.h
include/claws-mail/gtk/filesel.h
include/claws-mail/gtk/foldersort.h
-include/claws-mail/gtk/gdkkeysyms-new.h
include/claws-mail/gtk/gtkaspell.h
include/claws-mail/gtk/gtkcmclist.h
include/claws-mail/gtk/gtkcmctree.h
@@ -243,7 +244,6 @@ include/claws-mail/summaryview.h
include/claws-mail/textview.h
include/claws-mail/toolbar.h
include/claws-mail/undo.h
-include/claws-mail/unmime.h
include/claws-mail/uri_opener.h
include/claws-mail/vcard.h
include/claws-mail/viewtypes.h
@@ -361,6 +361,9 @@ share/locale/nb/LC_MESSAGES/claws-mail.m
share/locale/nl/LC_MESSAGES/claws-mail.mo
share/locale/pl/LC_MESSAGES/claws-mail.mo
share/locale/pt_BR/LC_MESSAGES/claws-mail.mo
+share/locale/pt_PT/
+share/locale/pt_PT/LC_MESSAGES/
+share/locale/pt_PT/LC_MESSAGES/claws-mail.mo
share/locale/ro/LC_MESSAGES/claws-mail.mo
share/locale/ru/LC_MESSAGES/claws-mail.mo
share/locale/sk/LC_MESSAGES/claws-mail.mo

Re: powerpc bulk build report

On Mon, 31 Dec 2018 00:15:56 -0700 (MST)
landry@openbsd.org wrote:

> http://build-failures.rhaalovely.net//powerpc/2018-12-14/x11/qt5/qtbase.log

Fails to #include <type_traits>, a C++ standard library header.
Is llvm missing a dependency on g++?

At the top of the log, gcc-4.9.4p15 and gcc-libs-4.9.4p15 and
llvm-6.0.1p19 get installed, but g++ doesn't.

> ===> Compiler link: gcc -> /usr/local/bin/clang
> ===> Compiler link: cc -> /usr/local/bin/clang
> ===> Compiler link: clang -> /usr/local/bin/clang
> ===> Compiler link: g++ -> /usr/local/bin/clang++
> ===> Compiler link: c++ -> /usr/local/bin/clang++
> ===> Compiler link: clang++ -> /usr/local/bin/clang++

The compiler is ports-clang from the llvm package. The compiler
needs a C++ library. It can't use libc++ from base-clang because
powerpc doesn't have base-clang. It can't use libstdc++ from
base-gcc because that lib can't do C++11.

It must use libestdc++ from ports-gcc. The library is in the
gcc-libs package but its headers are in the g++ package! This is
odd because most libraries have their headers in the same package.

lang/clang/clang.port.mk uses `MODULES += gcc4` to find libestdc++
and add the dependency on gcc-libs. Then lang/gcc/4.9/gcc4.port.mk
does `MODGCC4_LANGS += c` to add a dependency on the gcc package.
There seems to be no dependency on the g++ package. This might
allow dpb to junk g++ after it builds llvm (using g++) and before
it builds qtbase.

--
George Koehler <kernigh@gmail.com>

Re: UPDATE: sysutils/entr 4.0 -> 4.1

On Mon, Dec 31, 2018 at 10:55:14PM +0100, Sascha Paunovic wrote:
> here an update of sysutils/entr to 4.1. Tested on amd64, works ok for me.
Please name or link to changes so reviewers don't have to dig them up
each on their own.

https://bitbucket.org/eradman/entr/src/default/NEWS

== 4.1: June 13, 2018

- New '-n' non-interactive option disables keyboard input
- EV_TRACE environment variable enables file system event tracing
- Track changes to the inode number as a workaround for missing delete
events on the Linux kernel

Regress passes on amd64, OK kn if someone wants to commit it, otherwise
I'll take care of it after hearing from MAINTAINER (or timeout).

Re: UPDATE: sysutils/entr 4.0 -> 4.1

On Mon, Dec 31, 2018 at 10:55:14PM +0100, Sascha Paunovic wrote:
> Hi,
>
> here an update of sysutils/entr to 4.1. Tested on amd64, works ok for me.

Yes, this update is good.

> Index: Makefile
> ===================================================================
> RCS file: /cvs/ports/sysutils/entr/Makefile,v
> retrieving revision 1.12
> diff -u -p -r1.12 Makefile
> --- Makefile 20 Feb 2018 19:23:34 -0000 1.12
> +++ Makefile 31 Dec 2018 21:53:23 -0000
> @@ -2,8 +2,8 @@
>
> COMMENT = run arbitrary commands when files change
>
> -DISTNAME = entr-4.0
> -REV = d5110481f5b9
> +DISTNAME = entr-4.1
> +REV = f4e2cbe57708
>
> CATEGORIES = sysutils
>
> Index: distinfo
> ===================================================================
> RCS file: /cvs/ports/sysutils/entr/distinfo,v
> retrieving revision 1.9
> diff -u -p -r1.9 distinfo
> --- distinfo 20 Feb 2018 19:23:34 -0000 1.9
> +++ distinfo 31 Dec 2018 21:53:23 -0000
> @@ -1,2 +1,2 @@
> -SHA256 (entr-4.0.tar.gz) = StT+kQixeRmZUc/HilgaimlgKwc9rlm8rkuBD24fbIs=
> -SIZE (entr-4.0.tar.gz) = 24758
> +SHA256 (entr-4.1.tar.gz) = xQPJPM/8HmstqXnNy4jFpobiYRA1Ae7ecHf6CJzf73g=
> +SIZE (entr-4.1.tar.gz) = 25265

UPDATE: sysutils/entr 4.0 -> 4.1

Hi,

here an update of sysutils/entr to 4.1. Tested on amd64, works ok for
me.

