Re: Problem building py-cryptography
On Fri, May 21, 2021, at 10:29 AM, Dan Langille wrote:
> On Thu, May 20, 2021, at 7:18 PM, Simon Wright wrote:
> > Thanks Kubs, I'm travelling at the moment and will check further when I'm
> > back home. The original build log was with libressl and ccache. I'll
> > repeat without both and attach a new build log.
>
> FYI, I have build failures here today, based on FreeBSD 11.4. Builds
> fine on 12.2 though.
>
> https://services.unixathome.org/poudriere/data/114R-dvl/2021-05-21_13h56m56s/logs/errors/py38-cryptography-2.9.2.log
>
> ===> FAILED Applying FreeBSD patch-Fix-build-with-LibreSSL-3.3.2-5988
> ===> FAILED to apply cleanly FreeBSD patch(es)
> patch-Fix-build-with-LibreSSL-3.3.2-5988
>
> For me, on 11.4, fixed with:
>
> rm py-cryptography/files/patch-Fix-build-with-LibreSSL-3.3.2-5988
>
> With that file removed, 12.2 does not build.
Correction. It builds on 12.2. I was wrong, sorry.
>
>
> hope this helps.
>
> >
> > Apologies for the top post.
> >
> > Regards,
> >
> > Simon.
> >
> > 20 May 2021 13:50:54 Kubilay Kocak :
> >
> > > On 20/05/2021 2:17 pm, Simon Wright wrote:
> > >> On 20/05/2021 12:00 pm, Kubilay Kocak wrote:
> > >>> On 20/05/2021 1:21 pm, Simon Wright wrote:
> > >>>> Hi all,
> > >>>>
> > >>>> I've been unable to build security/py-cryptography for about 10 days
> > >>>> now. The build in Poudriere under 12.2 and 13.0 fail with a
> > >>>> "Bad_C++_code" error.
> > >>>>
> > >>>> I tried removing the libressl dependency but that made no
> > >>>> difference. Is
> > >>>> anyone else seeing this and can anyone point me in the right
> > >>>> direction
> > >>>> to get this fixed please?
> > >>>>
> > >>>> Below is my make.conf, list of poudriere-built ports and the full
> > >>>> poudriere log for py-cryptography
> > >>>>
> > >>>
> > >>> Hi Simon,
> > >>>
> > >>> Is the issue reproducible without ccache?
> > >>>
> > >>> Also, make.conf still shows:
> > >>>
> > >>> DEFAULT_VERSIONS+=ssl=libressl
> > >>>
> > >> Thanks Kubs. When I tested I removed libressl, tried the build again
> > >> and
> > >> it failed so I replaced libressl after the test.
> > >> Removing ccache (and with libressl) made no difference - still the
> > >> same
> > >> error. I then removed the libressl dependency and I reran the build -
> > >> no
> > >> ccache and no libressl - the build still failed, same error message.
> > >> Since no-one else has reported this I suppose it should be something
> > >> in
> > >> my environment . . . . But what? :)
> > >
> > > While the build error *with libressl* is known, matching that reported
> > > in:
> > >
> > > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=255241
> > >
> > > Failing to build without libressl is unexpected.
> > >
> > > Jump into the poudriere jail to confirm (or not) its libressl that's
> > > being installed and used. Could be:
> > >
> > > - Custom WRKDIRPREFIX?
> > > - An overriding jail or set (-z) specific poudriere foo-make.conf?
> > >
> > > Note, the OP build log contains:
> > >
> > > [pkg.home.santos-wright.net] | | `-- Installing libressl-3.3.3...
> > > [pkg.home.santos-wright.net] | | `-- Extracting libressl-3.3.3:
> > >
> > > With defaults, base openssl will be used, and you wont see libressl as
> > > a dependency in the build.
> > >
> > > [1] build/temp.freebsd-13.0-RELEASE-amd64-3.8/_openssl.c:2172:19:
> > > error:
> > > expected identifier or '('
> > > static const long SSL_OP_NO_DTLSv1 = 0;
> > >
> > >> Regards,
> > >> Simon.
> > >>
> > >>>>
> > >>>>
> > >>>> make.conf
> > >>>>
> > >>>> WRKDIRPREFIX=/usr/tmp
> > >>>> OPTIONS_SET=GECKO CUPS
> > >>>> NOI4B=1
> > >>>> OPTIONS_SET+=NO-X11
> > >>>> CUPS_OVERWRITE_BASE=YES
> > >>>> WITH_VIM_OPTIONS=yes
> > >>>> DEFAULT_VERSIONS+=ssl=libress
Re: Problem building py-cryptography
On Thu, May 20, 2021, at 7:18 PM, Simon Wright wrote:
> Thanks Kubs, I'm travelling at the moment and will check further when I'm
> back home. The original build log was with libressl and ccache. I'll
> repeat without both and attach a new build log.
FYI, I have build failures here today, based on FreeBSD 11.4. Builds fine on
12.2 though.
https://services.unixathome.org/poudriere/data/114R-dvl/2021-05-21_13h56m56s/logs/errors/py38-cryptography-2.9.2.log
===> FAILED Applying FreeBSD patch-Fix-build-with-LibreSSL-3.3.2-5988
===> FAILED to apply cleanly FreeBSD patch(es)
patch-Fix-build-with-LibreSSL-3.3.2-5988
For me, on 11.4, fixed with:
rm py-cryptography/files/patch-Fix-build-with-LibreSSL-3.3.2-5988
With that file removed, 12.2 does not build.
hope this helps.
>
> Apologies for the top post.
>
> Regards,
>
> Simon.
>
> 20 May 2021 13:50:54 Kubilay Kocak :
>
> > On 20/05/2021 2:17 pm, Simon Wright wrote:
> >> On 20/05/2021 12:00 pm, Kubilay Kocak wrote:
> >>> On 20/05/2021 1:21 pm, Simon Wright wrote:
> >>>> Hi all,
> >>>>
> >>>> I've been unable to build security/py-cryptography for about 10 days
> >>>> now. The build in Poudriere under 12.2 and 13.0 fail with a
> >>>> "Bad_C++_code" error.
> >>>>
> >>>> I tried removing the libressl dependency but that made no
> >>>> difference. Is
> >>>> anyone else seeing this and can anyone point me in the right
> >>>> direction
> >>>> to get this fixed please?
> >>>>
> >>>> Below is my make.conf, list of poudriere-built ports and the full
> >>>> poudriere log for py-cryptography
> >>>>
> >>>
> >>> Hi Simon,
> >>>
> >>> Is the issue reproducible without ccache?
> >>>
> >>> Also, make.conf still shows:
> >>>
> >>> DEFAULT_VERSIONS+=ssl=libressl
> >>>
> >> Thanks Kubs. When I tested I removed libressl, tried the build again
> >> and
> >> it failed so I replaced libressl after the test.
> >> Removing ccache (and with libressl) made no difference - still the
> >> same
> >> error. I then removed the libressl dependency and I reran the build -
> >> no
> >> ccache and no libressl - the build still failed, same error message.
> >> Since no-one else has reported this I suppose it should be something
> >> in
> >> my environment . . . . But what? :)
> >
> > While the build error *with libressl* is known, matching that reported
> > in:
> >
> > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=255241
> >
> > Failing to build without libressl is unexpected.
> >
> > Jump into the poudriere jail to confirm (or not) its libressl that's
> > being installed and used. Could be:
> >
> > - Custom WRKDIRPREFIX?
> > - An overriding jail or set (-z) specific poudriere foo-make.conf?
> >
> > Note, the OP build log contains:
> >
> > [pkg.home.santos-wright.net] | | `-- Installing libressl-3.3.3...
> > [pkg.home.santos-wright.net] | | `-- Extracting libressl-3.3.3:
> >
> > With defaults, base openssl will be used, and you wont see libressl as
> > a dependency in the build.
> >
> > [1] build/temp.freebsd-13.0-RELEASE-amd64-3.8/_openssl.c:2172:19:
> > error:
> > expected identifier or '('
> > static const long SSL_OP_NO_DTLSv1 = 0;
> >
> >> Regards,
> >> Simon.
> >>
> >>>>
> >>>>
> >>>> make.conf
> >>>>
> >>>> WRKDIRPREFIX=/usr/tmp
> >>>> OPTIONS_SET=GECKO CUPS
> >>>> NOI4B=1
> >>>> OPTIONS_SET+=NO-X11
> >>>> CUPS_OVERWRITE_BASE=YES
> >>>> WITH_VIM_OPTIONS=yes
> >>>> DEFAULT_VERSIONS+=ssl=libressl bdb=5
> >>>> VALID_CATEGORIES+=local
> >>>> SVN=svnlite
> >>>>
> >>>>
> >>>> Poudriere ports list:
> >>>>
> >>>> devel/git@lite
> >>>> dns/bind916
> >>>> emulators/open-vm-tools@nox11
> >>>> ftp/curl
> >>>> graphics/cairo
> >>>> local/vmserver-baseline
> >>>> mail/postfix
> >>>> print/hplip
> >>>> sysutils/facter
> >>>> sysutils/puppet6
> >>>> sysutils/rsyslog8
> >>>> sysutils/rubygem-puppetserver-ca
> >>>> www/apt-cacher-ng
> >>>> www/mod_log_sql
> >>>>
> >>>>
> > ___
> > freebsd-ports@freebsd.org mailing list
> > https://lists.freebsd.org/mailman/listinfo/freebsd-ports
> > To unsubscribe, send any mail to
> > "freebsd-ports-unsubscr...@freebsd.org"
> ___
> freebsd-ports@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-ports
> To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
>
--
Dan Langille
d...@langille.org
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
homepage for audio/faac
At present, the home page for audio/faac is set to a domain which is no longer relevant to the port. Should we set that to https://github.com/knik0/faac in pkg-descr? -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: FreeBSD Port: net/mosquitto pidfile or pid_file
On Fri, Mar 19, 2021, at 4:14 PM, Dan Langille wrote:
> Joe,
>
> I updated my mosquitto a few days ago and today it was restarted, but
> won't start. I think it's related to the update.
>
> Mar 14 01:47:40 supernews pkg[40249]: mosquitto upgraded: 1.6.7_1 -> 2.0.8
>
> The rc.d sript for net/mosquitto greps for pidfile.
>
> re:
> https://svnweb.freebsd.org/ports/head/net/mosquitto/files/mosquitto.in?revision=481153=markup#l30
>
> The documentation refers to pid_file, not pidfile.
>
> re: http://mosquitto.org/man/mosquitto-conf-5.html
>
> I think the rc.d script needs an update. I'd help with a patch, but I
> can't get a working solution.
>
> $ grep pid /usr/local/etc/rc.d/mosquitto
> pidfile=$(grep pid_file ${mosquitto_config} | awk '{print($2)}')
> echo "pidfile = '$pidfile'"
> rm -f $pidfile
>
> $ sudo service mosquitto start
> pidfile = '/var/run/mosquitto.pid'
> Starting mosquitto.
>
> $ ps auwwx | grep mos
> dan20215 0.0 0.0 11404 2260 1 S+ 20:12 0:00.00 grep
> mos
> $
>
>
>
> $ sudo /usr/local/sbin/mosquitto -c /usr/local/etc/mosquitto/mosquitto.conf -v
> 1616184446: Error: Unable to write pid file.
>
> $ ls -ld /var/run
> drwxr-xr-x 15 root wheel 1024 Mar 19 20:12 /var/run
>
> $ ls -l /var/run/mosquitto.pid
> ls: /var/run/mosquitto.pid: No such file or directory
Solved. At least for me.
In addition to the rc.d file change:
$ sudo mkdir /var/run/mosquitto
$ sudo chown mosquitto:mosquitto /var/run/mosquitto
Adjust /usr/local/etc/mosquitto/mosquitto.conf to use this new pid file
location.
Alter passwd file permissions:
$ sudo chgrp mosquitto /usr/local/etc/mosquitto/mosquitto.passwd
Modify certificate permissions:
$ ls -l /us/usr/local/etc/ssl/example.org.key
-rw--- 1 root wheel 1675 Apr 11 2018 /usr/local/etc/ssl/example.org.key
$ sudo chgrp mosquitto /usr/local/etc/ssl/example.org.key
$ sudo chmod g+r /usr/local/etc/ssl/example.org.key
$ ls -l /usr/local/etc/ssl/example.org.key
-rw-r- 1 root mosquitto 1675 Apr 11 2018
/usr/local/etc/ssl/example.org.key
Why all this? It seems that mosquitto is now dropping privs before writing the
PID file and before reading the certificate and password files.
Hope this helps
--
Dan Langille
d...@langille.org
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
FreeBSD Port: net/mosquitto pidfile or pid_file
Joe,
I updated my mosquitto a few days ago and today it was restarted, but won't
start. I think it's related to the update.
Mar 14 01:47:40 supernews pkg[40249]: mosquitto upgraded: 1.6.7_1 -> 2.0.8
The rc.d sript for net/mosquitto greps for pidfile.
re:
https://svnweb.freebsd.org/ports/head/net/mosquitto/files/mosquitto.in?revision=481153=markup#l30
The documentation refers to pid_file, not pidfile.
re: http://mosquitto.org/man/mosquitto-conf-5.html
I think the rc.d script needs an update. I'd help with a patch, but I can't
get a working solution.
$ grep pid /usr/local/etc/rc.d/mosquitto
pidfile=$(grep pid_file ${mosquitto_config} | awk '{print($2)}')
echo "pidfile = '$pidfile'"
rm -f $pidfile
$ sudo service mosquitto start
pidfile = '/var/run/mosquitto.pid'
Starting mosquitto.
$ ps auwwx | grep mos
dan20215 0.0 0.0 11404 2260 1 S+ 20:12 0:00.00 grep mos
$
$ sudo /usr/local/sbin/mosquitto -c /usr/local/etc/mosquitto/mosquitto.conf -v
1616184446: Error: Unable to write pid file.
$ ls -ld /var/run
drwxr-xr-x 15 root wheel 1024 Mar 19 20:12 /var/run
$ ls -l /var/run/mosquitto.pid
ls: /var/run/mosquitto.pid: No such file or directory
—
Dan Langille
http://langille.org/
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: Understanding different port counts
> On Nov 12, 2020, at 4:26 PM, Moritz Schmitt wrote: > > Hi, > > I noticed a big difference between the number of ports on > freebsd.org/ports/ and on freshports.org. Currently, it's 33348 vs. > 41346. > > The freebsd.org's number equals roughly the number of lines of a current > INDEX, but how does FreshPorts count? Hello, I'm glad you asked, because both sites are wrong. I'm going to fix FreshPorts because it is including commits to ports on branches in that total. That's not right. When ignoring branches, the FreshPorts count is 28,759 The FreeBSD total might be based on INDEX-12. In that count, each flavor counts as a port. Here is some data from that file: * py27-atspi-2.38.0|/usr/ports/accessibility/py-atspi * py37-atspi-2.38.0|/usr/ports/accessibility/py-atspi FreshPorts treats that as one port, accessibility/py-atspi If you remove the 'duplicates' from INDEX-12, the count is 28,759 which is much closer to what FreshPorts believes the count it. It is also close to this count: [dan@pkg01:~/ports/head] $ grep -h SUBDIR */Makefile | wc -l 28756 I put more information than you wished to read in this blog post: https://news.freshports.org/2020/11/18/how-many-ports-are-there/ Hope that helps. — Dan Langille http://langille <http://langille/>.org/ ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: Removal of net-mgmt/cnagios4
> On Nov 17, 2020, at 8:12 PM, Jonathan Chen wrote: > > Hi, > > The removal of net-mgmt/cnagios4 needs to reflected in net-mgmt/Makefile > The line: >SUBDIR += cnagios4 > needs to be removed. Thank you. Fixed. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: Aggressive ports removal (was: svn commit: r546907 - head/x11-clocks/wmtime)
On Sat, Aug 29, 2020, at 8:50 PM, Adam Weinberger wrote: > Just a disclaimer: I’m speaking for myself here, not on behalf of portmgr@. > Also please keep in mind that I am perpetually, pathologically optimistic. > >> On Aug 29, 2020, at 18:01, Warner Losh wrote: >> >> >> Basically, it draws the community back in, even if it's just small ways, >> rather than having them be a pure consumer giving nothing back. Make it easy >> to adopt a port, and we'll grow our base of labor. Make it easy to remove >> ports that no one can be bothered to adopt and we shrink our workload. And >> we'll spend our time on the things that we know net us at least one user who >> will pledge to keep it going. > > This is a serious issue that we’ve been dealing with for a long time. I’ve > advocated very strongly for a script that automatically notifies the > maintainer (+/- ports@?) when a port is marked BROKEN and/or DEPRECATED. > Community notification is always a good thing, and it opens the door for > objections and discussion. Nobody has written such a script, but I would be > thrilled to help deploy such a script if someone writes it. In the meantime, FreshPorts has a 'Watch ports I maintain' report under 'Watch ports I maintain' re https://github.com/FreshPorts/freshports/issues/138 -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: mail/mailman v3?