Index: Makefile
===================================================================
RCS file: /cvs/ports/sysutils/entr/Makefile,v
retrieving revision 1.12
diff -u -p -r1.12 Makefile
--- Makefile 20 Feb 2018 19:23:34 -0000 1.12
+++ Makefile 31 Dec 2018 21:53:23 -0000
@@ -2,8 +2,8 @@

COMMENT = run arbitrary commands when files change

-DISTNAME = entr-4.0
-REV = d5110481f5b9
+DISTNAME = entr-4.1
+REV = f4e2cbe57708

CATEGORIES = sysutils

Index: distinfo
===================================================================
RCS file: /cvs/ports/sysutils/entr/distinfo,v
retrieving revision 1.9
diff -u -p -r1.9 distinfo
--- distinfo 20 Feb 2018 19:23:34 -0000 1.9
+++ distinfo 31 Dec 2018 21:53:23 -0000
@@ -1,2 +1,2 @@
-SHA256 (entr-4.0.tar.gz) = StT+kQixeRmZUc/HilgaimlgKwc9rlm8rkuBD24fbIs=
-SIZE (entr-4.0.tar.gz) = 24758
+SHA256 (entr-4.1.tar.gz) = xQPJPM/8HmstqXnNy4jFpobiYRA1Ae7ecHf6CJzf73g=
+SIZE (entr-4.1.tar.gz) = 25265

UPDATE: sysutils/entr 4.0 -> 4.1

Hi,

here an update of sysutils/entr to 4.1.

Index: Makefile
===================================================================
RCS file: /cvs/ports/sysutils/entr/Makefile,v
retrieving revision 1.12
diff -u -p -r1.12 Makefile
--- Makefile 20 Feb 2018 19:23:34 -0000 1.12
+++ Makefile 31 Dec 2018 21:39:00 -0000
@@ -2,8 +2,8 @@

COMMENT = run arbitrary commands when files change

-DISTNAME = entr-4.0
-REV = d5110481f5b9
+DISTNAME = entr-4.1
+REV = f4e2cbe57708

CATEGORIES = sysutils

Index: distinfo
===================================================================
RCS file: /cvs/ports/sysutils/entr/distinfo,v
retrieving revision 1.9
diff -u -p -r1.9 distinfo
--- distinfo 20 Feb 2018 19:23:34 -0000 1.9
+++ distinfo 31 Dec 2018 21:39:00 -0000
@@ -1,2 +1,2 @@
-SHA256 (entr-4.0.tar.gz) = StT+kQixeRmZUc/HilgaimlgKwc9rlm8rkuBD24fbIs=
-SIZE (entr-4.0.tar.gz) = 24758
+SHA256 (entr-4.1.tar.gz) = xQPJPM/8HmstqXnNy4jFpobiYRA1Ae7ecHf6CJzf73g=
+SIZE (entr-4.1.tar.gz) = 25265

--
Sascha Paunovic <azarus>

Re: UPDATE: bsd.port.mk - Add GitLab support

On Mon, Dec 31, 2018 at 05:45:19PM +0100, Marc Espie wrote:
> Adding more fluff to bsd.port.mk to support this style of code is fairly
> disturbing.
>
> I don't like the github stuff too much, it's somewhat error-prone and there
> is regular traffic on ports-changes proving it.
>
> Adding a SECOND source of problems does not seem like the way to go.
>
> There's also the issue of getting reliable checksum tarballs... especially
> if that ends up involving REQUIRING dependencies just to be able to fetch
> things. That's something we tried to avoid and that is definitely bug-prone.
>
> That will lead, at the least, to quality-issue problems. And possibly to
> actual security issues.
>
> On platforms where it is possible to have release tarballs that don't change
> I would say that's still a much better choice.
>
> I would very much be in favor of people providing hosting services where this
> does not exist, and to have an actual FAQ of things to tell upstream so that
> they prepare actual properly tagged releases on platforms such as github.

That's what the GNOME project is currently doing (GitLab + some ftp space to
fetch tarballs). But if they end up changing and use generated tarballs from
GitLab directly, do not count on me to go see the 100+ different maintainers of
the GNOME subprojects to ask them to change their policy.

I am in favor of having suppor for gitlab in bsd.port.mk. Lots of projects are
moving there and again it's very convenient because portroach would be able to
warn you when there are updates available (which could also be security
updates).

--
Antoine

fix graphics/inkscape with newer poppler

This will fix inkscape after the update of print/poppler to 0.72.0.

Mainly based on cherry-picked archlinux diffs with some changes
directlry fetched from upstream.

ok (after the poppler update)?

Ciao,
Kili

Index: patches/patch-src_extension_internal_pdfinput_pdf-input_cpp
===================================================================
RCS file: patches/patch-src_extension_internal_pdfinput_pdf-input_cpp
diff -N patches/patch-src_extension_internal_pdfinput_pdf-input_cpp
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ patches/patch-src_extension_internal_pdfinput_pdf-input_cpp 31 Dec 2018 17:57:07 -0000
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+Fix with new poppler. From archlinux.
+
+Index: src/extension/internal/pdfinput/pdf-input.cpp
+--- src/extension/internal/pdfinput/pdf-input.cpp.orig
++++ src/extension/internal/pdfinput/pdf-input.cpp
+@@ -793,7 +793,7 @@ PdfInput::open(::Inkscape::Extension::Input * /*mod*/,
+ dlg->getImportSettings(prefs);
+
+ // Apply crop settings
+- PDFRectangle *clipToBox = NULL;
++ _POPPLER_CONST PDFRectangle *clipToBox = NULL;
+ double crop_setting;
+ sp_repr_get_double(prefs, "cropTo", &crop_setting);
+
Index: patches/patch-src_extension_internal_pdfinput_pdf-input_h
===================================================================
RCS file: patches/patch-src_extension_internal_pdfinput_pdf-input_h
diff -N patches/patch-src_extension_internal_pdfinput_pdf-input_h
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ patches/patch-src_extension_internal_pdfinput_pdf-input_h 31 Dec 2018 17:57:07 -0000
@@ -0,0 +1,15 @@
+$OpenBSD$
+
+Fix with new poppler. From archlinux.
+
+Index: src/extension/internal/pdfinput/pdf-input.h
+--- src/extension/internal/pdfinput/pdf-input.h.orig
++++ src/extension/internal/pdfinput/pdf-input.h
+@@ -15,6 +15,7 @@
+

Re: support new

On Mon, Dec 31, 2018 at 04:00:14PM +0100, info@oriented.net wrote:
> 0
> C Switzerland
> P Basel-Stadt
> T Basel
> Z 4001
> O oriented.net GmbH
> I Marc Bigler
> A Postfach 318
> M info@oriented.net
> U https://oriented.net/
> B +41 61 322 11 01
> X +41 61 322 11 02
> N We provide OpenBSD native virtual machine hosting using VMM, as well
> as consulting and support. We have a strong focus on using OpenBSD in
> network security appliances, such as firewalls, proxies, VPN,
> OpenSMTPD-based mail exchangers etc. We also offer managed UNIX-based
> services, including hosting, servers, support and consulting.

I have added your entry.

Cheers,
Stefan

Re: [update] x11/wxWidgets

On Mon, Dec 31 2018, Jeremie Courreges-Anglas <jca@wxcvbn.org> wrote:
> Hi,
>
> tl;dr, what the diff does:
>
> - update to wxWidgets-3.0.4, the latest version available
>
> - update to wxPython-3.0.2, the latest and probably last "classic"
> release - Aleksander Piotrowski
>
> - adapt consumer ports to cope. Those fall into five categories:
>
> 1. ports already fixed, with few or no patches
>
> 2. ports fixed, but broken at runtime
> - misc/rocrail,-gui (crash) - sebastia@
> - devel/codeblocks (hangs at startup) - Josh Grosse

[...]

> If you're the maintainer of a port among lists 2, 3 or 4, your feedback
> would be greatly appreciated. For rocview and codeblocks which are
> broken at runtime, I'll issue a follow-up with more details.

Regarding codeblocks, I noticed a hang when the editor autodetects
compilers available on the system. But it doesn't happen all the time,
maybe 1 time out of 5. The child process - supposedly the one that
attempts to run various compilers - ends up in a zombie state.

Josh, have you already spotted this?

--8<--
Valgrind plugin activated
Autosave plugin activated
BYO Games plugin activated
Koders query plugin activated
Code completion plugin activated
Code snippets plugin activated
Code statistics plugin activated
ReopenEditor plugin activated
Abbreviations plugin activated
Added compiler "GNU GCC Compiler"
<hangs>
-->8--

ps auxww output:

--8<--
jca 35289 0.0 0.0 0 0 p1 Z+ - 0:00.00 (codeblocks)
jca 96093 0.4 0.6 46184 99264 p1 I+ 6:15PM 0:02.22 codeblocks
-->8--

--
jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE

Re: [UPDATE] devel/p5-Class-Accessor-Grouped 0.10012 -> 0.10014

On Fri, Dec 21, 2018 at 06:34:04PM +0100, Charlene Wendling wrote:
> This is an update for Class::Accessor::Grouped, that will be needed once
> we ship Perl 5.28.

OK bluhm@

Can't commit until p5-Devel-Hide has been imported.

[update] x11/wxWidgets

Hi,

tl;dr, what the diff does:

- update to wxWidgets-3.0.4, the latest version available

- update to wxPython-3.0.2, the latest and probably last "classic"
release - Aleksander Piotrowski

- adapt consumer ports to cope. Those fall into five categories:

1. ports already fixed, with few or no patches

2. ports fixed, but broken at runtime
- misc/rocrail,-gui (crash) - sebastia@
- devel/codeblocks (hangs at startup) - Josh Grosse

3. ports updated to a newer version to cope with the wxWidgets update
- games/golly - Donovan Watteau
- net/filezilla - giovanni@

4. ports marked as broken, newer release available upstream
- audio/audacity (this one is for Antoine :)
- cad/kicad - I think Jérôme Kasper wanted to update it
- math/wxMaxima (already BROKEN) - edd@

5. ports marked as broken, no idea yet how to handle them
- graphics/discwrapper (last release in 2009)
- x11/p5-Wx

If you're the maintainer of a port among lists 2, 3 or 4, your feedback
would be greatly appreciated. For rocview and codeblocks which are
broken at runtime, I'll issue a follow-up with more details.

Regarding the wxWidgets update itself:

- work based on what was in openbsd-wip earlier in 2018
(either from Jérôme Kasper or from jasper@ ?)
- bite the bullet and replace wxWidgets 2 with wxWidgets 3. Lots of
abandonware used our ancient wxWidgets version, a cleanup was due.
- I've kept the -media subpackage as present in the ports tree to cut
down on dependencies.
- big changes: gtk2->gtk3, sdl->sdl2, odbc removed,
webkit->webkitgtk4->disabled for now
- at first I added a -webview subpackage for the parts depending on
webkit, but decided that it could be done afterwards. Please think
about !(x86) before suggesting that. ;)
- soname encoding is force-disabled for now, shouldn't be problem since
symlinks aren't used in the build.

Regarding runtime, nearly all consumers have been (lightly) tested:
- two nasty issues with rocview and codeblocks
- non-fatal assertion failures in poedit, rapidsvn and spatialite_gis
- ports not tested at all:
- lang/erlang/*,-wx
- print/printrun (wxPython consumer)
- comms/gnuradio (wxPython consumer)
- math/prover9/p9m4 is already broken at runtime

I'd like to push this in the tree soonish, so that it gets wider
testing. Yes, codeblocks and rocrail are currently unusable, but IMHO
it's unfair to block progress here for just two programs. There's
plenty of time for us to find fixes before 6.5. ;)

I've asked Antoine if he could add this to an amd64 bulk on exopi.
Reviews and oks welcome.

Re: UPDATE: bsd.port.mk - Add GitLab support

Adding more fluff to bsd.port.mk to support this style of code is fairly
disturbing.

I don't like the github stuff too much, it's somewhat error-prone and there
is regular traffic on ports-changes proving it.

Adding a SECOND source of problems does not seem like the way to go.

There's also the issue of getting reliable checksum tarballs... especially
if that ends up involving REQUIRING dependencies just to be able to fetch
things. That's something we tried to avoid and that is definitely bug-prone.

That will lead, at the least, to quality-issue problems. And possibly to
actual security issues.

On platforms where it is possible to have release tarballs that don't change
I would say that's still a much better choice.

I would very much be in favor of people providing hosting services where this
does not exist, and to have an actual FAQ of things to tell upstream so that
they prepare actual properly tagged releases on platforms such as github.