On Fri, Apr 24, 2020, at 9:32 AM, Dan Langille wrote: > > On Apr 24, 2020, at 9:04 AM, Kurt Jaeger wrote: > > > > Hi! > > > >> With mail/mailman being Python 2.7 (which is end-of-life), and mailman 3 > >> being Python 3 compatible: > >> > >> Do you know of any plans to port Mailman 3? > > > > There's already a PR about that: > > > > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=225543 > > Now 2 years old I see. Good start. > > > The patch itself is fine, but we need run-tests. > > > > This means: If you want to help, > > - use that patch, > > - build mailman3, > > - and install it somewhere and > > - test all the use-cases that you can think of > > - then write some docs on how to move an existing mailman2 site > > to mailman3 > > I'm guessing that's over and above what I found at: > > https://docs.mailman3.org/en/latest/migration.html > > > > - and give ideas how to handle list archives > > *especially* keeping the URLs identical (!) > > I think the existing archives are static HTML. I have some archives dating > back to 1999: https://www.unixathome.org/adsl/ > > There might be some server-side rewrites or aliases to ensure that this > URL always works, before and after mailman3: > > https://lists.freebsd.org/pipermail/freebsd-ports/2020-April/118352.html > <https://lists.freebsd.org/pipermail/freebsd-ports/2020-April/118352.html> > > Let's compare 2 and 3 lists: > > https://mail.python.org/pipermail/mailman-users/ > <https://mail.python.org/pipermail/mailman-users/> > > https://lists.mailman3.org/archives/list/mailman-us...@mailman3.org/ > <https://lists.mailman3.org/archives/list/mailman-us...@mailman3.org/> > > They are distinct (ignoring the hostname differences) so keeping the > old alongside the new should be safe. > > > And, speaking as one of the postmaster@ team: > > As lists.freebsd.org uses mailman2, we need this! > > > > postmaster@ has not yet decided if we really want to move to mailman3, > > so we are open to other options. The mail archive is the biggest hurdle 8-( > > Yes, we can't lose those. I have my own archives to support. I see the mailman lists themselves are now on Mailman 3: https://mail.python.org/archives/list/mailman-annou...@python.org/thread/HHQN7V6NY7G5CTOSC3WBU7VXW5KEBGVO/ -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: mail/mailman v3?
> On Apr 24, 2020, at 9:04 AM, Kurt Jaeger wrote: > > Hi! > >> With mail/mailman being Python 2.7 (which is end-of-life), and mailman 3 >> being Python 3 compatible: >> >> Do you know of any plans to port Mailman 3? > > There's already a PR about that: > > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=225543 Now 2 years old I see. Good start. > The patch itself is fine, but we need run-tests. > > This means: If you want to help, > - use that patch, > - build mailman3, > - and install it somewhere and > - test all the use-cases that you can think of > - then write some docs on how to move an existing mailman2 site > to mailman3 I'm guessing that's over and above what I found at: https://docs.mailman3.org/en/latest/migration.html > - and give ideas how to handle list archives > *especially* keeping the URLs identical (!) I think the existing archives are static HTML. I have some archives dating back to 1999: https://www.unixathome.org/adsl/ There might be some server-side rewrites or aliases to ensure that this URL always works, before and after mailman3: https://lists.freebsd.org/pipermail/freebsd-ports/2020-April/118352.html <https://lists.freebsd.org/pipermail/freebsd-ports/2020-April/118352.html> Let's compare 2 and 3 lists: https://mail.python.org/pipermail/mailman-users/ <https://mail.python.org/pipermail/mailman-users/> https://lists.mailman3.org/archives/list/mailman-us...@mailman3.org/ <https://lists.mailman3.org/archives/list/mailman-us...@mailman3.org/> They are distinct (ignoring the hostname differences) so keeping the old alongside the new should be safe. > And, speaking as one of the postmaster@ team: > As lists.freebsd.org uses mailman2, we need this! > > postmaster@ has not yet decided if we really want to move to mailman3, > so we are open to other options. The mail archive is the biggest hurdle 8-( Yes, we can't lose those. I have my own archives to support. — Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
mail/mailman v3?
Hello Matthias, With mail/mailman being Python 2.7 (which is end-of-life), and mailman 3 being Python 3 compatible: Do you know of any plans to port Mailman 3? If you were planning to do that, I can help. Thank you. — Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: FreeBSD Port: devel/py-subversion - Python 3 flavor?
On Thu, Apr 23, 2020, at 2:09 PM, Lev Serebryakov wrote: > Hello Dan, > > Thursday, April 23, 2020, 7:03:10 PM, you wrote: > > Huh. There are no problems, as we now have "current" subversion and "lts" > one, version 1.9 was removed. And 1.14.x will be LTS, so BOTH ports versions > will be 1.14 with python3 support! I must misunderstand completely. So long as devel/py-subversion soon supports Python 3, I'm happy. > > > This sounds like two ports are required new & legacy, instead of flavors. > > > On Thu, Apr 23, 2020, at 11:37 AM, Lev Serebryakov wrote: > >> On 23.04.2020 17:43, Dan Langille wrote: > >> > >> subversion 1.14.0 which is RC2 now will be python3 only. > >> > >> But we are used to have old versions of subversion in ports tree, and > >> I'm not sure to do. > >> > > > -- > Best regards, > Levmailto:l...@freebsd.org > > ___ > freebsd-ports@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-ports > To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org" > -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: FreeBSD Port: devel/py-subversion - Python 3 flavor?
This sounds like two ports are required new & legacy, instead of flavors. On Thu, Apr 23, 2020, at 11:37 AM, Lev Serebryakov wrote: > On 23.04.2020 17:43, Dan Langille wrote: > > subversion 1.14.0 which is RC2 now will be python3 only. > > But we are used to have old versions of subversion in ports tree, and > I'm not sure to do. > > > Hello Lev, > > > > Do you know of anything preventing devel/py-subversion from working with > > Python 3? > > > > I ask because devel/py-subversion at present requires Python 2.7 which is > > past EOL. > > > > While I was typing this, I thought: I should just try it first. > > > > I failed: > > > > https://services.unixathome.org/poudriere/data/121amd64-dvl/2020-04-23_14h34m22s/logs/errors/py37-subversion-1.13.0.log > > > > ModuleNotFoundError: No module named 'libsvn._core' > > > > Ideas? > > > > Thank you. > > > > — > > Dan Langille > > http://langille.org/ > > > > > > > > > > > > > -- > // Lev Serebryakov > > > Attachments: > * signature.asc -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
FreeBSD Port: devel/py-subversion - Python 3 flavor?
Hello Lev, Do you know of anything preventing devel/py-subversion from working with Python 3? I ask because devel/py-subversion at present requires Python 2.7 which is past EOL. While I was typing this, I thought: I should just try it first. I failed: https://services.unixathome.org/poudriere/data/121amd64-dvl/2020-04-23_14h34m22s/logs/errors/py37-subversion-1.13.0.log ModuleNotFoundError: No module named 'libsvn._core' Ideas? Thank you. — Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: New bacula release
On Fri, Feb 28, 2020, at 10:43 AM, Andrea Venturoli wrote: > On 2020-02-28 16:37, Chris wrote: > > > > What revision is your ports(7) tree at? I mention this, because according > > the site that maintainer of bacula runs: > > > > https://www.freshports.org/search.php?query=bacula=name=match=100=excludedeleted=category=asc=Search > > > > > > > > indicates bacula is already at 9.6.2. :) > > Wow, guess "a few more days" went down to 6 minutes! > Congratulation and thanks! We use Bacula at $WORK so I could do that there... ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: New bacula release
On Fri, Feb 28, 2020, at 5:20 AM, Andrea Venturoli wrote: > Hello. > > Sorry to bother you, but I'm planning some upgrades of old Bacula > installation and I'd benefit from some info. > > While I was going to update to 9.4.x, I saw 9.6.2 is out. > Do you have any schedule for the FreeBSD port? > > Just to decide whether to wait for it or go ahead with 9.4 in the meantime. 9.6.2 was announced about 48 hours ago. It may take me a few more days to get it ported and tested. If all goes well.. Hope that helps. -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
FreeBSD ports: MASTER_SITES=GHC
Hello, MASTER_SITES= GHC is broken and is causing fetch errors. This appears to be a GitHub issue which will not be resolved. Usually the fix is to use GH not GHC. See https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=242268 The ports affected are: converters/lua-iconv/Makefile:MAINTAINER= vani...@freebsd.org converters/lua51-iconv/Makefile:MAINTAINER= f...@freebsd.org deskutils/devd-notifier/Makefile:MAINTAINER=f.ungl...@googlemail.com devel/libcli/Makefile:MAINTAINER= po...@freebsd.org graphics/glosm/Makefile:MAINTAINER= amd...@freebsd.org java/netty-tcnative/Makefile:MAINTAINER=d...@freebsd.org lang/rhino/Makefile:MAINTAINER= ni...@freebsd.org mail/libsieve/Makefile:MAINTAINER= po...@freebsd.org math/liblbfgs/Makefile:MAINTAINER= g...@hychen.org sysutils/bsdinfo/Makefile:MAINTAINER= freebsd-po...@samu.pl sysutils/confctl/Makefile:MAINTAINER= tr...@freebsd.org textproc/syck/Makefile:MAINTAINER= po...@freebsd.org www/ach/Makefile:MAINTAINER=dan...@freebsd.org www/mod_xsendfile/Makefile:MAINTAINER= potatosal...@gmail.com www/p5-Ark/Makefile:MAINTAINER= kuriy...@freebsd.org www/tinytinyhttpd/Makefile:MAINTAINER= smat...@karashi.org -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: Many ports recently marked BROKEN/unfetchable that aren't broken?
On Wed, Nov 27, 2019, at 12:26 PM, Dan Langille wrote: > On Wed, Nov 27, 2019, at 8:36 AM, Dan Langille wrote: > > This appears to be a DNS issue. > > > > [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ > > sudo make makesum > > ===> License BSD2CLAUSE accepted by the user > > ===> License BSD2CLAUSE accepted by the user > > ===> zfs-stats-1.2.2_1 depends on file: /usr/local/sbin/pkg - found > > => zfs-stats-1.2.2.tar.gz doesn't seem to exist in > > /usr/local/poudriere/ports/default/distfiles/. > > => Attempting to fetch > > https://cloud.github.com/downloads/mmatuska/zfs-stats/zfs-stats-1.2.2.tar.gz > > fetch: > > https://cloud.github.com/downloads/mmatuska/zfs-stats/zfs-stats-1.2.2.tar.gz: > > No address record > > => Attempting to fetch > > http://distcache.FreeBSD.org/ports-distfiles/zfs-stats-1.2.2.tar.gz > > zfs-stats-1.2.2.tar.gz 9 kB 36 MBps > >00s > > ===> Fetching all distfiles required by zfs-stats-1.2.2_1 for building > > > > > > [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ > > host cloud.github.com > > cloud.github.com is an alias for d24z2fz21y4fag.cloudfront.net. > > > > [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ > > host d24z2fz21y4fag.cloudfront.net > > [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ > > > Further discussions via IRC indicate that we should stop using GHC because > GitHub isn't going to do anything about this issue. > > cloud.github.com isn't 'public', it's backend. I have opened this PR: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=242268 -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: Many ports recently marked BROKEN/unfetchable that aren't broken?
On Wed, Nov 27, 2019, at 8:36 AM, Dan Langille wrote: > On Fri, Nov 15, 2019, at 10:44 AM, Matthew Seaman wrote: > > On 15/11/2019 15:25, Nick Rogers wrote: > > > For me the affected ports were sysutils/pftop, sysutils/zfs-stats, and > > > sysutils/stress, which worked just fine before this commit. Apologies if I > > > am missing something, but it seems like they are not actually broken? > > > > If you were building these ports previously, then you probably have the > > distfiles in your local distfiles cache already. That will give the > > impression that everything is fine if you just try and build the port. > > If you didn't have a cached copy of the distfiles, then you'ld probably > > have a rather different experience. > > > > Although if you can show that the distfiles are now fetchable, eg. by: > > > > make distclean > > make fetch > > > > then by all means please do report it. A PR with patches is preferred, > > but any reasonable mechanism for bringing it to the attention of > > developers is acceptable. > > This appears to be a DNS issue. > > [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ > sudo make makesum > ===> License BSD2CLAUSE accepted by the user > ===> License BSD2CLAUSE accepted by the user > ===> zfs-stats-1.2.2_1 depends on file: /usr/local/sbin/pkg - found > => zfs-stats-1.2.2.tar.gz doesn't seem to exist in > /usr/local/poudriere/ports/default/distfiles/. > => Attempting to fetch > https://cloud.github.com/downloads/mmatuska/zfs-stats/zfs-stats-1.2.2.tar.gz > fetch: > https://cloud.github.com/downloads/mmatuska/zfs-stats/zfs-stats-1.2.2.tar.gz: > No address record > => Attempting to fetch > http://distcache.FreeBSD.org/ports-distfiles/zfs-stats-1.2.2.tar.gz > zfs-stats-1.2.2.tar.gz 9 kB 36 MBps >00s > ===> Fetching all distfiles required by zfs-stats-1.2.2_1 for building > > > [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ > host cloud.github.com > cloud.github.com is an alias for d24z2fz21y4fag.cloudfront.net. > > [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ > host d24z2fz21y4fag.cloudfront.net > [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ Further discussions via IRC indicate that we should stop using GHC because GitHub isn't going to do anything about this issue. cloud.github.com isn't 'public', it's backend. -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: Many ports recently marked BROKEN/unfetchable that aren't broken?
On Fri, Nov 15, 2019, at 10:44 AM, Matthew Seaman wrote: > On 15/11/2019 15:25, Nick Rogers wrote: > > For me the affected ports were sysutils/pftop, sysutils/zfs-stats, and > > sysutils/stress, which worked just fine before this commit. Apologies if I > > am missing something, but it seems like they are not actually broken? > > If you were building these ports previously, then you probably have the > distfiles in your local distfiles cache already. That will give the > impression that everything is fine if you just try and build the port. > If you didn't have a cached copy of the distfiles, then you'ld probably > have a rather different experience. > > Although if you can show that the distfiles are now fetchable, eg. by: > > make distclean > make fetch > > then by all means please do report it. A PR with patches is preferred, > but any reasonable mechanism for bringing it to the attention of > developers is acceptable. This appears to be a DNS issue. [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ sudo make makesum ===> License BSD2CLAUSE accepted by the user ===> License BSD2CLAUSE accepted by the user ===> zfs-stats-1.2.2_1 depends on file: /usr/local/sbin/pkg - found => zfs-stats-1.2.2.tar.gz doesn't seem to exist in /usr/local/poudriere/ports/default/distfiles/. => Attempting to fetch https://cloud.github.com/downloads/mmatuska/zfs-stats/zfs-stats-1.2.2.tar.gz fetch: https://cloud.github.com/downloads/mmatuska/zfs-stats/zfs-stats-1.2.2.tar.gz: No address record => Attempting to fetch http://distcache.FreeBSD.org/ports-distfiles/zfs-stats-1.2.2.tar.gz zfs-stats-1.2.2.tar.gz 9 kB 36 MBps00s ===> Fetching all distfiles required by zfs-stats-1.2.2_1 for building [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ host cloud.github.com cloud.github.com is an alias for d24z2fz21y4fag.cloudfront.net. [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ host d24z2fz21y4fag.cloudfront.net [dan@pkg01:/usr/local/poudriere/ports/default/sysutils/zfs-stats] $ -- Dan Langille d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: sysutils/iocage: recent update to 1.2 removed all commands except migrate
Hello,
I think the port no longer requires dulwich:
The release nodes mention dulwich:
https://github.com/iocage/iocage/releases/tag/1.2
and references this commit:
https://github.com/iocage/iocage/commit/c7110e195261430dbdbd41dfa935ab706fe13c7b
--
Dan Langille
d...@langille.org
On Sun, Sep 29, 2019, at 10:52 PM, Marcelo Araujo wrote:
> Hello all,
>
> Thanks for the prompt report about this issue, the fix is here:
> https://svnweb.freebsd.org/changeset/ports/513279
> Please, let me know if there is anything broken at your side.