Re: UPDATE: bsd.port.mk - Add GitLab support

On Sun, 2018-12-30 at 23:57 +0100, Antoine Jacoutot wrote:
> On Sun, Dec 30, 2018 at 10:48:41PM +0000, Stuart Henderson wrote:
> > On 2018/12/29 15:44, Edward Lopez-Acosta wrote:
> > > Stuart,
> > >
> > > I am not sure I understand the question or the issues you refer to. Can you clarify for me so I can look more into it please?
> > >
> > > Are you also proposing the GitHub directives already provided are bad as well?
> > >
> > > On December 29, 2018 2:40:24 PM UTC, Stuart Henderson <stu@spacehopper.org> wrote:
> > > > On 2018/12/29 08:24, Edward Lopez-Acosta wrote:
> > > > > Any feedback for this?
> > > >
> > > > How is gitlab doing at keeping stable distfiles? If it's even worse
> > > > than
> > > > github (and I have a feeling it might be) then I wouldn't really want
> > > > to
> > > > encourage people using it directly as a source.
> >
> > See https://marc.info/?l=openbsd-ports&m=151973450514279&w=2 for more about
> > the problem.

For better or for worse, GitHub, Gitlab, Bitbucket, et. al. are the New
World Order of software development. I don't think not supporting
directives for these platforms is going to solve anything; the
alternative is just "don't package those things". The problem at hand
is the lack of proper releases from upstream.

I was just looking at porting Lollypop[1] since it looks neat and I
believe we have all the dependencies. Even the big guys like GNOME seem
to be using $GIT_FRONTEND_OF_CHOICE to handle their releases nowadays.

At a certain point, we either need to start mirroring all of these
releases somewhere, or find a solution to work around the problems of
this style of software release.

I don't have a good solution to offer up, but I would suggest that
keeping GitLab (or other git hosting platform) support out of
bsd.port.mk just shuffles the problem under the rug, and makes it
harder for people to write ports for software hosted there.

I would further cite GitLab issue 38830[2] in support of merging
Edward's changes; this appears to be something that the GitLab folks
are aware of and have responded to in the past. In particular, GitLab
MR 17225[3] specifically adds support for `project-ref.ext` style
tarballs to support packaging.

Perhaps it would be constructive to reach out to the GitLab development
team directly to see if their archive generations and handling is
compatible with the requirements of the OpenBSD ports tree?

~ Charles

1 - https://wiki.gnome.org/Apps/Lollypop
2 - https://gitlab.com/gitlab-org/gitlab-ce/issues/38830
3 - https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/17225

Re: [NEW] devel/p5-Devel-Hide 0.0010

On Fri, Dec 21, 2018 at 04:32:06PM +0100, Charlene Wendling wrote:
> > I've found out that Devel::Hide is required as a test dep by more
> > than one ports in the tree, so i'm proposing it, and myself as
> > MAINTAINER.
> Fixed tarball with proper licensing infos. Sorry for the noise.

I prefer Makefile variables sorted and indented as in
/usr/ports/infrastructure/templates/Makefile.template or as created
by portgen(1). Then I can easly compare the manually provided
Makefile with the machine generated suggestion.

anyway, OK bluhm@

Re: UPDATE: graphics/opencv 3.4.5

On Sun Dec 30, 2018 at 03:30:00PM +0100, Raphael Graf wrote:
> Hi
>
> Attached is an update to OpenCV 3.4.5, based on the version in openbsd-wip.
>
> It works surprisingly well, on amd64.
>
> The extra modules from https://github.com/opencv/opencv_extra are not
> included.
>
> I'm not sure about the tests, 500MB of testdata needs to be downloaded,
> this seems a bit excessive.. (see attached test.log)

After some tweaks Raphael Graf pushed the last version to openbsd-wip[1].
This version looks really really good and works quite well for me. Thanks
Raphael for your work!

BUT we need to port opencv_contrib[2], more specifically: libface. We
need libface to update x11/kde-applications/libkface and
graphics/digikam-kde4.

I would prefer to build all in opencv. I already stared with some
bits[3] but I need time or more eyes and hands.

Have a good New Year!

[1]: https://github.com/jasperla/openbsd-wip/tree/master/graphics/opencv
[2]: https://github.com/opencv/opencv_contrib
[3]: https://github.com/jasperla/openbsd-wip/commit/f9e6ce4547a6ce5e1236bc5081658eec4c0551f0

support new

0
C Switzerland
P Basel-Stadt
T Basel
Z 4001
O oriented.net GmbH
I Marc Bigler
A Postfach 318
M info@oriented.net
U https://oriented.net/
B +41 61 322 11 01
X +41 61 322 11 02
N We provide OpenBSD native virtual machine hosting using VMM, as well
as consulting and support. We have a strong focus on using OpenBSD in
network security appliances, such as firewalls, proxies, VPN,
OpenSMTPD-based mail exchangers etc. We also offer managed UNIX-based
services, including hosting, servers, support and consulting.

Re: Missing libraries after upgrade to 6.4

On 2018-12-20, John Ankarström <john@ankarstrom.se> wrote:
> Hello all,
>
> I have this port [1] that installed fine on 6.3, but after I upgraded to
> 6.4, following the FAQ, I'm getting weird errors.
>
> When running make install, it fails because qtbase-5.9.4 can't be
> installed, which is weird that it wants to do, because the version in
> ports is 5.9.6p1. Here's the output: http://sprunge.us/z6d97h
>
> When I try to run make rebuild, it complains about missing libraries for
> Qt5 stuff: http://sprunge.us/oEDaKE
>
> Some proof that I've actually upgraded:
>
> $ uname -a
> OpenBSD lbsd.home 6.4 GENERIC.MP#364 amd64
>
> $ cd /usr/ports; cvs status Makefile
> [...]
> Sticky Tag: OPENBSD_6_4 (branch: 1.77.2)
> [...]
>
> Anybody got any idea as to what's going on?
>
> Best regards,
> John
>
> [1]:
> https://files.jkvinge.net/packages/strawberry/strawberry-openbsd-port.tar.gz
>
>

Most likely you have old packages lying around from building
strawberry before. Try "make clean; make clean=all; make repackage reinstall"

[M. UPDATE] devel/py-ipykernel 4.9.0 -> 4.10.0

Index: devel/py-ipykernel/Makefile
===================================================================
RCS file: /cvs/ports/devel/py-ipykernel/Makefile,v
retrieving revision 1.6
diff -u -p -u -r1.6 Makefile
--- devel/py-ipykernel/Makefile 31 Oct 2018 12:38:38 -0000 1.6
+++ devel/py-ipykernel/Makefile 31 Dec 2018 13:37:08 -0000
@@ -2,9 +2,11 @@

COMMENT = IPython kernel for Jupyter

-MODPY_EGG_VERSION = 4.9.0
+MODPY_EGG_VERSION = 4.10.0
DISTNAME = ipykernel-${MODPY_EGG_VERSION}
PKGNAME = py-${DISTNAME}
+
+PORTROACH= limit:^4.*$$

CATEGORIES = devel

Index: devel/py-ipykernel/distinfo
===================================================================
RCS file: /cvs/ports/devel/py-ipykernel/distinfo,v
retrieving revision 1.4
diff -u -p -u -r1.4 distinfo
--- devel/py-ipykernel/distinfo 31 Oct 2018 12:38:38 -0000 1.4
+++ devel/py-ipykernel/distinfo 31 Dec 2018 13:37:08 -0000
@@ -1,2 +1,2 @@
-SHA256 (ipykernel-4.9.0.tar.gz) = xaSYxw93ZcNPM5fPlDsGkFf1vvTgIY5M+7cz6fOPpfo=
-SIZE (ipykernel-4.9.0.tar.gz) = 95255
+SHA256 (ipykernel-4.10.0.tar.gz) = aZEDyOZIhuPscFPypqqDu5BCYGNSb2OoGHMv84UgK60=
+SIZE (ipykernel-4.10.0.tar.gz) = 96242
Changelog:
https://github.com/ipython/ipykernel/tree/5.1.0

4.10.0:
- Fix compatibility with IPython 7.0
- Fix compatibility in cases where sys.stdout can be None

These changes shouldn't affect the current consumers.

Versions >= 5.0 are python3 only, so we are sticking with 4.X for now.
Adding the following option so portroach only gets the 4.X branch:
limit:^4.*$$

Regression tests looking good:
Ran 95 (1 skipped)

Also the following regression tests were executed to check for problems:
devel/ipython => OK
devel/py-jupyter_client => OK
devel/py-jupyter_core => OK
devel/py-nbconvert => OK
shells/py-qtconsole => OK
www/jupyter-notebook => OK

devel/spyder runs OK with the change.

Cheers.
Elias.

Re: powerpc lang/gcc/6 can't build devel/llvm

On Sun, Dec 30 2018, George Koehler <kernigh@gmail.com> wrote:
> Hello OpenBSD ports list!

Hi!

> I want to report a problem with lang/gcc/6 on powerpc, where it can't
> build devel/llvm. I have reproduced the problem in a small C++ file
> that errors when the powerpc option -mlongcall is passed to eg++.
>
> $ cat loopy.cpp
> #include <functional>
>
> struct Thing {
> int get() const { return 0; }
> };
>
> int f(Thing t) {
> return std::mem_fn(&Thing::get)(t);
> }
> $ eg++ --version
> eg++ (GCC) 6.5.0
> ...
> $ eg++ -mlongcall -c loopy.cpp
> ...
> (see attached loopy-error.txt)
>
> The error goes away if I do any _one_ of the following:
> - remove -mlongcall

Using -mlongcall dates back from 2017, but it did not actually solve
relocation overflows, iirc due to code in /usr/src/lib/csu/. Later the
size of the llvm executables on powerpc was scaled down using

CONFIGURE_ARGS += -DLLVM_TARGETS_TO_BUILD="PowerPC"

so maybe -mlongcall is not needed at all now.

If you remove CFLAGS/CXXFLAGS = -mlongcall and restart the build from
scratch, are clang and friends properly linked? Do you get a working
llvm package?

--
jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE

[new] lang/gcc/8

Here's a port for GCC 8.2.0. Fully tested on aarch64, amd64, i386, arm,
sparc64, hppa. Unfortunately, both my powerpc and mips64 crashed during
the build, so new adastraps will have to wait until I'm back home.

Notable changes:

* gcj has been *removed*
* Ada has switched its exception handling to ZCX (instead of sjlj).
This does not yet work on sparc64, so I've disabled the -ada subpackage
for now.
* aarch64 and armv7 now use devel/gas from ports at build and run time
* C++17 and experimental C++2a support

Re: howto set terminus font in .Xresources for xterm

niyalevi@gmail.com (shadrock uhuru), 2019.12.31 (Mon) 11:01 (CET):
> what is the correct command to put in .Xresources for the terminus font,

Works for me:

XTerm*faceName:Terminus*
XTerm*faceSize:12

Marcus

> the following is my Xresources file,
> i've tried a few variation but all i get when i start xterm is cannot
> load font,
> font loading is new to me so i have only try examples off the web
> ---
>
>
> XTerm*utf8: 1
> ! XTerm*font: -*-terminus-medium-*-*-*-18-*-*-*-*-*-iso10646-1
> XTerm*font: terminus-12
> XTerm*italicFont: terminus-12
> XTerm*selectToClipboard: true
>
>
> !    ! Use a nice truetype font and size by default...
> !    xterm*faceName: DejaVu Sans Mono Book
> !    xterm*faceSize: 11
>
> xterm*loginshell: true
>
> xterm*savelines: 16384
>
> ! double-click to select whole URLs :D
> xterm*charClass: 33:48,36-47:48,58-59:48,61:48,63-64:48,95:48,126:48
> XTerm*on3Clicks: regex
> ([[:alpha:]]+://)?([[:alnum:]!#+,./=?@_~-]|(%[[:xdigit:]][[:xdigit:]]))+
> *VT100*translations: #override Shift <Btn1Up>:
> exec-formatted("google-chrome '%t'", PRIMARY)
>
> ! DOS-box colours...
> !    xterm*foreground: rgb:a8/a8/a8
>     xterm*foreground: rgb:ff/ff/00
>     xterm*background: rgb:00/00/00
>     xterm*color0: rgb:00/00/00
>     xterm*color1: rgb:a8/00/00
>     xterm*color2: rgb:00/a8/00
>     xterm*color3: rgb:a8/54/00
>     xterm*color4: rgb:00/00/a8
>     xterm*color5: rgb:a8/00/a8
>     xterm*color6: rgb:00/a8/a8
>     xterm*color7: rgb:a8/a8/a8
>     xterm*color8: rgb:54/54/54
>     xterm*color9: rgb:fc/54/54
>     xterm*color10: rgb:54/fc/54
>     xterm*color11: rgb:fc/fc/54
>     xterm*color12: rgb:54/54/fc
>     xterm*color13: rgb:fc/54/fc
>     xterm*color14: rgb:54/fc/fc
>     xterm*color15: rgb:fc/fc/fc
>
> ! right hand side scrollbar...
>     xterm*rightScrollBar: true
>     xterm*ScrollBar: true
>
> ! stop output to terminal from jumping down to bottom of scroll again
>     xterm*scrollTtyOutput: false
>
> ---
>
> thanks shadrock
>

howto set terminus font in .Xresources for xterm

hi everyone

what is the correct command to put in .Xresources for the terminus font,
the following is my Xresources file,
i've tried a few variation but all i get when i start xterm is cannot
load font,
font loading is new to me so i have only try examples off the web
---