>
> Best Regards,
>
> Em seg, 30 de set de 2019 às 03:18, Dan Langille escreveu:
>> On Sun, Sep 29, 2019, at 12:59 PM, Rebecca Cran wrote:
>> > I just upgraded to iocage version 1.2 and now I don't have any commands
>> > other than "migrate":
>> >
>> > root@muon:/home/bcran # iocage
>> > Usage: iocage [OPTIONS] COMMAND [ARGS]...
>> >
>> > A jail manager.
>> >
>> > Options:
>> > -v, --version Display iocage's version and exit.
>> > -f, --force Allow iocage to rename datasets.
>> > -D, --debug Log debug output to the console.
>> > --help Show this message and exit.
>> >
>> > Commands:
>> > migrate Migrate all iocage_legacy develop basejails to new clonejails.
>> >
>> >
>> > It looks to be the same issue as reported in
>> > https://github.com/iocage/iocage/issues/894 and
>> > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=239759 .
>>
>>
>> Comparing sysutils/iocage and sysutils/iocage-devel I notice these are
>> missing from the former:
>>
>>
>> ${PYTHON_PKGNAMEPREFIX}GitPython>=2.1.10:devel/py-gitpython@${PY_FLAVOR} \
>> git-lite>0:devel/git-lite
>>
>>
>> Also, I am sure that this is no longer required:
>>
>> ${PYTHON_PKGNAMEPREFIX}dulwich>0:devel/dulwich@${PY_FLAVOR} \
>>
>>
>> --
>> Dan Langille
>> d...@langille.org
>>
>
>
> --
>
> --
Marcelo Araujo(__)
> ara...@freebsd.org \\\'',)
> http://www.FreeBSD.org <http://www.freebsd.org/> \/ \ ^
Power To Server. .\. /_)
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: sysutils/iocage: recent update to 1.2 removed all commands except migrate
On Sun, Sep 29, 2019, at 12:59 PM, Rebecca Cran wrote:
> I just upgraded to iocage version 1.2 and now I don't have any commands
> other than "migrate":
>
> root@muon:/home/bcran # iocage
> Usage: iocage [OPTIONS] COMMAND [ARGS]...
>
> A jail manager.
>
> Options:
> -v, --version Display iocage's version and exit.
> -f, --force Allow iocage to rename datasets.
> -D, --debug Log debug output to the console.
> --help Show this message and exit.
>
> Commands:
> migrate Migrate all iocage_legacy develop basejails to new clonejails.
>
>
> It looks to be the same issue as reported in
> https://github.com/iocage/iocage/issues/894 and
> https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=239759 .
Comparing sysutils/iocage and sysutils/iocage-devel I notice these are missing
from the former:
${PYTHON_PKGNAMEPREFIX}GitPython>=2.1.10:devel/py-gitpython@${PY_FLAVOR} \
git-lite>0:devel/git-lite
Also, I am sure that this is no longer required:
${PYTHON_PKGNAMEPREFIX}dulwich>0:devel/dulwich@${PY_FLAVOR} \
--
Dan Langille
d...@langille.org
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
www/privatebin DATADBDIR not defined
Hello,
Post install, index.php contains: define('DATADBDIR', '%% DATADBDIR%%');
I think the intention was to define DATADBDIR and perform a SUB_LIST operation.
$ grep -r DATADBDIR *
Makefile: @${REINPLACE_CMD} -e 's:%%DATADBDIR%%:/var/db/${PORTNAME}:' \
files/patch-index.php:+define('DATADBDIR', '%%DATADBDIR%%');
files/patch-cfg_conf.sample.php:+dir = DATADBDIR
files/patch-cfg_conf.sample.php:+dir = DATADBDIR
files/patch-cfg_conf.sample.php:+dir = DATADBDIR
files/patch-cfg_conf.sample.php:+;dsn = "sqlite:" DATADBDIR "/db.sq3"
Is the Makefile missing a 'DATADBDIR= privatebin' for example?
—
Dan Langille
http://langille.org/
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: depends blacklist qa check
> On Apr 8, 2019, at 4:38 PM, Mathieu Arnold wrote: > > On Mon, Apr 08, 2019 at 01:55:19PM -0400, Dan Langille wrote: >> re: https://reviews.freebsd.org/D19562 >> >> I understand the purpose of the blacklist is to restrict the use of certain >> meta ports to end-users only. That is, the meta-ports in question should >> not be used as a depends in the FreeBSD ports tree. >> >> For example, do not use lang/python as a depends. Let the user install that >> install. >> >> With my user hat on, who has a in-house meta-port which includes >> lang/python2 as a depends, because we want it, and list of other stuff >> installed. >> >> We can't do that now. Our meta-port is broken and can no longer be used. Is >> there another way? The list appears hardcoded and no way for the end user to >> configure. > > Well, that is not correct. QA checks are only run when using poudriere > testport or bulk -t, or when DEVELOPER=yes. So, while it is true that > such meta port will no longer pass testing, it will not impact a normal > poudriere bulk. That's good news. Yes, I discovered this on a testport, while amending an in-house port. Phew. Sorry for the noise. This got me completely kerfuffled. cheers. signature.asc Description: Message signed with OpenPGP
depends blacklist qa check
re: https://reviews.freebsd.org/D19562 I understand the purpose of the blacklist is to restrict the use of certain meta ports to end-users only. That is, the meta-ports in question should not be used as a depends in the FreeBSD ports tree. For example, do not use lang/python as a depends. Let the user install that install. With my user hat on, who has a in-house meta-port which includes lang/python2 as a depends, because we want it, and list of other stuff installed. We can't do that now. Our meta-port is broken and can no longer be used. Is there another way? The list appears hardcoded and no way for the end user to configure. Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: https://www.freshports.org/ is down
> On Feb 28, 2019, at 7:01 PM, Yuri wrote: > > It resolves to the IP, and pings work, but it can't be contacted on 443 or 80. I only just saw this email, while looking for other things. I have details: it was a software error. The site was offline from about 630 PM EST until about 9:20 AM the next day. This is the FreshPorts status page: https://freshports.wordpress.com/ This is the Twitter account: https://twitter.com/freshports Thank you. -- Dan Langille - Freshp\Ports & BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
PORTREVISION bump on lang/perl5.26
Mat, I think we should do a PORTREVISION bump on lang/perl5.26 because of DEPRECATED. 5.26.3 came in on 4 Dec with https://svnweb.freebsd.org/ports?view=revision=486599 DEPRECATED was commented out. DEPRECATED came back in on 17 Dec with https://svnweb.freebsd.org/ports/head/lang/perl5.26/Makefile?r1=487685=487686; This affects me, and probably others. I have multiple hosts with perl5-5.26.3, some deprecated, some not, depending on when perl was built. [dan@pg11:~] $ pkg annotate -a -S expiration_date perl5-5.26.3: Tag: expiration_date Value: 2020-05-31 [dan@pg10:~] $ pkg annotate -a -S expiration_date [dan@pg10:~] $ pkg info -x perl perl5-5.26.3 [dan@pg10:~] $ I'm happy to do the bump. Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: dns/bind911 - statistics-file
> On Jan 20, 2019, at 5:42 AM, Mathieu Arnold wrote: > > On Sat, Jan 19, 2019 at 07:50:45PM -0500, Dan Langille wrote: >> Mat, >> >> I encountered an odd situation where my stats file kept changing >> permissions. With every reinstall of bind911, >> the permissions on var/run/named/stats change to chown root:bind which >> prevents bind from updating the file. >> >> This is what I need: >> >> $ ls -l /var/run/named/stats >> -rw-r--r-- 1 bind bind 11507 Jan 20 00:45 /var/run/named/stats >> >> Could that change be carried out by this file? >> >> >> https://svnweb.freebsd.org/ports/head/dns/bind911/files/BIND.chroot.dist?view=markup#l24 >> >> I don't see a reference to /var/run/named/stats in BIND.chroot.dist but >> can't help but wonder if it's something similar. >> >> I have been using these options: >> >>directory "/usr/local/etc/namedb/working"; >>pid-file"/var/run/named/pid"; >>dump-file "/var/dump/named_dump.db"; >>statistics-file "/var/run/named/stats"; >>zone-statistics yes; >> >> When researching this tonight, I noticed the sample configuration uses >> /var/run/named.stats. Perhaps I'm doing this wrong. >> I am happy to change my configuration, but first I write in case the script >> is doing something unexpected. > > I do not think anything in the BIND9 ports would change the file permissions. > > The mtree file only touches the directories to make sure they have the > correct permissions, so it is not it. Moreover the mtree file is ONLY > used when using named_chrootdir to chroot named, which does not appear > to be your case. > The BIND9 ports have not had a pkg-install script for years, so it's not > it either. > The rc file does not chown anything, so it's not it doing it either. > > Side note, the sample configuration uses /var/stats/named.stats, not > /var/run/named.stats. And it was ever since it was added to the base > system named.conf file back in 2004 (in src r135918). > > So I'd say something else on your system "fixes" the file's permissions. This is it: [dan@toiler:~] $ grep cleanvar /etc/defaults/rc.conf cleanvar_enable="YES" # Clean the /var directory That clears the file, then bind creates it chown root:bind. Why it creates it like that, I don't know yet, but that's outside scope of this post. Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
Re: dns/bind911 - statistics-file
> On Jan 20, 2019, at 5:42 AM, Mathieu Arnold wrote: > > On Sat, Jan 19, 2019 at 07:50:45PM -0500, Dan Langille wrote: >> Mat, >> >> I encountered an odd situation where my stats file kept changing >> permissions. With every reinstall of bind911, >> the permissions on var/run/named/stats change to chown root:bind which >> prevents bind from updating the file. >> >> This is what I need: >> >> $ ls -l /var/run/named/stats >> -rw-r--r-- 1 bind bind 11507 Jan 20 00:45 /var/run/named/stats >> >> Could that change be carried out by this file? >> >> >> https://svnweb.freebsd.org/ports/head/dns/bind911/files/BIND.chroot.dist?view=markup#l24 >> >> I don't see a reference to /var/run/named/stats in BIND.chroot.dist but >> can't help but wonder if it's something similar. >> >> I have been using these options: >> >>directory "/usr/local/etc/namedb/working"; >>pid-file"/var/run/named/pid"; >>dump-file "/var/dump/named_dump.db"; >>statistics-file "/var/run/named/stats"; >>zone-statistics yes; >> >> When researching this tonight, I noticed the sample configuration uses >> /var/run/named.stats. Perhaps I'm doing this wrong. >> I am happy to change my configuration, but first I write in case the script >> is doing something unexpected. > > I do not think anything in the BIND9 ports would change the file permissions. > > The mtree file only touches the directories to make sure they have the > correct permissions, so it is not it. Moreover the mtree file is ONLY > used when using named_chrootdir to chroot named, which does not appear > to be your case. > The BIND9 ports have not had a pkg-install script for years, so it's not > it either. > The rc file does not chown anything, so it's not it doing it either. > > Side note, the sample configuration uses /var/stats/named.stats, not > /var/run/named.stats. And it was ever since it was added to the base > system named.conf file back in 2004 (in src r135918). Noted. Thank you. > So I'd say something else on your system "fixes" the file's permissions. It seems to be upon system start up. I did a test in a jail. # From the host, verify the file in the jail exists: [dan@knew:~] $ ls -l /iocage/jails/toiler/root/var/run/named/stats -rw-r--r-- 1 bind bind 1 Jan 20 16:30 /iocage/jails/toiler/root/var/run/named/stats [dan@knew:~] $ sudo iocage stop toiler * Stopping toiler + Running prestop OK + Stopping services OK + Refusing to remove protected devfs_ruleset: 7 + Removing jail process OK + Running poststop OK [dan@knew:~] $ ls -l /iocage/jails/toiler/root/var/run/named/stats -rw-r--r-- 1 bind bind 1 Jan 20 16:30 /iocage/jails/toiler/root/var/run/named/stats [dan@knew:~] $ sudo iocage start toiler * Starting toiler + Started OK + Using devfs_ruleset: 7 + Starting services OK [dan@knew:~] $ ls -l /iocage/jails/toiler/root/var/run/named/stats ls: /iocage/jails/toiler/root/var/run/named/stats: No such file or directory [dan@knew:~] $ The file soon gets created: [dan@knew:~] $ ls -l /iocage/jails/toiler/root/var/run/named/stats -rw-r--r-- 1 root bind 1 Jan 20 16:35 /iocage/jails/toiler/root/var/run/named/stats Presumably by this: 20-Jan-2019 16:35:12.819 received control channel command 'stats' 20-Jan-2019 16:35:12.819 could not open statistics dump file '/var/run/named/stats': permission denied 20-Jan-2019 16:35:12.819 dumpstats failed: permission denied I will keep looking. Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
dns/bind911 - statistics-file
Mat, I encountered an odd situation where my stats file kept changing permissions. With every reinstall of bind911, the permissions on var/run/named/stats change to chown root:bind which prevents bind from updating the file. This is what I need: $ ls -l /var/run/named/stats -rw-r--r-- 1 bind bind 11507 Jan 20 00:45 /var/run/named/stats Could that change be carried out by this file? https://svnweb.freebsd.org/ports/head/dns/bind911/files/BIND.chroot.dist?view=markup#l24 I don't see a reference to /var/run/named/stats in BIND.chroot.dist but can't help but wonder if it's something similar. I have been using these options: directory "/usr/local/etc/namedb/working"; pid-file"/var/run/named/pid"; dump-file "/var/dump/named_dump.db"; statistics-file "/var/run/named/stats"; zone-statistics yes; When researching this tonight, I noticed the sample configuration uses /var/run/named.stats. Perhaps I'm doing this wrong. I am happy to change my configuration, but first I write in case the script is doing something unexpected. Ideas welcome. Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
Re: FreshPorts - CONFLICTS
> On Oct 5, 2018, at 10:56 PM, Dan Langille wrote: > > A short while ago I added CONFLICTS to FreshPorts (see > https://news.freshports.org/2018/09/26/adding-and-populating-conflicts/). > > Tonight I added conflict matches to https://dev.freshports.org so if you > know a port with CONFLICTS please have a look. > > For example: > > https://dev.freshports.org/security/openssl/ conflicts with > > * libressl > * libressl-devel > * openssl-devel > * openssl111 > > > https://dev.freshports.org/sysutils/bacula-server conflicts with: > > * bacula5-server > > Thank you. I have feedback from one developer saying all good. Thanks. I ask each port maintainer to review this list and see if their ports give expected results. At this page you will find your port with a conflict, and the ports it conflicts with. I just realized this page does not distinguish between CONFLICTS, CONFLICTS_BUILD, and CONFLICTS_INSTALL, but I will do that soon https://dev.freshports.org/ports-conflicts.html <https://dev.freshports.org/ports-conflicts.html> If results are correct, no need to reply to me. If the results are incorrect, please reply offlist to d...@freebsd.org <mailto:d...@freebsd.org> Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
FreshPorts - CONFLICTS
A short while ago I added CONFLICTS to FreshPorts (see https://news.freshports.org/2018/09/26/adding-and-populating-conflicts/). Tonight I added conflict matches to https://dev.freshports.org so if you know a port with CONFLICTS please have a look. For example: https://dev.freshports.org/security/openssl/ conflicts with * libressl * libressl-devel * openssl-devel * openssl111 https://dev.freshports.org/sysutils/bacula-server conflicts with: * bacula5-server Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
Re: Bacula 9.2.1 fails on 10.4:
> On Aug 29, 2018, at 10:59 AM, Jason E. Hale wrote: > > On Mon, Aug 27, 2018 at 7:42 PM Dan Langille wrote: >> >>> On Aug 27, 2018, at 7:26 PM, Yuri Pankov wrote: >>> >>> Dan Langille wrote: >>>> Why would Bacula 9.2.1 compile on 11.2 but fail on 10.4? >>>> The error is: >>>> bsock.c:439:20: error: use of undeclared identifier 'ENODATA' >>>> The complete build logs are at the following URLs. Can you see the cause. >>>> 11.2: >>>> https://services.unixathome.org/poudriere/data/112amd64-default/2018-08-27_21h15m53s/logs/bacula9-client-9.2.1.log >>>> 10.4: >>>> https://services.unixathome.org/poudriere/data/104amd64-default/2018-08-27_21h43m31s/logs/errors/bacula9-client-9.2.1.log >>>> It doesn't make any sense to me. Thanks. >>> >>> I'd say it's libc++ missing its errno.h having ENODATA defined if the >>> following is true: >>> >>> - both builds are using clang++ >>> - both builds are using libc++ >>> >>> That header defining ENODATA exists in 11.2 and doesn't exist in 10.4 >>> (contrib/libc++/include/errno.h). >> >> What is a decent solution? Patch upstream? Patch the port? >> >> Thank you. >> > > You can patch the port for now and submit it upstream. I think they have patched upstream already, just not released it: http://www.bacula.org/git/cgit.cgi/bacula/commit/?h=Branch-9.2=eea817f035d32221ab46d1a4f95193cd4d6dbda7 > Looks like they > had ENODATA defined as EPIPE and missed it when they refactored: > http://www.bacula.org/git/cgit.cgi/bacula/commit/bacula/src/lib/bsock.c?h=Release-9.2.1=c3c124adbe784a55bb31c66401630b2d22768baf. > Now that code is in: > http://www.bacula.org/git/cgit.cgi/bacula/tree/bacula/src/lib/bsockcore.c?h=Release-9.2.1#n44, > but it needs to be in both files. > I think it would make more sense to factor it out to src/baconfig.h. > That way, they could remove the #if block from bsockcore.c. I've > attached a patch and put it in plain text below in case it gets > scrubbed. Thank you Jason, this is very helpful. I will pass this upsteam, and in the meantime, fix the port. cheers. > > Cheers, > Jason > > Index: sysutils/bacula9-server/files/patch-src_baconfig.h > === > --- sysutils/bacula9-server/files/patch-src_baconfig.h (nonexistent) > +++ sysutils/bacula9-server/files/patch-src_baconfig.h (working copy) > @@ -0,0 +1,13 @@ > +--- src/baconfig.h.orig2018-08-29 14:09:36 UTC > src/baconfig.h > +@@ -38,6 +38,10 @@ > + #define have_tls 0 > + #endif > + > ++#ifndef ENODATA > ++#define ENODATA EPIPE > ++#endif > ++ > + #ifndef ETIME > + #define ETIME ETIMEDOUT > + #endif > > Property changes on: sysutils/bacula9-server/files/patch-src_baconfig.h > ___ > Added: fbsd:nokeywords > ## -0,0 +1 ## > +yes > \ No newline at end of property > Added: svn:eol-style > ## -0,0 +1 ## > +native > \ No newline at end of property > Added: svn:mime-type > ## -0,0 +1 ## > +text/plain > \ No newline at end of property > -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: Bacula 9.2.1 fails on 10.4:
> On Aug 27, 2018, at 7:26 PM, Yuri Pankov wrote: > > Dan Langille wrote: >> Why would Bacula 9.2.1 compile on 11.2 but fail on 10.4? >> The error is: >> bsock.c:439:20: error: use of undeclared identifier 'ENODATA' >> The complete build logs are at the following URLs. Can you see the cause. >> 11.2: >> https://services.unixathome.org/poudriere/data/112amd64-default/2018-08-27_21h15m53s/logs/bacula9-client-9.2.1.log >> 10.4: >> https://services.unixathome.org/poudriere/data/104amd64-default/2018-08-27_21h43m31s/logs/errors/bacula9-client-9.2.1.log >> It doesn't make any sense to me. Thanks. > > I'd say it's libc++ missing its errno.h having ENODATA defined if the > following is true: > > - both builds are using clang++ > - both builds are using libc++ > > That header defining ENODATA exists in 11.2 and doesn't exist in 10.4 > (contrib/libc++/include/errno.h). What is a decent solution? Patch upstream? Patch the port? Thank you. -- Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Bacula 9.2.1 fails on 10.4:
Why would Bacula 9.2.1 compile on 11.2 but fail on 10.4? The error is: bsock.c:439:20: error: use of undeclared identifier 'ENODATA' The complete build logs are at the following URLs. Can you see the cause. 11.2: https://services.unixathome.org/poudriere/data/112amd64-default/2018-08-27_21h15m53s/logs/bacula9-client-9.2.1.log 10.4: https://services.unixathome.org/poudriere/data/104amd64-default/2018-08-27_21h43m31s/logs/errors/bacula9-client-9.2.1.log It doesn't make any sense to me. Thanks. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: svn commit: r475644 - head/databases/mantis
> On Aug 24, 2018, at 5:56 PM, Tobias Kortkamp wrote: > > On Mon, Aug 20, 2018, at 22:49, Dan Langille wrote: >>> On Jul 29, 2018, at 6:51 AM, Jochen Neumeister wrote: >>> >>> Author: joneum >>> Date: Sun Jul 29 10:51:37 2018 >>> New Revision: 475644 >>> URL: https://svnweb.freebsd.org/changeset/ports/475644 >>> >>> Log: >>> databases/mantis: Update to 2.15.0 >>> >>> Changelog: >>> https://mantisbt.org/bugs/changelog_page.php?project=mantisbt=2.15.0 >>> >>> PR: 229880 >>> Submitted by: Nathan >>> MFH:2018Q3 >>> Security: 0822a4cf-9318-11e8-8d88-00e04c1ea73d >>> >>> Modified: >>> head/databases/mantis/Makefile >>> head/databases/mantis/distinfo >> >> This version fails to run for me. The vendor directory is missing. >> This is the error I get: >> >> [Mon Aug 20 20:40:14.528249 2018] [:error] [pid 73365] [client redacted] >> PHP Warning: require_once(/usr/local/www/mantis/vendor/autoload.php): >> failed to open stream: No such file or directory in /usr/local/www/ >> mantis/core.php on line 71 >> [Mon Aug 20 20:40:14.528317 2018] [:error] [pid 73365] [client redacted] >> PHP Fatal error: require_once(): Failed opening required '/usr/local/ >> www/mantis/vendor/autoload.php' (include_path='/usr/local/share/pear') >> in /usr/local/www/mantis/core.php on line 71 >> >> Is this working for anyone? >> >> There was no vendor directory in the 2.6.0 tarball. There isn't one in >> the 2.15.0 tarball either, but the code definitely references it. >> >> Upstream inquires >> (https://mantisbt.org/forums/viewtopic.php?p=66190#p66190 >> <https://mantisbt.org/forums/viewtopic.php?p=66190#p66190>) indicate we >> should be using Sourceforge downloads, not GitHub >> >> The vendor directory is present in the SF tarball at >> https://sourceforge.net/projects/mantisbt/files/mantis-stable/2.16.0/ >> <https://sourceforge.net/projects/mantisbt/files/mantis-stable/2.16.0/> >> >> I tried last night to update the port, but failed. >> >> Ideas please? > > Hi, > > could you test the update to 2.16.0 on https://reviews.freebsd.org/D16890 > <https://reviews.freebsd.org/D16890> ? > I've switched it over to the SF tarball and it seems to work ok > for me, but would appreciate a confirmation from someone else that > it works before committing it. > > Thanks! The patch does not apply cleanly to the Makefile for me, so I applied it manually. It builds, it runs. A patch against head is belowl I create diffs against head and uploaded them here: https://gist.github.com/dlangille/a9de194209795a4261eebbd3de65e841 <https://gist.github.com/dlangille/a9de194209795a4261eebbd3de65e841> One is for the Makefile, the other is for the whole port. Hope this helps. Thank you, your work got me running again. I look forward to the commit. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
Re: svn commit: r475644 - head/databases/mantis
> On Jul 29, 2018, at 6:51 AM, Jochen Neumeister wrote: > > Author: joneum > Date: Sun Jul 29 10:51:37 2018 > New Revision: 475644 > URL: https://svnweb.freebsd.org/changeset/ports/475644 > > Log: > databases/mantis: Update to 2.15.0 > > Changelog: > https://mantisbt.org/bugs/changelog_page.php?project=mantisbt=2.15.0 > > PR: 229880 > Submitted by:Nathan > MFH: 2018Q3 > Security:0822a4cf-9318-11e8-8d88-00e04c1ea73d > > Modified: > head/databases/mantis/Makefile > head/databases/mantis/distinfo This version fails to run for me. The vendor directory is missing. This is the error I get: [Mon Aug 20 20:40:14.528249 2018] [:error] [pid 73365] [client redacted] PHP Warning: require_once(/usr/local/www/mantis/vendor/autoload.php): failed to open stream: No such file or directory in /usr/local/www/mantis/core.php on line 71 [Mon Aug 20 20:40:14.528317 2018] [:error] [pid 73365] [client redacted] PHP Fatal error: require_once(): Failed opening required '/usr/local/www/mantis/vendor/autoload.php' (include_path='/usr/local/share/pear') in /usr/local/www/mantis/core.php on line 71 Is this working for anyone? There was no vendor directory in the 2.6.0 tarball. There isn't one in the 2.15.0 tarball either, but the code definitely references it. Upstream inquires (https://mantisbt.org/forums/viewtopic.php?p=66190#p66190 <https://mantisbt.org/forums/viewtopic.php?p=66190#p66190>) indicate we should be using Sourceforge downloads, not GitHub The vendor directory is present in the SF tarball at https://sourceforge.net/projects/mantisbt/files/mantis-stable/2.16.0/ <https://sourceforge.net/projects/mantisbt/files/mantis-stable/2.16.0/> I tried last night to update the port, but failed. Ideas please? -- Dan Langille - BSDCan / PGCon d...@langille.org > > Modified: head/databases/mantis/Makefile > == > --- head/databases/mantis/MakefileSun Jul 29 10:42:23 2018 > (r475643) > +++ head/databases/mantis/MakefileSun Jul 29 10:51:37 2018 > (r475644) > @@ -2,7 +2,7 @@ > # $FreeBSD$ > > PORTNAME= mantis > -PORTVERSION= 2.9.0 > +PORTVERSION= 2.15.0 > CATEGORIES= databases www > > MAINTAINER= po...@freebsd.org > > Modified: head/databases/mantis/distinfo > == > --- head/databases/mantis/distinfoSun Jul 29 10:42:23 2018 > (r475643) > +++ head/databases/mantis/distinfoSun Jul 29 10:51:37 2018 > (r475644) > @@ -1,6 +1,6 @@ > -TIMESTAMP = 1512569589 > -SHA256 > (mantisbt-mantisbt-2.9.0-798792eebed2e699a98c4e94f4aeb4c35eba55f8_GH0.tar.gz) > = a6b1bbd8fa7bc7f541e77a59e9f8a6fd496b8f61f2cd0423750ad6731af93522 > -SIZE > (mantisbt-mantisbt-2.9.0-798792eebed2e699a98c4e94f4aeb4c35eba55f8_GH0.tar.gz) > = 4547294 > +TIMESTAMP = 1532858621 > +SHA256 > (mantisbt-mantisbt-2.15.0-798792eebed2e699a98c4e94f4aeb4c35eba55f8_GH0.tar.gz) > = a6b1bbd8fa7bc7f541e77a59e9f8a6fd496b8f61f2cd0423750ad6731af93522 > +SIZE > (mantisbt-mantisbt-2.15.0-798792eebed2e699a98c4e94f4aeb4c35eba55f8_GH0.tar.gz) > = 4547294 > SHA256 > (mantisbt-swagger-ui-0beaf61f94c3e6359ee0cdaf7b23ad4b365d761f_GH0.tar.gz) = > 5c67f724da6d2daa8931920b82f92319098badc47b57848c385a942a1420bdf4 > SIZE > (mantisbt-swagger-ui-0beaf61f94c3e6359ee0cdaf7b23ad4b365d761f_GH0.tar.gz) = > 1401641 > SHA256 (mantisbt-ADOdb-cc2c13e337f1441e3aa1e17497722c1cd98b281b_GH0.tar.gz) = > a9334857b2dfefb42273f5296ef89166e6d65fcca26857702bce579ba1add102 > ___ > svn-ports-h...@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/svn-ports-head > To unsubscribe, send any mail to "svn-ports-head-unsubscr...@freebsd.org" signature.asc Description: Message signed with OpenPGP
Re: databases/mantis - require_once(/usr/local/www/mantis/vendor/autoload.php): failed to open stream
> On Aug 19, 2018, at 9:40 PM, Dan Langille wrote:
>
>> On Aug 19, 2018, at 5:13 PM, Rozhuk Ivan wrote:
>>
>> On Sun, 19 Aug 2018 12:38:03 -0400
>> Dan Langille wrote:
>>
>>> There is no vendor directory in the pkg-plist for this port.
>>>
>>> What are other mantis users doing?
>>
>> IMHO makefile must be switch back to use vendor site instead github: vendor
>> package contain things that not included in github.
>
> Thank you for this very helpful information.
>
> I'll work on a port upgrade for 2.16.
I've hit my first block.
Some of the configuration items for mantis are from GitHub. I do not know how
to combine both SF and GitHub into MASTER_SITES.
For now, I tried to ignore GitHub and commented out USE_GITHUB
Then I started trying this:
DISTNAME=
${PORTNAME}bt-${DISTVERSIONPREFIX}${DISTVERSION}${DISTVERSIONSUFFIX}
MASTER_SITES= SF
MASTER_SITE_SUBDIR= mantisbt/files/mantis-stable/${PORTVERSION}
But so far I've been unable to get the download URL correct for make makesum.
It keeps trying to download from
https://downloads.sourceforge.net/project/mantisbt/files/mantis-stable/2.16.0/mantisbt-2.16.0.tar.gz
But I know I can get it from
https://sourceforge.net/projects/mantisbt/files/mantis-stable/2.16.0/mantisbt-2.16.0.tar.gz/download
Or even just
https://sourceforge.net/projects/mantisbt/files/mantis-stable/2.16.0/mantisbt-2.16.0.tar.gz
I often find this part of DISTFILES confusing and difficult when it comes to
the macros.