XTerm*utf8: 1
! XTerm*font: -*-terminus-medium-*-*-*-18-*-*-*-*-*-iso10646-1
XTerm*font: terminus-12
XTerm*italicFont: terminus-12
XTerm*selectToClipboard: true

!    ! Use a nice truetype font and size by default...
!    xterm*faceName: DejaVu Sans Mono Book
!    xterm*faceSize: 11

xterm*loginshell: true

xterm*savelines: 16384

! double-click to select whole URLs :D
xterm*charClass: 33:48,36-47:48,58-59:48,61:48,63-64:48,95:48,126:48
XTerm*on3Clicks: regex
([[:alpha:]]+://)?([[:alnum:]!#+,./=?@_~-]|(%[[:xdigit:]][[:xdigit:]]))+
*VT100*translations: #override Shift <Btn1Up>:
exec-formatted("google-chrome '%t'", PRIMARY)

! DOS-box colours...
!    xterm*foreground: rgb:a8/a8/a8
    xterm*foreground: rgb:ff/ff/00
    xterm*background: rgb:00/00/00
    xterm*color0: rgb:00/00/00
    xterm*color1: rgb:a8/00/00
    xterm*color2: rgb:00/a8/00
    xterm*color3: rgb:a8/54/00
    xterm*color4: rgb:00/00/a8
    xterm*color5: rgb:a8/00/a8
    xterm*color6: rgb:00/a8/a8
    xterm*color7: rgb:a8/a8/a8
    xterm*color8: rgb:54/54/54
    xterm*color9: rgb:fc/54/54
    xterm*color10: rgb:54/fc/54
    xterm*color11: rgb:fc/fc/54
    xterm*color12: rgb:54/54/fc
    xterm*color13: rgb:fc/54/fc
    xterm*color14: rgb:54/fc/fc
    xterm*color15: rgb:fc/fc/fc

! right hand side scrollbar...
    xterm*rightScrollBar: true
    xterm*ScrollBar: true

! stop output to terminal from jumping down to bottom of scroll again
    xterm*scrollTtyOutput: false

---

thanks shadrock

[UPDATE] www/py-beautifulsoup4

Index: Makefile
===================================================================
RCS file: /cvs/ports/www/py-beautifulsoup4/Makefile,v
retrieving revision 1.8
diff -u -p -u -p -r1.8 Makefile
--- Makefile 14 May 2017 14:20:46 -0000 1.8
+++ Makefile 31 Dec 2018 09:30:51 -0000
@@ -2,7 +2,7 @@

COMMENT = HTML/XML parser that supports invalid markup

-MODPY_EGG_VERSION = 4.6.0
+MODPY_EGG_VERSION = 4.6.3
DISTNAME = beautifulsoup4-${MODPY_EGG_VERSION}
PKGNAME = py-${DISTNAME}

Index: distinfo
===================================================================
RCS file: /cvs/ports/www/py-beautifulsoup4/distinfo,v
retrieving revision 1.6
diff -u -p -u -p -r1.6 distinfo
--- distinfo 14 May 2017 14:20:46 -0000 1.6
+++ distinfo 31 Dec 2018 09:30:51 -0000
@@ -1,2 +1,2 @@
-SHA256 (beautifulsoup4-4.6.0.tar.gz) = gItqyTLcywpBJlWPff3PQXEN1EpO9Jegu1mnf58Hjok=
-SIZE (beautifulsoup4-4.6.0.tar.gz) = 160846
+SHA256 (beautifulsoup4-4.6.3.tar.gz) = kPjmESHWrlg2LOO+2M2ZfvsAyRTq4P89Njwy+amCLRA=
+SIZE (beautifulsoup4-4.6.3.tar.gz) = 167469
Hi,

this is the diff to update beautifulsoup4 to latest release.

Ok?

Cheers,

Remi.

Re: [UPDATE] sysutils/glide 0.13.1 -> 0.13.2

ping :-)

On 10/27/18 8:40 AM, fredl wrote:
> ping
>
> On 10/2/18 10:15 AM, fredl wrote:
>> Hey,
>> attached diff updates sysutils/glide from 0.13.1 to 0.13.2
>>
>> release notes:
>> https://github.com/Masterminds/glide/releases/tag/v0.13.2
>>
>

Re: CoC loving Linux programmers swear the GPLv2 is irrevocable. They are wrong. (As are the women they wish to empower).

Thank you very much for this message, friend. We the people should
learn it and stop taking Cock of Conduct up our ass

пн, 31 дек. 2018 г. в 05:07, <vnsndalce@memeware.net>:
>
> Their take is that if you lent (licensed) them a lawnmower and told them
> not to wreck it, the fact that they did not wreck it entitles them to
> keep the lawnmower forever(they followed your instruction regarding the
> use of your property: "thus consideration, thus irrevocable license").
>
> They are wrong. You can revoke the license.
>
> You were paid nothing for this grant.
> The takers have nothing.
>
> Yet they are taking your positions, your seniority, controlling your
> speech, and ejecting you from your communities.
>
> You must fight back. They have already taken from you all the natural
> pleasures in life (which include pretty young girls as brides, ruling
> over your girls and family, being the master - not the servant of the
> woman and her state).
>
> Now they are taking this edifice, this refuge, which you have built for
> yourself.
>
> They see you as replaceable, expendable, removable employees.
> (See: Bruce Perens statements on slashdot and this and other mailing
> lists about how those who do not like the Code of Conduct can be easily
> replaced)
>
> But you are property owners and may elect to enforce your rights as
> such. And you should do so.
>
> And yes, You can trust my words. I am a licensed attorney, and I'm on
> your side. Not the side of the women. Not the side of people like Bruce
> Perens who see you as a replaceable cog: a worker... a chattel. Your
> side. The side of the dispossessed men who are ruled over by the women
> and the supporters of the women: who will not tolerate any corner of
> advertence not fixated on them.
>
>