--
Dan Langille - BSDCan / PGCon
d...@langille.org
signature.asc
Description: Message signed with OpenPGP
Re: databases/mantis - require_once(/usr/local/www/mantis/vendor/autoload.php): failed to open stream
> On Aug 19, 2018, at 5:13 PM, Rozhuk Ivan wrote: > > On Sun, 19 Aug 2018 12:38:03 -0400 > Dan Langille wrote: > >> There is no vendor directory in the pkg-plist for this port. >> >> What are other mantis users doing? > > IMHO makefile must be switch back to use vendor site instead github: vendor > package contain things that not included in github. Thank you for this very helpful information. I'll work on a port upgrade for 2.16. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
databases/mantis - require_once(/usr/local/www/mantis/vendor/autoload.php): failed to open stream
I'm using databases/mantis on FreeBSD 11.2 with PHP 5.6.37 The error I see is: [Sun Aug 19 16:21:51.240380 2018] [:error] [pid 13018] [client 50.243.42.139:50002] PHP Warning: require_once(/usr/local/www/mantis/vendor/autoload.php): failed to open stream: No such file or directory in /usr/local/www/mantis/core.php on line 71 [Sun Aug 19 16:21:51.240434 2018] [:error] [pid 13018] [client 50.243.42.139:50002] PHP Fatal error: require_once(): Failed opening required '/usr/local/www/mantis/vendor/autoload.php' (include_path='/usr/local/share/pear') in /usr/local/www/mantis/core.php on line 71 Searching online, it seems I must run: composer install https://stackoverflow.com/questions/41209349/requirevendor-autoload-php-failed-to-open-stream#41209642 I see nothing mentioned in the docs: https://mantisbt.org/docs/master/en-US/Admin_Guide/html/admin.install.config.html ... etc There is no vendor directory in the pkg-plist for this port. What are other mantis users doing? -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: getting PKGNAME from CONFLICTS
> On Aug 15, 2018, at 8:46 AM, Matthew Seaman wrote: > > On 15/08/2018 00:35, Dan Langille wrote: >>> On Aug 14, 2018, at 2:55 PM, Mark Millard via freebsd-ports >>> wrote: >>> >>> >>> Dan Langille dan at langille.org wrote on >>> Tue Aug 14 17:54:01 UTC 2018 : >>> >>>> . . . >>>> At https://dev.freshports.org/www/p5-CGI/ you can see: >>>> >>>> CONFLICTS: p5-CGI.pm-[1-3]* >>>> . . . >>>> To extract the PKGNAME values from the CONFLICTS I will need to remove >>>> everything after the trailing dash. >>>> . . . >>> >>> p5- >>> vs. >>> p5-CGI.pm- >>> vs. >>> p5-CGI.pm-[1- >>> >>> It looks to me like "trailing dash" probably has a >>> complicated definition where some "-"(s) may exist >>> that are to be ignored after the one of interest. >>> In the example I'm guessing that the middle >>> "-" is intended (so "p5-CGI.pm-"). >> >> Agreed. The hard part is identifying the regex and deleting it from >> consideration. >> > > If you don't mind spawning a new process, you can just do: > > % pkg search -qg 'p5-CGI.pm-[1-3]*' > p5-CGI.pm-3.63_1,1 > > This does assume your pkg(8) is configured to use a repository with all > possible packages available. The default FreeBSD repositories are a > good choice in that regard. They are good, but the FreshPorts database has a collection of PKGNAME values for every port. I want to query those values. > Or if you already have a database table with all of the package names > and versions, then you'll presumably want to change the glob expression > into a regex match (in this case something like '^p5-CGI\.pm-[1-3].*') freshports.dev=# SELECT categoryport(id) FROM ( SELECT distinct P.id FROM ports P JOIN commit_log_ports CLP on P.id = CLP.port_id WHERE P.package_name || '-' || CLP.port_version ~ 'p5-CGI.pm-[1-3].*') TMP; categoryport --- www/p5-CGI.pm This queries all ports, all commits on those ports, and the revisions for each port from that commit. It takes about 2 seconds. This query takes about 1 second: freshports.dev=# SELECT distinct categoryport(id) FROM freshports.dev-# (SELECT P.id, P.package_name || '-' || CLP.port_version AS release freshports.dev(# FROM ports P JOIN commit_log_ports CLP on P.id = CLP.port_id) TMP freshports.dev-# WHERE TMP.release like 'p5-CGI.pm%'; categoryport --- www/p5-CGI.pm (1 row) I think this might be the best solution for searching for a distinct match. Everyone feel free to correct my thinking here: ### Ideally, the result for this query would be static, never changing. But it can change as ports are added/removed. Assuming FreshPorts runs the above query while processing a commit to www/p5-CGI, then the data collected *could* be invalidated during the next commit. But this only affects PKGNAME values. Thus, we only have to requery after a new port is added or when a change occurs to the PKGNAME for a port. ### We can requery in the background as required. > Unless there's a PG extension that allows using glob(3) to match > strings? I can't see one after a pretty cursory search. (sqlite has > glob(3) support, which is what the pkg(8) command above is using under > the hood.) I found this: https://www.postgresql.org/docs/10/static/functions-matching.html#FUNCTIONS-POSIX-REGEXP <https://www.postgresql.org/docs/10/static/functions-matching.html#FUNCTIONS-POSIX-REGEXP> It seems to be pretty close. Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
Re: getting PKGNAME from CONFLICTS
> On Aug 15, 2018, at 9:17 AM, Rodrigo Osorio wrote: > > On 08/15/18 14:46, Matthew Seaman wrote: >> On 15/08/2018 00:35, Dan Langille wrote: >>>> On Aug 14, 2018, at 2:55 PM, Mark Millard via freebsd-ports >>>> wrote: >>>> >>>> >>>> Dan Langille dan at langille.org wrote on >>>> Tue Aug 14 17:54:01 UTC 2018 : >>>> >>>>> . . . >>>>> At https://dev.freshports.org/www/p5-CGI/ you can see: >>>>> >>>>> CONFLICTS: p5-CGI.pm-[1-3]* >>>>> . . . >>>>> To extract the PKGNAME values from the CONFLICTS I will need to remove >>>>> everything after the trailing dash. >>>>> . . . >>>> p5- >>>> vs. >>>> p5-CGI.pm- >>>> vs. >>>> p5-CGI.pm-[1- >>>> >>>> It looks to me like "trailing dash" probably has a >>>> complicated definition where some "-"(s) may exist >>>> that are to be ignored after the one of interest. >>>> In the example I'm guessing that the middle >>>> "-" is intended (so "p5-CGI.pm-"). >>> Agreed. The hard part is identifying the regex and deleting it from >>> consideration. >>> >> If you don't mind spawning a new process, you can just do: >> >> % pkg search -qg 'p5-CGI.pm-[1-3]*' >> p5-CGI.pm-3.63_1,1 >> >> This does assume your pkg(8) is configured to use a repository with all >> possible packages available. The default FreeBSD repositories are a >> good choice in that regard. >> >> Or if you already have a database table with all of the package names >> and versions, then you'll presumably want to change the glob expression >> into a regex match (in this case something like '^p5-CGI\.pm-[1-3].*') >> Unless there's a PG extension that allows using glob(3) to match >> strings? I can't see one after a pretty cursory search. (sqlite has >> glob(3) support, which is what the pkg(8) command above is using under >> the hood.) >> >> Cheers, >> >> Matthew >> > Hi, > > Why do you uses regexp instead of evaluating them with fnmatch ? > The function is available (at least) in php, python and ruby. I want to extract PKGNAME from CONFLICTS. I was not trying to match anything with the raw CONFLICTS field. With PKGNAME, the application can then search the database. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
Re: getting PKGNAME from CONFLICTS
> On Aug 14, 2018, at 2:55 PM, Mark Millard via freebsd-ports > wrote: > > > Dan Langille dan at langille.org wrote on > Tue Aug 14 17:54:01 UTC 2018 : > >> . . . >> At https://dev.freshports.org/www/p5-CGI/ you can see: >> >> CONFLICTS: p5-CGI.pm-[1-3]* >> . . . >> To extract the PKGNAME values from the CONFLICTS I will need to remove >> everything after the trailing dash. >> . . . > > p5- > vs. > p5-CGI.pm- > vs. > p5-CGI.pm-[1- > > It looks to me like "trailing dash" probably has a > complicated definition where some "-"(s) may exist > that are to be ignored after the one of interest. > In the example I'm guessing that the middle > "-" is intended (so "p5-CGI.pm-"). Agreed. The hard part is identifying the regex and deleting it from consideration. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: getting PKGNAME from CONFLICTS
> On Aug 14, 2018, at 3:15 PM, Adam Weinberger wrote: > > On Tue, Aug 14, 2018 at 1:13 PM Adam Weinberger wrote: >> >> On Tue, Aug 14, 2018 at 11:54 AM Dan Langille wrote: >>> >>> I am in the process of adding FreshPorts support for CONFLICTS[1]. >>> >>> I reference www/p5-CGI only because it had a recent commit and it contains >>> a CONFLICTS directive. >>> >>> At https://dev.freshports.org/www/p5-CGI/ you can see: >>> >>>CONFLICTS: p5-CGI.pm-[1-3]* >>> >>> My goal is to allow a link to the search page or to the page for that port. >>> >>> The CONFLICTS field is a 'A list of package name patterns'[2] >>> >>> I interpret this to mean the values returned by make -V PKGNAME and the >>> documentation[3] indicates this is correct. >>> >>> Thus, the CONFLICTS matches https://dev.freshports.org/www/p5-CGI.pm [4] >>> >>> To extract the PKGNAME values from the CONFLICTS I will need to remove >>> everything after the trailing dash. >>> >>> Can someone do me up a regex for that please? regex is a weak point for me. >>> Or point me into the Mk infrastructure to locate an existing solution. >> >> Hi Dan, >> >> I suspect that the best you can get is to make some assumptions about >> what versions look like. This will catch most, but not all: >> >> /^(.*?)-[-_\.,0-9ab]+$/ > > On second thought, 'p' should be in there too. [-_\.,0-9abp]. I think we are trying to match the regex instead of noticing a regex and removing it. Am I doing it wrong? $ cat ~/tmp/pkg-name.php p5-CGI.pm-[ [1] => p5-CGI.pm ) It's close, but not quite. Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
getting PKGNAME from CONFLICTS
I am in the process of adding FreshPorts support for CONFLICTS[1]. I reference www/p5-CGI only because it had a recent commit and it contains a CONFLICTS directive. At https://dev.freshports.org/www/p5-CGI/ you can see: CONFLICTS: p5-CGI.pm-[1-3]* My goal is to allow a link to the search page or to the page for that port. The CONFLICTS field is a 'A list of package name patterns'[2] I interpret this to mean the values returned by make -V PKGNAME and the documentation[3] indicates this is correct. Thus, the CONFLICTS matches https://dev.freshports.org/www/p5-CGI.pm [4] To extract the PKGNAME values from the CONFLICTS I will need to remove everything after the trailing dash. Can someone do me up a regex for that please? regex is a weak point for me. Or point me into the Mk infrastructure to locate an existing solution. Thank you. refs: [1] - https://github.com/FreshPorts/freshports/issues/84 [2] - https://svnweb.freebsd.org/ports/head/Mk/bsd.port.mk?view=markup#l436 [3] - https://www.freebsd.org/doc/en/books/porters-handbook/conflicts.html [4] - https://dev.freshports.org/search.php?stype=package=prefix=p5-CGI.pm=10=category=asc=Search=html=head -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
Re: PHP composer for upgrading net-mgmt/librenms
If you want what I have so far, you are welcome to it. This builds, but does not run, because of missing vendor files. https://gist.github.com/dlangille/ae3eddf4a76bb9e6189ef8a2773cdf65 <https://gist.github.com/dlangille/ae3eddf4a76bb9e6189ef8a2773cdf65> Happy with help more. BTW, I heard rumors that 1.42 is not far away, but I reckon that will be easier to port. Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org > On Jul 27, 2018, at 10:53 AM, Muhammad Moinur Rahman wrote: > > Hi Dan, > I have already started upgrading my outdated ports. Hopefully will start on > librenms tonight lets see what can be done. > > Kind regards, > bofh > > On Fri, Jul 27, 2018 at 8:37 PM Dan Langille <mailto:d...@langille.org>> wrote: > Hello, > > I tried to create a patch for net-mgmt/librenms earlier this week. > > I got caught up on composer. > > The latest version of net-mgmt/librenms no longer ships with a vendor > directory which means we have to somehow include all the dependencies > directly. > > I stopped there. > > Has anyone already done such composer magic for other PHP packages? > > Thank you. > > -- > Dan Langille - BSDCan / PGCon > d...@langille.org <mailto:d...@langille.org> > > signature.asc Description: Message signed with OpenPGP
PHP composer for upgrading net-mgmt/librenms
Hello, I tried to create a patch for net-mgmt/librenms earlier this week. I got caught up on composer. The latest version of net-mgmt/librenms no longer ships with a vendor directory which means we have to somehow include all the dependencies directly. I stopped there. Has anyone already done such composer magic for other PHP packages? Thank you. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP
Re: www/rubygem-passenger creating slave ports
Sergey,
Romain Tartière (romain@) has just commented at
https://reviews.freebsd.org/D13889:
"Great, this has annoyed me for a while, I fixed this upstream but this has not
been released yet. Patching will therefore become useless in the future, but
for now, thanks for fixing this!"
Given the port will need to be redone when this new feature is released, I
would like to proceed, with your approval, with the two slave ports as an
interim measure.
This will all be redundant once ruby FLAVORs are in.
Thank you.
--
Dan Langille - BSDCan / PGCon
d...@langille.org
> On Jan 12, 2018, at 8:11 PM, Sergey A. Osokin <o...@freebsd.org> wrote:
>
> Hi Dan,
>
> my suggestion in this case is to try to play with next options
> in www/rubygem-passenger:
>
> OPTIONS_SINGLE= SERVER
> OPTIONS_SINGLE_SERVER= APACHE22 NGINX
>
> Again, I believe it's possible to modify www/rubygem-passenger to
> have ability to install necessary files to support one of them
> (apache or nginx) or even both technologies.
>
> --
> Sergey A. Osokin
>
> On Fri, Jan 12, 2018 at 08:02:08PM -0500, Dan Langille wrote:
>> Hello Sergey,
>>
>> The proposed solution is a short term solution which goes away when FLAVORS
>> are done for ruby. This is not soon.
>>
>> We (feld@ & I) choose this solution as it has very little affect on the
>> existing port.
>>
>> The need arose today & we hope to proceed quickly with whatever can be done.
>>
>> How do you suggest we proceed with the work?
>>
>> Thank you
>>
>> --
>> Dan Langille
>> http://langille.org/
>>
>>
>>> On Jan 12, 2018, at 7:33 PM, Sergey A. Osokin <o...@freebsd.org> wrote:
>>>
>>> Hi Dan,
>>>
>>> I don't see any reason to create two separate ports in this case and support
>>> them then cause potentially it's possible to modify the existing port to get
>>> the same result.
>>>
>>> --
>>> Sergey A. Osokin
>>>
>>>> On Fri, Jan 12, 2018 at 10:18:37AM -0500, Dan Langille wrote:
>>>> Sergey,
>>>>
>>>> We have a need to use rubygem-passenger with both Nginx and with Apache
>>>> (on different servers).
>>>>
>>>> I propose to create two new slave ports:
>>>>
>>>> * www/rubygem-passenger-nginx
>>>> * www/rubygem-passenger-passenger
>>>>
>>>> To do this, I would like to make a one-line change to the master port:
>>>>
>>>>
>>>> Index: Makefile
>>>> ===
>>>> --- Makefile(revision 458852)
>>>> +++ Makefile(working copy)
>>>> @@ -24,7 +24,7 @@
>>>> APACHE22_DESC=Build with apache22 support
>>>> NGINX_DESC=Build with nginx support
>>>> SYMLINK_DESC=Create passenger symlink
>>>> -OPTIONS_DEFAULT=APACHE22
>>>> +OPTIONS_DEFAULT=?APACHE22
>>>>
>>>> USE_RUBY=yes
>>>> RAKE_BIN=${LOCALBASE}/bin/rake
>>>>
>>>> This will allow the slave port to set OPTIONS_DEFAULT and build for the
>>>> desired web server.
>>>>
>>>> If you agree, I am happen to make the commit / submit a patch for your
>>>> approval.
>>>>
>>>> Thank you.
>>>>
>>>> --
>>>> Dan Langille - BSDCan / PGCon
>>>> d...@langille.org
>>>>
>>>>
>>
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: www/rubygem-passenger creating slave ports
Hello Sergey,
The proposed solution is a short term solution which goes away when FLAVORS are
done for ruby. This is not soon.
We (feld@ & I) choose this solution as it has very little affect on the
existing port.
The need arose today & we hope to proceed quickly with whatever can be done.
How do you suggest we proceed with the work?
Thank you
--
Dan Langille
http://langille.org/
> On Jan 12, 2018, at 7:33 PM, Sergey A. Osokin <o...@freebsd.org> wrote:
>
> Hi Dan,
>
> I don't see any reason to create two separate ports in this case and support
> them then cause potentially it's possible to modify the existing port to get
> the same result.
>
> --
> Sergey A. Osokin
>
>> On Fri, Jan 12, 2018 at 10:18:37AM -0500, Dan Langille wrote:
>> Sergey,
>>
>> We have a need to use rubygem-passenger with both Nginx and with Apache (on
>> different servers).
>>
>> I propose to create two new slave ports:
>>
>> * www/rubygem-passenger-nginx
>> * www/rubygem-passenger-passenger
>>
>> To do this, I would like to make a one-line change to the master port:
>>
>>
>> Index: Makefile
>> ===
>> --- Makefile(revision 458852)
>> +++ Makefile(working copy)
>> @@ -24,7 +24,7 @@
>> APACHE22_DESC=Build with apache22 support
>> NGINX_DESC=Build with nginx support
>> SYMLINK_DESC=Create passenger symlink
>> -OPTIONS_DEFAULT=APACHE22
>> +OPTIONS_DEFAULT=?APACHE22
>>
>> USE_RUBY=yes
>> RAKE_BIN=${LOCALBASE}/bin/rake
>>
>> This will allow the slave port to set OPTIONS_DEFAULT and build for the
>> desired web server.
>>
>> If you agree, I am happen to make the commit / submit a patch for your
>> approval.
>>
>> Thank you.
>>
>> --
>> Dan Langille - BSDCan / PGCon
>> d...@langille.org
>>
>>
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: www/rubygem-passenger creating slave ports
> On Jan 12, 2018, at 11:48 AM, Dan Langille <d...@langille.org> wrote: > > >> On Fri, Jan 12, 2018 at 7:23 AM, Dan Langille <d...@langille.org> wrote: >>> On Jan 12, 2018, at 10:18 AM, Dan Langille <d...@langille.org> wrote: >>> >>> Sergey, >>> >>> We have a need to use rubygem-passenger with both Nginx and with Apache (on >>> different servers). >>> >>> I propose to create two new slave ports: >>> >>> * www/rubygem-passenger-nginx >>> * www/rubygem-passenger-passenger >> >> That should be: >> >>> * www/rubygem-passenger-apache >>> * www/rubygem-passenger-nginx >> >> Sorry. >> > >> On Jan 12, 2018, at 11:16 AM, Ultima <ultima1...@gmail.com> wrote: >> >> This is an example of a perfect use case for flavors. Instead of >> a slave port, creating an nginx flavor and apache flavor would >> be a better approach. >> > > Agreed. > > Flavors was our first choice, but is part of a much bigger project which will > not arrive soon enough to meet this need. https://reviews.freebsd.org/D13889 -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: www/rubygem-passenger creating slave ports
> On Fri, Jan 12, 2018 at 7:23 AM, Dan Langille <d...@langille.org> wrote: > > On Jan 12, 2018, at 10:18 AM, Dan Langille <d...@langille.org> wrote: > > > > Sergey, > > > > We have a need to use rubygem-passenger with both Nginx and with Apache (on > > different servers). > > > > I propose to create two new slave ports: > > > > * www/rubygem-passenger-nginx > > * www/rubygem-passenger-passenger > > That should be: > > > * www/rubygem-passenger-apache > > * www/rubygem-passenger-nginx > > Sorry. > > On Jan 12, 2018, at 11:16 AM, Ultima <ultima1...@gmail.com> wrote: > > This is an example of a perfect use case for flavors. Instead of > a slave port, creating an nginx flavor and apache flavor would > be a better approach. > Agreed. Flavors was our first choice, but is part of a much bigger project which will not arrive soon enough to meet this need. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: www/rubygem-passenger creating slave ports
> On Jan 12, 2018, at 10:18 AM, Dan Langille <d...@langille.org> wrote: > > Sergey, > > We have a need to use rubygem-passenger with both Nginx and with Apache (on > different servers). > > I propose to create two new slave ports: > > * www/rubygem-passenger-nginx > * www/rubygem-passenger-passenger That should be: > * www/rubygem-passenger-apache > * www/rubygem-passenger-nginx Sorry. ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
www/rubygem-passenger creating slave ports
Sergey,
We have a need to use rubygem-passenger with both Nginx and with Apache (on
different servers).
I propose to create two new slave ports:
* www/rubygem-passenger-nginx
* www/rubygem-passenger-passenger
To do this, I would like to make a one-line change to the master port:
Index: Makefile
===
--- Makefile(revision 458852)
+++ Makefile(working copy)
@@ -24,7 +24,7 @@
APACHE22_DESC= Build with apache22 support
NGINX_DESC=Build with nginx support
SYMLINK_DESC= Create passenger symlink
-OPTIONS_DEFAULT= APACHE22
+OPTIONS_DEFAULT=? APACHE22
USE_RUBY= yes
RAKE_BIN= ${LOCALBASE}/bin/rake
This will allow the slave port to set OPTIONS_DEFAULT and build for the desired
web server.
If you agree, I am happen to make the commit / submit a patch for your approval.
Thank you.
--
Dan Langille - BSDCan / PGCon
d...@langille.org
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: Procmail Vulnerabilities check
> On Dec 13, 2017, at 12:27 PM, Christoph Brinkhaus <c.brinkh...@t-online.de> > wrote: > > On Wed, Dec 13, 2017 at 11:35:55AM +0100, Jos Chrispijn wrote: >> On 8-12-2017 17:58, Warren Block wrote: >>> procmail is ancient, and has had known quality issues for much of the >>> time. Consider maildrop as a more powerful and more maintained >>> replacement that is pretty easy to implement: >> I know - but I can remember that procmail should be installed also when >> using Postfix. >> Might be wrong here... > > Dear Joe, > > I have replaced procmail by maildrop recently using it with Postfix. > There has been just one single obstacle. I run fetchmail as suer > fetchmail started with the entry in /etc/rc.conf. The mails have been > delivered to Postfix which involked procmail to distribute the mail. > > With maildrop this did not work initially. Adding the user fetchmail > to /etc/aliases with a proper alias address followed by the command > newaliases fixed that. I like such replacements. However, if third party code is required, there is little we can do in the short term. Case in point: security/logcheck. I went upstream looking to see why Debian uses that. I cannot recall exactly what it was, but it wasn't procmail, but another utility provide by procmail. I stopped there. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
FreshPorts and FLAVORS
I've been reading up on FLAVORS
(https://wiki.freebsd.org/Ports/FlavorsAndSubPackages)
and creating a plan for FreshPorts.
Since FLAVORS came in, FreshPorts has been reporting false sanity test failures
like this:
https://www.freshports.org/sanity_test_failures.php?message_id=201712041212.vb4ccwu4059...@repo.freebsd.org
e.g.
###
devel/py-flake8:
NOTE: this particular sanity test is very experimental
A port specified in the BUILD_DEPENDS of devel/py-flake8 does not exist:
'devel/py-pytest-runner@py27' on branch 'head'.
NOTE: this particular sanity test is very experimental
A port specified in the BUILD_DEPENDS of devel/py-flake8 does not exist:
'devel/py-setuptools@py27' on branch 'head'.
NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of devel/py-flake8 does not exist:
'devel/py-mccabe@py27' on branch 'head'.
...
###
Some background based on the Makefile for the above commit:
https://svnweb.freebsd.org/ports/head/devel/py-flake8/Makefile?view=markup=455509
FreshPorts converts this line:
BUILD_DEPENDS=
${PYTHON_PKGNAMEPREFIX}pytest-runner>=0:devel/py-pytest-runner@${FLAVOR}
to:
py27-pytest-runner>=0:devel/py-pytest-runner@py27
The part before the : is dropped, leaving us with the port upon which
devel/py-flake8 is dependent:
py-pytest-runner@py27
FreshPorts looks for that in the ports tree, and fails. As such, FreshPorts
needs to learn about FLAVORS.
Or more precisely, at least in the short term, how to ignore FLAVORS.
The solution I have is:
###
[dan@pro02:~/tmp] $ cat depends.pl
#!/usr/local/bin/perl
my $depends = 'devel/py-setuptools@py27';
($depends, undef) = split /[@\:]/, $depends;
print "'$depends'\n";
[dan@pro02:~/tmp] $ perl depends.pl
'devel/py-setuptools'
###
This results in a port which FreshPorts can find in the tree. I plan to start
coding and testing this
approach later today.
This solution also caters for depends such as devel/git:configure which also
created false positives.
I suspect this will only be a short term solution.
short term issues:
* I am not yet sure about duplicate depends. If a port has
devel/py-setuptools@py27 and devel/py-setuptools@py36, what will happen?
* looking at the depends on FreshPorts, you cannot tell which FLAVORS are
available / in use.
long term issues:
* Should FreshPorts discern between different FLAVORS? How useful would that be?
* How should FreshPorts display FLAVORS available in a port.
Comments and suggestions welcome. So are patches.
--
Dan Langille - BSDCan / PGCon
d...@langille.org
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: FreshPorts API
> On Sep 21, 2017, at 4:10 PM, Torsten Zuehlsdorff > <mailingli...@toco-domains.de> wrote: > > > > On 21.09.2017 14:31, Dan Langille wrote: >> FreshPorts contains more than just ports. It contains all commits to the >> FreeBSD repository. >> This is all stored in a database. It can be queried. We just need an API. >> The goal is to start gradually, based on the ideas started at >> https://duck.co/ideas/idea/936/grab-bsd-port-summary-from-freshports >> <https://duck.co/ideas/idea/936/grab-bsd-port-summary-from-freshports> >> and later at https://github.com/FreshPorts/freshports/issues/4 >> <https://github.com/FreshPorts/freshports/issues/4> >> I did start using https://github.com/marcj/php-rest-service >> <https://github.com/marcj/php-rest-service> (I didn't get to the >> point of having methods invoked), I'm looking for others to help get this >> started. >> The goal is a PHP-based API. Why? Existing PHP code for database access. > > But what problem to you want to solve with it? > > I regularly dump the ports-tree into a database for some research or hacking. > But beside this very specific cases i never needed it. So what do you want to > accomplish? This API is for third party use. We can try to anticipate their use cases by providing a general API, but in general, it's unknown. In the duck.co <http://duck.co/> link above, it shows one such use case: search. It allows a third party to search FreshPorts. to find results. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
FreshPorts API
FreshPorts contains more than just ports. It contains all commits to the FreeBSD repository. This is all stored in a database. It can be queried. We just need an API. The goal is to start gradually, based on the ideas started at https://duck.co/ideas/idea/936/grab-bsd-port-summary-from-freshports <https://duck.co/ideas/idea/936/grab-bsd-port-summary-from-freshports> and later at https://github.com/FreshPorts/freshports/issues/4 <https://github.com/FreshPorts/freshports/issues/4> I did start using https://github.com/marcj/php-rest-service <https://github.com/marcj/php-rest-service> (I didn't get to the point of having methods invoked), I'm looking for others to help get this started. The goal is a PHP-based API. Why? Existing PHP code for database access. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: installing a kibana plugin (re textproc/kibana43)
> On Nov 1, 2016, at 1:14 PM, Thierry Thomas <thie...@freebsd.org> wrote: > > Hello, > > Le mar. 1 nov. 16 à 17:23:49 +0100, Dan Langille <d...@langille.org> > écrivait : > >> I am using Kibana 4.3 on FreeBSD 9.3 and I want to use >> textproc/elasticsearch2-plugin-marvel 2.1.0 >> >> I have everything installed from packages and I'm reading the Elasticseach >> documentation >> (see #5 at https://www.elastic.co/guide/en/marvel/2.1/installing-marvel.html >> <https://www.elastic.co/guide/en/marvel/2.1/installing-marvel.html>) >> >> I'm getting stuck on 'Install the Marvel app into Kibana'. >> >> The instructions look for a kibana binary (I think). The command is: >> >> bin/kibana plugin --install elasticsearch/marvel/2.1.2 >> >> Where is that kibana binary? It's not references by the port that I could >> find. > > I guess that this the port is too old. If you look at > <https://github.com/elastic/kibana> (ATM this is 5.0.0) > there is is a bin directory, providing a kebana script. We are using 4.3 with Elasticsearch 2.1. I suspect we are on 4.3 for compatibility reasons. The binary/script is in the tarball. The Makefile deletes it. re: https://svnweb.freebsd.org/ports/head/textproc/kibana45/Makefile?revision=419775=markup#l28 <https://svnweb.freebsd.org/ports/head/textproc/kibana45/Makefile?revision=419775=markup#l28> (link is to master port). I'm happy to patch if the maintainer is OK with that. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
installing a kibana plugin (re textproc/kibana43)
I am using Kibana 4.3 on FreeBSD 9.3 and I want to use textproc/elasticsearch2-plugin-marvel 2.1.0 I have everything installed from packages and I'm reading the Elasticseach documentation (see #5 at https://www.elastic.co/guide/en/marvel/2.1/installing-marvel.html <https://www.elastic.co/guide/en/marvel/2.1/installing-marvel.html>) I'm getting stuck on 'Install the Marvel app into Kibana'. The instructions look for a kibana binary (I think). The command is: bin/kibana plugin --install elasticsearch/marvel/2.1.2 Where is that kibana binary? It's not references by the port that I could find. How does one install a Kibana plugin on FreeBSD? Thanks. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: 2016Q2 - no Mk/Uses/mysql.mk
> On Jun 28, 2016, at 7:56 AM, Mathieu Arnold <m...@freebsd.org> wrote: > > > > +--On 27 juin 2016 17:22:56 -0400 Dan Langille <d...@langille.org> wrote: > | I've been working with the 2016Q2 branch with respect to FreshPorts. > | > | I tried running make -V on branches/2016Q2 and it failed with: > | > | Error message is: make: "/usr/local/repos/PORTS-2016Q2/Mk/bsd.port.mk" > | line 1433: Cannot open /usr/local/repos/PORTS-2016Q2/Mk/Uses/mysql.mk > | > | It seems that recent changes to Mk also need to be backported to branches > | if there is a commit on that branch. > > Well, no, people merging things are supposed to check what they are doing > and not commit stupid patches, which is the case here. > > Which port has USES=mysql on 2016Q2 ? Log: MFH: r416066 net-mgmt/cacti: 0.8.8g -> 0.8.8h It appears to have already been fixed in Revision 416207: - no USES=mysql allowed in the quarterly branch re https://svnweb.freebsd.org/ports/branches/2016Q2/net-mgmt/cacti/Makefile?view=log <https://svnweb.freebsd.org/ports/branches/2016Q2/net-mgmt/cacti/Makefile?view=log> Ooops. Sorry for the noise. I'll look at the new commits on that port. I was analyzing commits which FreshPorts had difficulty processing. Nothing to see here. Please move along... -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP using GPGMail
2016Q2 - no Mk/Uses/mysql.mk
I've been working with the 2016Q2 branch with respect to FreshPorts. I tried running make -V on branches/2016Q2 and it failed with: Error message is: make: "/usr/local/repos/PORTS-2016Q2/Mk/bsd.port.mk" line 1433: Cannot open /usr/local/repos/PORTS-2016Q2/Mk/Uses/mysql.mk It seems that recent changes to Mk also need to be backported to branches if there is a commit on that branch. Agreed? -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
UPDATING entry has no AFFECTS clause
re revision 383476 to UPDATING https://svnweb.freebsd.org/ports/head/UPDATING?r1=383474=383476 <https://svnweb.freebsd.org/ports/head/UPDATING?r1=383474=383476> This entry has no AFFECTS clause. Should it? I ask because I encounter this issue when fixing something in FreshPorts. ### 20150331: AUTHOR: oha...@freebsd.org Subversion modules are no longer activated in httpd.conf! To activate the subversion apache modules a dedicated file will be installed as modules.d/220_subversion.conf ### Should it be devel/subversion? www/apache*? I'm happy to do the fix. I just want to check to see which port[s] you think it should be. -- Dan Langille - BSDCan / PGCon d...@langille.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: New addition to /usr/ports/MOVED file breaks portupgrade
> On Mar 10, 2016, at 1:14 PM, Manfred Antar <manfredan...@yahoo.com> wrote: > > This addition to MOVED breaks portupgrade: > > elasticsearch-plugin-marvel2|elasticsearch2-plugin-marvel|Rename to match > other Elasticsearch 2 plugins > > I think it needs to be: > > elasticsearch-plugin-marvel|2016-03-10|elasticsearch2-plugin-marvel|Rename to > match other Elasticsearch 2 plugins I think it needs to be: textproc/elasticsearch-plugin-marvel2|textproc/elasticsearch2-plugin-marvel|Rename to match other Elasticsearch 2 plugins NOTE: original name was elasticsearch-plugin-marvel2, which was a repo-copy from elasticsearch-plugin-marvel New name is elasticsearch2-plugin-marvel Thanks for ehaupt and antoine for their fixes. -- Dan Langille - BSDCan / PGCon d...@langille.org signature.asc Description: Message signed with OpenPGP using GPGMail
need a particular version of django
I'm writing about this problem:
http://beefy2.nyi.freebsd.org/data/93amd64-default/388198/logs/py27-crits-3.1.0.log
Background:
security/py-crits needs www/py-django16 and www/py-django-tastypie
However, www/py-django-tastypie wants to install www/py-django (which is v
1.8).