Sunday, December 30, 2018

powerpc bulk build report

Re: UPDATE: bsd.port.mk - Add GitLab support

On Sun, Dec 30, 2018 at 10:48:41PM +0000, Stuart Henderson wrote:
> On 2018/12/29 15:44, Edward Lopez-Acosta wrote:
> > Stuart,
> >
> > I am not sure I understand the question or the issues you refer to. Can you clarify for me so I can look more into it please?
> >
> > Are you also proposing the GitHub directives already provided are bad as well?
> >
> > On December 29, 2018 2:40:24 PM UTC, Stuart Henderson <stu@spacehopper.org> wrote:
> > >On 2018/12/29 08:24, Edward Lopez-Acosta wrote:
> > >> Any feedback for this?
> > >
> > >How is gitlab doing at keeping stable distfiles? If it's even worse
> > >than
> > >github (and I have a feeling it might be) then I wouldn't really want
> > >to
> > >encourage people using it directly as a source.
> >
>
> See https://marc.info/?l=openbsd-ports&m=151973450514279&w=2 for more about
> the problem.

I agree the situation is far from nice when using on-demand generated tarballs.
But, it's really handy for portroach, so we know when there's an update
available.

--
Antoine

Re: UPDATE: bsd.port.mk - Add GitLab support

On 2018/12/29 15:44, Edward Lopez-Acosta wrote:
> Stuart,
>
> I am not sure I understand the question or the issues you refer to. Can you clarify for me so I can look more into it please?
>
> Are you also proposing the GitHub directives already provided are bad as well?
>
> On December 29, 2018 2:40:24 PM UTC, Stuart Henderson <stu@spacehopper.org> wrote:
> >On 2018/12/29 08:24, Edward Lopez-Acosta wrote:
> >> Any feedback for this?
> >
> >How is gitlab doing at keeping stable distfiles? If it's even worse
> >than
> >github (and I have a feeling it might be) then I wouldn't really want
> >to
> >encourage people using it directly as a source.
>

See https://marc.info/?l=openbsd-ports&m=151973450514279&w=2 for more about
the problem.

Re: update/fix: devel/boost

On 2018/12/30 19:05, Rafael Sadowski wrote:
> +Index: tools/build/src/engine/Jambase
> +--- tools/build/src/engine/Jambase.orig
> ++++ tools/build/src/engine/Jambase
> +@@ -51,7 +51,7 @@ BOOST_BUILD_PATH.user-value = $(BOOST_BUILD_PATH) ;
> + # Policy.
> + if ! $(BOOST_BUILD_PATH) && $(UNIX)
> + {
> +- BOOST_BUILD_PATH = /usr/share/boost-build ;
> ++ BOOST_BUILD_PATH = ${LOCALBASE}/share/boost-build ;

from bsd.port.mk(5) :

LOCALBASE
where other ports have already been installed. Default: /usr/local.

PREFIX Base directory for the current port installation.

this refers to "current port" so should be prefix.

Re: NEW: net/spoofer-1.4.2

On 2018/12/22 14:49, Denis Fondras wrote:
> The CAIDA Spoofer project measures the Internet's susceptibility to spoofed
> source address IP packets. The Spoofer client attempts to send a series of
> spoofed UDP packets that test the ability to send different classes of spoofed
> traffic
>
> http://spoofer.caida.org/

Initial tweaks. It's protobuf not protobuf-c, c++ ports need to set
COMPILER except in special cases, and don't hardcode /usr/local.

Additionally either the Qt bits need adding properly or disabling
(build fails if qt3-mt is installed. it builds extra things if qt5
is installed. haven't looked at qt4).

diff --git Makefile Makefile
index 3f06033..1395bc6 100644
--- Makefile
+++ Makefile
@@ -7,18 +7,20 @@ CATEGORIES= net

HOMEPAGE= https://www.caida.org/projects/spoofer/

-MASTER_SITES= ${HOMEPAGE}downloads/
+MASTER_SITES= https://www.caida.org/projects/spoofer/downloads/

MAINTAINER= Denis Fondras <denis@openbsd.org>

# GPLv3
PERMIT_PACKAGE_CDROM= Yes

-WANTLIB+= protobuf-c
+WANTLIB += ${COMPILER_LIBCXX} c crypto m pcap protobuf-lite ssl

-LIB_DEPENDS= devel/protobuf-c
+COMPILER= base-clang ports-gcc
+
+LIB_DEPENDS= devel/protobuf

CONFIGURE_STYLE= gnu
-CONFIGURE_ARGS= --with-protobuf=/usr/local
+CONFIGURE_ARGS= --with-protobuf=${LOCALBASE}