This is the line from www/py-django-tastypie in question:
RUN_DEPENDS:= ${BUILD_DEPENDS} \
${PYTHON_PKGNAMEPREFIX}django=1.5:${PORTSDIR}/www/py-django
Is there a nice way to say www/py-django16 is also OK?
—
Dan Langille
http://langille.org/
signature.asc
Description: Message signed with OpenPGP using GPGMail
Re: need a particular version of django
On Jun 1, 2015, at 9:27 AM, Mathieu Arnold m...@freebsd.org wrote:
+--On 1 juin 2015 08:35:04 -0400 Dan Langille d...@langille.org wrote:
| I'm writing about this problem:
| http://beefy2.nyi.freebsd.org/data/93amd64-default/388198/logs/py27-crits
| -3.1.0.log
|
| Background:
|
| security/py-crits needs www/py-django16 and www/py-django-tastypie
|
| However, www/py-django-tastypie wants to install www/py-django (which is
| v 1.8).
|
| This is the line from www/py-django-tastypie in question:
|
| RUN_DEPENDS:= ${BUILD_DEPENDS} \
|
| ${PYTHON_PKGNAMEPREFIX}django=1.5:${PORTSDIR}/www/py-django
You can't use := here because PYTHON_PKGNAMEPREFIX is unset at that time.
| Is there a nice way to say www/py-django16 is also OK?
No.
It seems practical to create a new port, based on www/py-django-tastypie, which
requires www/py-django16
instead of www/py-django.
Said port could be a slave of www/py-django-tastypie
Agreed?
—
Dan Langille
http://langille.org/
signature.asc
Description: Message signed with OpenPGP using GPGMail
Re: pkg-static: Insufficient privileges to add packages
This reply has been rearranged.. On Tue, Apr 21, 2015 at 6:56 AM Ben Woods woods...@gmail.com mailto:woods...@gmail.com wrote: On Apr 20, 2015, at 7:18 PM, Ben Woods woods...@gmail.com wrote: On Tue, Apr 21, 2015 at 4:02 AM Dan Langille d...@langille.org mailto:d...@langille.org wrote: I am running a testport on www/py-django-tastypie My goal it to add an option for using www/django16 as well as www/django (1.7). The error I get is: $ sudo poudriere testport -v -j 101amd64 -p testing -z testing -i -o www/py-django-tastypie ... ===Verifying install for py27-django=1.5 in /usr/ports/www/py-django16 === Installing existing package /packages/All/py27-django16-1.6.11.txz pkg-static: Insufficient privileges to add packages But because I'm running testport interactively, I see I can install the port it mentions: root@101amd64-testing-testing:~ # pkg install /packages/All/py27-django16-1.6.11.txz Updating local repository catalogue... [101amd64-testing-testing] Fetching meta.txz: 100%816 B 0.8kB/s00:01 [101amd64-testing-testing] Fetching packagesite.txz: 100% 50 KiB 51.4kB/s 00:01 Processing entries: 100% local repository update completed. 205 packages processed. Updating database digests format: 100% Checking integrity... done (0 conflicting) The following 1 package(s) will be affected (of 0 checked): New packages to be INSTALLED: py27-django16: 1.6.11 The process will require 27 MiB more space. Proceed with this action? [y/N]: y [101amd64-testing-testing] [1/1] Installing py27-django16-1.6.11... [101amd64-testing-testing] [1/1] Extracting py27-django16-1.6.11: 100% root@101amd64-testing-testing:~ # Why does the testport fail? Full build log here: http://services.unixathome.org/poudriere/data/101amd64-testing-testing/2015-04-20_19h53m23s/logs/errors/py27-django-tastypie-0.12.1.log http://services.unixathome.org/poudriere/data/101amd64-testing-testing/2015-04-20_19h53m23s/logs/errors/py27-django-tastypie-0.12.1.log That error is generated in the pkg exec_add function in src/add.c here: https://github.com/freebsd/pkg/blob/master/src/add.c https://github.com/freebsd/pkg/blob/master/src/add.c It occurs because it checks the package database for read, write, create access using the pkgdb_access and pkgdb_check_access functions in libpkg/pkgdb.c here: https://github.com/freebsd/pkg/blob/master/libpkg/pkgdb.c https://github.com/freebsd/pkg/blob/master/libpkg/pkgdb.c I believe this is inside the jail created by poudriere testport. It would be interesting to add the -i interactive flag to the testport run and check the permissions on/in the pkg database. Mine are here for reference: % ls -lah /var/db/pkg/ total 37484 drwxr-xr-x 2 root wheel 8B 21 Apr 06:00 . drwxr-xr-x 13 root wheel19B 21 Apr 06:13 .. -rw-r--r-- 1 root wheel 246B 4 Dec 06:22 FreeBSD.meta -rw-r--r-- 1 root wheel11M 21 Apr 06:00 local.sqlite -rw-r--r-- 1 root wheel 246B 21 Apr 03:05 poudriere.meta -rw-r--r-- 1 root wheel49M 4 Dec 06:22 repo-FreeBSD.sqlite -rw-r--r-- 1 root wheel 1.6M 21 Apr 03:05 repo-poudriere.sqlite -r--r--r-- 1 root wheel 3.3M 21 Apr 03:17 vuln.xml Regards, Ben Hand on a minute now my poudriere system is doing the same thing - looks like we have a bug somewhere. I am running pkg 1.5.1 and poudriere 3.1.4. According to freshports, pkg was updated on 16th April, and poudriere on 10th April. Inside my poudriere testport jail (using the -i interactive mode flag), I have: # ls -lah /var/db/pkg/ total 7953 drwxr-xr-x 2 root wheel 3B Apr 20 23:17 . drwxr-xr-x 12 root wheel14B Apr 20 23:17 .. -rw-r--r-- 1 root wheel 7.7M Apr 20 23:17 local.sqlite Thoughts? Here's what I have: Edit /etc/motd to change this login announcement. root@101amd64-testing-testing:~ # ls -lah /var/db/pkg/ total 2991 drwxr-xr-x 2 root wheel 3B Apr 21 13:28 . drwxr-xr-x 10 root wheel12B Mar 14 19:18 .. -rw-r--r-- 1 root wheel 2.8M Apr 21 13:28 local.sqlite root@101amd64-testing-testing:~ # — Dan Langille http://langille http://langille/.org/ signature.asc Description: Message signed with OpenPGP using GPGMail
Re: pkg-static: Insufficient privileges to add packages
On Apr 21, 2015, at 12:34 PM, Baptiste Daroussin b...@freebsd.org wrote: On Tue, Apr 21, 2015 at 11:44:01PM +0800, Ben Woods wrote: On Tue, Apr 21, 2015 at 4:02 AM Dan Langille d...@langille.org wrote: I am running a testport on www/py-django-tastypie My goal it to add an option for using www/django16 as well as www/django (1.7). The error I get is: $ sudo poudriere testport -v -j 101amd64 -p testing -z testing -i -o www/py-django-tastypie ... ===Verifying install for py27-django=1.5 in /usr/ports/www/py-django16 === Installing existing package /packages/All/py27-django16-1.6.11.txz pkg-static: Insufficient privileges to add packages But because I'm running testport interactively, I see I can install the port it mentions: root@101amd64-testing-testing:~ # pkg install /packages/All/py27-django16-1.6.11.txz Updating local repository catalogue... [101amd64-testing-testing] Fetching meta.txz: 100%816 B 0.8kB/s 00:01 [101amd64-testing-testing] Fetching packagesite.txz: 100% 50 KiB 51.4kB/s00:01 Processing entries: 100% local repository update completed. 205 packages processed. Updating database digests format: 100% Checking integrity... done (0 conflicting) The following 1 package(s) will be affected (of 0 checked): New packages to be INSTALLED: py27-django16: 1.6.11 The process will require 27 MiB more space. Proceed with this action? [y/N]: y [101amd64-testing-testing] [1/1] Installing py27-django16-1.6.11... [101amd64-testing-testing] [1/1] Extracting py27-django16-1.6.11: 100% root@101amd64-testing-testing:~ # Why does the testport fail? Full build log here: http://services.unixathome.org/poudriere/data/101amd64-testing-testing/2015-04-20_19h53m23s/logs/errors/py27-django-tastypie-0.12.1.log That error is generated in the pkg exec_add function in src/add.c here: https://github.com/freebsd/pkg/blob/master/src/add.c It occurs because it checks the package database for read, write, create access using the pkgdb_access and pkgdb_check_access functions in libpkg/pkgdb.c here: https://github.com/freebsd/pkg/blob/master/libpkg/pkgdb.c I believe this is inside the jail created by poudriere testport. It would be interesting to add the -i interactive flag to the testport run and check the permissions on/in the pkg database. Mine are here for reference: % ls -lah /var/db/pkg/ total 37484 drwxr-xr-x 2 root wheel 8B 21 Apr 06:00 . drwxr-xr-x 13 root wheel19B 21 Apr 06:13 .. -rw-r--r-- 1 root wheel 246B 4 Dec 06:22 FreeBSD.meta -rw-r--r-- 1 root wheel11M 21 Apr 06:00 local.sqlite -rw-r--r-- 1 root wheel 246B 21 Apr 03:05 poudriere.meta -rw-r--r-- 1 root wheel49M 4 Dec 06:22 repo-FreeBSD.sqlite -rw-r--r-- 1 root wheel 1.6M 21 Apr 03:05 repo-poudriere.sqlite -r--r--r-- 1 root wheel 3.3M 21 Apr 03:17 vuln.xml Regards, Ben Hand on a minute now my poudriere system is doing the same thing - looks like we have a bug somewhere. I am running pkg 1.5.1 and poudriere 3.1.4. According to freshports, pkg was updated on 16th April, and poudriere on 10th April. Inside my poudriere testport jail (using the -i interactive mode flag), I have: # ls -lah /var/db/pkg/ total 7953 drwxr-xr-x 2 root wheel 3B Apr 20 23:17 . drwxr-xr-x 12 root wheel14B Apr 20 23:17 .. -rw-r--r-- 1 root wheel 7.7M Apr 20 23:17 local.sqlite Thoughts? Here's what I have: Edit /etc/motd to change this login announcement. root@101amd64-testing-testing:~ # ls -lah /var/db/pkg/ total 2991 drwxr-xr-x 2 root wheel 3B Apr 21 13:28 . drwxr-xr-x 10 root wheel12B Mar 14 19:18 .. -rw-r--r-- 1 root wheel 2.8M Apr 21 13:28 local.sqlite root@101amd64-testing-testing:~ # — Dan Langille http://langille.org/ For the record, downgrading pkg to 1.4.12 and poudriere to 3.1.2 did not fix this issue for me. The port that I am trying to testport on FreeBSD 10.1 is math/plplot-ada. The testport permission issues do not occur with the port as it is currently - only after making the changes I have proposed to math/plplot here: https://bugs.freebsd.org/bugzilla/attachment.cgi?id=155771action=diff I found the reason for Dan by checking the logs, If I could see your own logs then I will tell you where you mistake is. actually you are having poudriere building as a user (internally to the jail) which is a very good thing as it allows to catch things like this :D) I mean it is dropping roots priviledges to build, stage and make package. but still try to install a dependency there, probably because the dependency line you have in your port is buggy. In dan case he was testing if py27-django=1.5 but was refering to the www/py-django16 port which result in a package named py27-django16 which is not py27-django aka the test will always fail and the ports tree will try to reinstall again and again even in phases when
Re: pkg-static: Insufficient privileges to add packages
On Apr 21, 2015, at 11:44 AM, Ben Woods woods...@gmail.com wrote: On Tue, Apr 21, 2015 at 4:02 AM Dan Langille d...@langille.org wrote: I am running a testport on www/py-django-tastypie My goal it to add an option for using www/django16 as well as www/django (1.7). The error I get is: $ sudo poudriere testport -v -j 101amd64 -p testing -z testing -i -o www/py-django-tastypie ... ===Verifying install for py27-django=1.5 in /usr/ports/www/py-django16 === Installing existing package /packages/All/py27-django16-1.6.11.txz pkg-static: Insufficient privileges to add packages But because I'm running testport interactively, I see I can install the port it mentions: root@101amd64-testing-testing:~ # pkg install /packages/All/py27-django16-1.6.11.txz Updating local repository catalogue... [101amd64-testing-testing] Fetching meta.txz: 100%816 B 0.8kB/s 00:01 [101amd64-testing-testing] Fetching packagesite.txz: 100% 50 KiB 51.4kB/s00:01 Processing entries: 100% local repository update completed. 205 packages processed. Updating database digests format: 100% Checking integrity... done (0 conflicting) The following 1 package(s) will be affected (of 0 checked): New packages to be INSTALLED: py27-django16: 1.6.11 The process will require 27 MiB more space. Proceed with this action? [y/N]: y [101amd64-testing-testing] [1/1] Installing py27-django16-1.6.11... [101amd64-testing-testing] [1/1] Extracting py27-django16-1.6.11: 100% root@101amd64-testing-testing:~ # Why does the testport fail? Full build log here: http://services.unixathome.org/poudriere/data/101amd64-testing-testing/2015-04-20_19h53m23s/logs/errors/py27-django-tastypie-0.12.1.log That error is generated in the pkg exec_add function in src/add.c here: https://github.com/freebsd/pkg/blob/master/src/add.c It occurs because it checks the package database for read, write, create access using the pkgdb_access and pkgdb_check_access functions in libpkg/pkgdb.c here: https://github.com/freebsd/pkg/blob/master/libpkg/pkgdb.c I believe this is inside the jail created by poudriere testport. It would be interesting to add the -i interactive flag to the testport run and check the permissions on/in the pkg database. Mine are here for reference: % ls -lah /var/db/pkg/ total 37484 drwxr-xr-x 2 root wheel 8B 21 Apr 06:00 . drwxr-xr-x 13 root wheel19B 21 Apr 06:13 .. -rw-r--r-- 1 root wheel 246B 4 Dec 06:22 FreeBSD.meta -rw-r--r-- 1 root wheel11M 21 Apr 06:00 local.sqlite -rw-r--r-- 1 root wheel 246B 21 Apr 03:05 poudriere.meta -rw-r--r-- 1 root wheel49M 4 Dec 06:22 repo-FreeBSD.sqlite -rw-r--r-- 1 root wheel 1.6M 21 Apr 03:05 repo-poudriere.sqlite -r--r--r-- 1 root wheel 3.3M 21 Apr 03:17 vuln.xml Regards, Ben Hand on a minute now my poudriere system is doing the same thing - looks like we have a bug somewhere. I am running pkg 1.5.1 and poudriere 3.1.4. According to freshports, pkg was updated on 16th April, and poudriere on 10th April. Inside my poudriere testport jail (using the -i interactive mode flag), I have: # ls -lah /var/db/pkg/ total 7953 drwxr-xr-x 2 root wheel 3B Apr 20 23:17 . drwxr-xr-x 12 root wheel14B Apr 20 23:17 .. -rw-r--r-- 1 root wheel 7.7M Apr 20 23:17 local.sqlite Thoughts? Here's what I have: Edit /etc/motd to change this login announcement. root@101amd64-testing-testing:~ # ls -lah /var/db/pkg/ total 2991 drwxr-xr-x 2 root wheel 3B Apr 21 13:28 . drwxr-xr-x 10 root wheel12B Mar 14 19:18 .. -rw-r--r-- 1 root wheel 2.8M Apr 21 13:28 local.sqlite root@101amd64-testing-testing:~ # — Dan Langille http://langille.org/ For the record, downgrading pkg to 1.4.12 and poudriere to 3.1.2 did not fix this issue for me. The port that I am trying to testport on FreeBSD 10.1 is math/plplot-ada. The testport permission issues do not occur with the port as it is currently - only after making the changes I have proposed to math/plplot here: https://bugs.freebsd.org/bugzilla/attachment.cgi?id=155771action=diff I am also testing local changes to a port. The problem occurs on a 10.1 but NOT on a 9.3 jail (both amd64). Here is the good 9.3 run: http://services.unixathome.org/poudriere/build.html?mastername=93amd64-testing-testingbuild=2015-04-21_16h17m45s Compare that to the failed 10.1 run: Full build log here: http://services.unixathome.org/poudriere/data/101amd64-testing-testing/2015-04-20_19h53m23s/logs/errors/py27-django-tastypie-0.12.1.log I am using pkg-1.5.1 and poudriere-3.1.4 — Dan Langille http://langille.org/ signature.asc Description: Message signed with OpenPGP using GPGMail