.include <bsd.port.mk>

powerpc lang/gcc/6 can't build devel/llvm

loopy.cpp: In function 'int f(Thing)':
loopy.cpp:8:35: error: no match for call to '(std::_Mem_fn<int (Thing::*)() const>) (Thing&)'
return std::mem_fn(&Thing::get)(t);
^
In file included from loopy.cpp:1:0:
/usr/local/include/c++/6.5.0/functional:625:2: note: candidate: template<class _Tp> decltype (std::__invoke(((const std::_Mem_fn_base<_MemObjPtr, false>*)this)->std::_Mem_fn_base<_MemObjPtr, false>::_M_pm, forward<_Tp>(__obj))) std::_Mem_fn_base<_MemObjPtr, false>::operator()(_Tp&&) const [with _Tp = _Tp; _MemObjPtr = int (Thing::*)() const]
operator()(_Tp&& __obj) const
^~~~~~~~
/usr/local/include/c++/6.5.0/functional:625:2: note: template argument deduction/substitution failed:
/usr/local/include/c++/6.5.0/functional: In substitution of 'template<class _Tp> decltype (std::__invoke(((const std::_Mem_fn_base<_MemObjPtr, false>*)this)->std::_Mem_fn_base<_MemObjPtr, false>::_M_pm, forward<_Tp>(__obj))) std::_Mem_fn_base<_MemObjPtr, false>::operator()(_Tp&&) const [with _Tp = Thing&]':
loopy.cpp:8:35: required from here
/usr/local/include/c++/6.5.0/functional:627:27: error: no matching function for call to '__invoke(int (Thing::* const&)() const, Thing&)'
-> decltype(std::__invoke(_M_pm, std::forward<_Tp>(__obj)))
~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/usr/local/include/c++/6.5.0/functional:245:5: note: candidate: template<class _Callable, class ... _Args> typename std::result_of<_Callable&&(_Args&& ...)>::type std::__invoke(_Callable&&, _Args&& ...)
__invoke(_Callable&& __fn, _Args&&... __args)
^~~~~~~~
/usr/local/include/c++/6.5.0/functional:245:5: note: template argument deduction/substitution failed:
/usr/local/include/c++/6.5.0/functional: In substitution of 'template<class _Callable, class ... _Args> typename std::result_of<_Callable&&(_Args&& ...)>::type std::__invoke(_Callable&&, _Args&& ...) [with _Callable = int (Thing::* const&)() const; _Args = {Thing&}]':
/usr/local/include/c++/6.5.0/functional:627:27: required by substitution of 'template<class _Tp> decltype (std::__invoke(((const std::_Mem_fn_base<_MemObjPtr, false>*)this)->std::_Mem_fn_base<_MemObjPtr, false>::_M_pm, forward<_Tp>(__obj))) std::_Mem_fn_base<_MemObjPtr, false>::operator()(_Tp&&) const [with _Tp = Thing&]'
loopy.cpp:8:35: required from here
/usr/local/include/c++/6.5.0/functional:245:5: error: no type named 'type' in 'class std::result_of<int (Thing::* const&(Thing&))() const>'
Index: infrastructure/mk/fortran.port.mk
===================================================================
RCS file: /cvs/ports/infrastructure/mk/fortran.port.mk,v
retrieving revision 1.14
diff -u -p -u -r1.14 fortran.port.mk
--- infrastructure/mk/fortran.port.mk 2 Sep 2016 13:02:24 -0000 1.14
+++ infrastructure/mk/fortran.port.mk 30 Dec 2018 21:10:58 -0000
@@ -14,9 +14,9 @@ MODFORTRAN_WANTLIB += g2c
MODULES += gcc4
MODGCC4_ARCHS ?= *
MODGCC4_LANGS += fortran
-MODFORTRAN_BUILD_DEPENDS += lang/gcc/4.9,-f95>=4.9.4p0,<4.10
+MODFORTRAN_BUILD_DEPENDS += lang/gcc/6,-f95>=6,<7
MODFORTRAN_LIB_DEPENDS += ${MODGCC4_CPPLIBDEP}
-MODFORTRAN_WANTLIB += gfortran>=3
+MODFORTRAN_WANTLIB += gfortran>=7
. if ${MACHINE_ARCH} == "amd64" || ${MACHINE_ARCH} == "i386"
MODFORTRAN_WANTLIB += quadmath
. endif
Index: infrastructure/mk/gcc4.port.mk
===================================================================
RCS file: /cvs/ports/infrastructure/mk/gcc4.port.mk,v
retrieving revision 1.12
diff -u -p -u -r1.12 gcc4.port.mk
--- infrastructure/mk/gcc4.port.mk 8 Mar 2016 16:46:05 -0000 1.12
+++ infrastructure/mk/gcc4.port.mk 30 Dec 2018 21:10:58 -0000
@@ -1,2 +1,2 @@
-MODGCC4_VERSION?=4.9
+MODGCC4_VERSION?=6
.include "${PORTSDIR}/lang/gcc/${MODGCC4_VERSION}/gcc4.port.mk"
Index: lang/gcc/Makefile
===================================================================
RCS file: /cvs/ports/lang/gcc/Makefile,v
retrieving revision 1.15
diff -u -p -u -r1.15 Makefile
--- lang/gcc/Makefile 19 Oct 2018 09:19:40 -0000 1.15
+++ lang/gcc/Makefile 30 Dec 2018 21:10:58 -0000
@@ -1,6 +1,6 @@
# $OpenBSD: Makefile,v 1.15 2018/10/19 09:19:40 sthen Exp $

SUBDIR =
- SUBDIR += 4.9
+ SUBDIR += 6

.include <bsd.port.subdir.mk>
Hello OpenBSD ports list!

I want to report a problem with lang/gcc/6 on powerpc, where it can't
build devel/llvm. I have reproduced the problem in a small C++ file
that errors when the powerpc option -mlongcall is passed to eg++.

$ cat loopy.cpp
#include <functional>

struct Thing {
int get() const { return 0; }
};

int f(Thing t) {
return std::mem_fn(&Thing::get)(t);
}
$ eg++ --version
eg++ (GCC) 6.5.0
...
$ eg++ -mlongcall -c loopy.cpp
...
(see attached loopy-error.txt)

The error goes away if I do any _one_ of the following:
- remove -mlongcall
- remove const in `int get() const`
- change `std::mem_fn(&Thing::get)(t)` to `(t.*(&Thing::get))()`

The default ports-gcc is lang/gcc/4.9; I am using a diff (attached
as gcc-6-2.diff) to switch ports-gcc to lang/gcc/6. This diff was
from https://marc.info/?l=openbsd-ports&m=154165621429861&w=2 but
I removed the chunk that disables Ada.

I also edited devel/llvm/Makefile to change GCC_VER from 4.9.0 to
6.4.0; this was before I built gcc-6.5.0p0.

My OpenBSD/macppc kernel has the powerpc swap fix from Mark Kettenis
in https://marc.info/?l=openbsd-bugs&m=154559013731861&w=2 so programs
don't crash while I am building devel/llvm. With the swap fix,
lang/gcc/6 can build much of devel/llvm but can't build
LoopLoadElimination.cpp where that file uses std::mem_fn.

devel/llvm uses -mlongcall because its port Makefile adds -mlongcall
to CFLAGS and CXXFLAGS for powerpc. -mlongcall also seems to cause
a bunch of "warning: ignoring attributes on template argument '...'`.
(I guess that -mlongcall sets an attribute on everything, but some
template parameters ignore the attribute?) I got the error from
LoopLoadElimination.cpp with gcc-6.4.0; then I used loopy.cpp to
reproduce the error in both 6.4.0 and gcc-6.5.0p0.

--
George Koehler <kernigh@gmail.com>