pkg-static: Insufficient privileges to add packages
I am running a testport on www/py-django-tastypie My goal it to add an option for using www/django16 as well as www/django (1.7). The error I get is: $ sudo poudriere testport -v -j 101amd64 -p testing -z testing -i -o www/py-django-tastypie ... ===Verifying install for py27-django=1.5 in /usr/ports/www/py-django16 === Installing existing package /packages/All/py27-django16-1.6.11.txz pkg-static: Insufficient privileges to add packages But because I'm running testport interactively, I see I can install the port it mentions: root@101amd64-testing-testing:~ # pkg install /packages/All/py27-django16-1.6.11.txz Updating local repository catalogue... [101amd64-testing-testing] Fetching meta.txz: 100%816 B 0.8kB/s00:01 [101amd64-testing-testing] Fetching packagesite.txz: 100% 50 KiB 51.4kB/s 00:01 Processing entries: 100% local repository update completed. 205 packages processed. Updating database digests format: 100% Checking integrity... done (0 conflicting) The following 1 package(s) will be affected (of 0 checked): New packages to be INSTALLED: py27-django16: 1.6.11 The process will require 27 MiB more space. Proceed with this action? [y/N]: y [101amd64-testing-testing] [1/1] Installing py27-django16-1.6.11... [101amd64-testing-testing] [1/1] Extracting py27-django16-1.6.11: 100% root@101amd64-testing-testing:~ # Why does the testport fail? Full build log here: http://services.unixathome.org/poudriere/data/101amd64-testing-testing/2015-04-20_19h53m23s/logs/errors/py27-django-tastypie-0.12.1.log — Dan Langille http://langille.org/ signature.asc Description: Message signed with OpenPGP using GPGMail
Re: Bacula without IPv6
On Apr 12, 2015, at 12:38 PM, Andrea Venturoli m...@netfence.it wrote: Hello. Running Bacula on a system where IPv6 was disabled in the kernel yields lot of errors: # bconsole Connecting to Director localhost:9101 bconsole: bsock.c:227-0 Socket open error. proto=28 port=9101. ERR=Protocol not supported ... *autodisplay on *status client ... 12-Apr 17:34 martin-xxx JobId 0: Error: bsock.c:225 Socket open error. proto=28 port=9102. ERR=Protocol not supported While these are harmless, any succesful job will report: Termination:Backup OK -- with warnings This can lead to overlooking more serious problems. Solving it all is as easy as adding --disable-ipv6 to CONFIGURE_ARGS. I'm reporting this hoping it will help others, but I also ask: can we have an option to disable IPv6, like many other ports have? That is a good idea. Please submit a PR so we do not lose track of this. I ask because it may be some time before I work on this. We'll have the default with IPV6 on. — Dan Langille http://langille.org/ signature.asc Description: Message signed with OpenPGP using GPGMail
Re: FreeBSD Port: security/py-kerberos
On Mar 27, 2015, at 6:06 PM, Roland van Laar rol...@micite.net wrote: On 27-03-15 23:02, Dan Langille wrote: On Mar 27, 2015, at 5:58 PM, Roland van Laar rol...@micite.net wrote: On 27-03-15 22:56, Roland van Laar wrote: On 27-03-15 22:39, Dan Langille wrote: On Mar 27, 2015, at 5:36 PM, Roland van Laar rol...@micite.net wrote: On 27-03-15 21:11, Dan Langille wrote: Hello, Why does security/py-kerberos require security/krb5? I ask because we already have Kerberos in base. Why do we need two installations of Kerberos? I'm confused. :/ I ask because I just added security/py-requests-kerberos to the tree. py-kerberos depends on gssapi/gssapi_generic.h. This file isn't included in the base kerberos, but it is in the security/krb5 port. I see. So we need it only for compiling, not for running? ;-) I haven't tried running it without the krb5 package. The resulting python module is: kerberos.so. Which actually depends on libkrb5support.so from security/krb5 See: # ldd ./kerberos.so ./kerberos.so: libpython2.7.so.1 = /usr/local/lib/libpython2.7.so.1 (0x801606000) libgssapi_krb5.so = /usr/local/lib/libgssapi_krb5.so (0x8019c1000) libkrb5.so = /usr/local/lib/libkrb5.so (0x801c08000) libk5crypto.so = /usr/local/lib/libk5crypto.so (0x801ee8000) libcom_err.so = /usr/local/lib/libcom_err.so (0x802117000) libc.so.7 = /lib/libc.so.7 (0x80081f000) libthr.so.3 = /lib/libthr.so.3 (0x80231a000) libintl.so.9 = /usr/local/lib/libintl.so.9 (0x80253f000) libutil.so.9 = /lib/libutil.so.9 (0x802748000) libm.so.5 = /lib/libm.so.5 (0x80295a000) libkrb5support.so = /usr/local/lib/libkrb5support.so (0x802b82000) That looks like what I found on my server after installing. I want to avoid having two Kerberos installed on the server. :) I wonder why it needs security/krb5 and can't use Kerberos from base. It seems odd to me... In what way is it odd? The base kerberos is a different one than the MIT one. py-kerberos is developed by Apple for their calendarserver software and I think they just used the MIT implementation. Here's a patch so it works with Base and with heimdal too. https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=199123 — Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
FreeBSD Port: security/py-kerberos
Hello, Why does security/py-kerberos require security/krb5? I ask because we already have Kerberos in base. Why do we need two installations of Kerberos? I'm confused. :/ I ask because I just added security/py-requests-kerberos to the tree. — Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: FreeBSD Port: security/py-kerberos
On Mar 27, 2015, at 5:36 PM, Roland van Laar rol...@micite.net wrote: On 27-03-15 21:11, Dan Langille wrote: Hello, Why does security/py-kerberos require security/krb5? I ask because we already have Kerberos in base. Why do we need two installations of Kerberos? I'm confused. :/ I ask because I just added security/py-requests-kerberos to the tree. py-kerberos depends on gssapi/gssapi_generic.h. This file isn't included in the base kerberos, but it is in the security/krb5 port. I see. So we need it only for compiling, not for running? I want to avoid having two Kerberos installed on the server. :) — Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: FreeBSD Port: security/py-kerberos
On Mar 27, 2015, at 5:58 PM, Roland van Laar rol...@micite.net wrote: On 27-03-15 22:56, Roland van Laar wrote: On 27-03-15 22:39, Dan Langille wrote: On Mar 27, 2015, at 5:36 PM, Roland van Laar rol...@micite.net wrote: On 27-03-15 21:11, Dan Langille wrote: Hello, Why does security/py-kerberos require security/krb5? I ask because we already have Kerberos in base. Why do we need two installations of Kerberos? I'm confused. :/ I ask because I just added security/py-requests-kerberos to the tree. py-kerberos depends on gssapi/gssapi_generic.h. This file isn't included in the base kerberos, but it is in the security/krb5 port. I see. So we need it only for compiling, not for running? ;-) I haven't tried running it without the krb5 package. The resulting python module is: kerberos.so. Which actually depends on libkrb5support.so from security/krb5 See: # ldd ./kerberos.so ./kerberos.so: libpython2.7.so.1 = /usr/local/lib/libpython2.7.so.1 (0x801606000) libgssapi_krb5.so = /usr/local/lib/libgssapi_krb5.so (0x8019c1000) libkrb5.so = /usr/local/lib/libkrb5.so (0x801c08000) libk5crypto.so = /usr/local/lib/libk5crypto.so (0x801ee8000) libcom_err.so = /usr/local/lib/libcom_err.so (0x802117000) libc.so.7 = /lib/libc.so.7 (0x80081f000) libthr.so.3 = /lib/libthr.so.3 (0x80231a000) libintl.so.9 = /usr/local/lib/libintl.so.9 (0x80253f000) libutil.so.9 = /lib/libutil.so.9 (0x802748000) libm.so.5 = /lib/libm.so.5 (0x80295a000) libkrb5support.so = /usr/local/lib/libkrb5support.so (0x802b82000) That looks like what I found on my server after installing. I want to avoid having two Kerberos installed on the server. :) I wonder why it needs security/krb5 and can't use Kerberos from base. It seems odd to me... — Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
removing a dir, regardless of content
A webapp, www/bacula-web puts cache files at %%BASE%%/application/view/cache and on deinstall, that directory should be removed, even though there are files there. Can that be done? I’ve tried this in pkg-plist: @unexec rm -rf %%BASE%%/application/view/cache — Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: removing a dir, regardless of content
On Dec 20, 2014, at 6:46 PM, A.J. 'Fonz' van Werven free...@skysmurf.nl
wrote:
Dan Langille wrote:
A webapp, www/bacula-web puts cache files at
%%BASE%%/application/view/cache and on deinstall, that directory should
be removed, even though there are files there.
Can that be done?
I’ve tried this in pkg-plist:
@unexec rm -rf %%BASE%%/application/view/cache
Just to be sure: did you remember to add BASE to PLIST_SUB in the
Makefile? You probably did, but I figured I'd ask anyway.
Is this what you mean?
$ grep PLIST_SUB Makefile
PLIST_SUB= WWWOWN=${WWWOWN} WWWGRP=${WWWGRP} BASE=${BASE}
—
Dan Langille
http://langille.org/
signature.asc
Description: Message signed with OpenPGP using GPGMail
Re: removing a dir, regardless of content
On Dec 20, 2014, at 8:04 PM, A.J. 'Fonz' van Werven free...@skysmurf.nl
wrote:
Dan Langille wrote:
Just to be sure: did you remember to add BASE to PLIST_SUB in the
Makefile? You probably did, but I figured I'd ask anyway.
Is this what you mean?
$ grep PLIST_SUB Makefile
PLIST_SUB= WWWOWN=${WWWOWN} WWWGRP=${WWWGRP} BASE=${BASE}
Yes. And it seems to be okay, but if I'm not mistaken pkg-plist must also
be added to SUB_FILES.
I don't have this port installed, but can you say what actually happens
when trying to uninstall it? If it is indeed a PLIST_SUB/SUB_FILES issue
then surely there must be an error message of some kind.
There is no error.
After install:
$ ls -l /usr/local/www/bacula-web/application/view/cache
total 10
-r--r--r-- 1 www www 51 Jul 29 12:23 README
Add a file:
[dan@empty ~]$ sudo touch
/usr/local/www/bacula-web/application/view/cache/things
[dan@empty ~]$ ls -l /usr/local/www/bacula-web/application/view/cache
total 10
-r--r--r-- 1 www www 51 Jul 29 12:23 README
-rw-r--r-- 1 root www 0 Dec 20 23:25 things
Delete the pkg:
[dan@empty ~]$ sudo pkg delete www/bacula-web
Checking integrity... done (0 conflicting)
Deinstallation has been requested for the following 1 packages (of 0 packages
in the universe):
Installed packages to be REMOVED:
bacula-web-6.0.1_1
The operation will free 1 MB.
Proceed with deinstalling packages? [y/N]: y
[empty] [1/1] Deinstalling bacula-web-6.0.1_1...
[empty] [1/1] Deleting files for bacula-web-6.0.1_1: 100%
And our file is still there:
[dan@empty ~]$ ls -l /usr/local/www/bacula-web/application/view/cache
total 1
-rw-r--r-- 1 root www 0 Dec 20 23:25 things
—
Dan Langille
http://langille.org/
___
freebsd-ports@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
pkg upgrade gives: Assertion failed
I am installing from my own repo onto FreeBSD 9.3 with pkg 1.4.0 $ sudo pkg upgrade -f Updating local repository catalogue... local repository is up-to-date. Updating postgresql91 repository catalogue... postgresql91 repository is up-to-date. All repositories are up-to-date. Checking for upgrades (422 candidates): 92% db47 has no direct installation candidates, change it to db48? [Y/n]: y Checking for upgrades (422 candidates): 93% db42 has no direct installation candidates, change it to db48? [Y/n]: y Checking for upgrades (422 candidates): 100% Processing candidates (422 candidates): 100% pkg: sqlite error while executing UPDATE packages SET name=?1 WHERE name=?2; in file pkg_jobs.c:1601: UNIQUE constraint failed: packages.name pkg: sqlite error while executing UPDATE packages SET name=?1 WHERE name=?2; in file pkg_jobs.c:1601: UNIQUE constraint failed: packages.name Checking integrity...pkg: cannot load files from pkg to check integrity pkg: cannot load files from zoo to check integrity pkg: cannot load files from xtail to check integrity pkg: cannot load files from xproto to check integrity pkg: cannot load files from xorg-macros to check integrity [snip] pkg: cannot load files from OpenSP to check integrity pkg: cannot load files from 3dm to check integrity Assertion failed: (pkgdb_ensure_loaded(j-db, p2, PKG_LOAD_FILES|PKG_LOAD_DIRS) == EPKG_OK), function pkg_conflicts_need_conflict, file pkg_jobs_conflicts.c, line 211. Child process pid=78554 terminated abnormally: Abort trap: 6 Ideas? — Dan Langille http://langille.org/ signature.asc Description: Message signed with OpenPGP using GPGMail
Re: pkg upgrade gives: Assertion failed
On Dec 17, 2014, at 6:06 PM, Andriy Gapon a...@freebsd.org wrote: On 18/12/2014 00:49, Dan Langille wrote: I am installing from my own repo onto FreeBSD 9.3 with pkg 1.4.0 $ sudo pkg upgrade -f Have you just switched from pkg 1.3.x? yes. If yes, please try pkg update -f before pkg upgrade. Hmmm, I usually do that every time. But, success. Thank you. Next time you see me, tell me this story and that I owe you a beer. ;) — Dan Langille http://langille.org/ signature.asc Description: Message signed with OpenPGP using GPGMail
Re: Resetting maintainership
On Dec 10, 2014, at 5:27 PM, Pietro Cerutti g...@freebsd.org wrote: All, unfortunately, I don't foresee having time to take care of my ports as they deserve in the near future. Hence, I'm dropping maintainership of all of them. Of course, I'll be occasionally patching and updating stuff here and there anyway. Here's the list of ports. Please be greedy :) That’s a whole lot of ports. Thanks for your work. Best wishes. — Dan Langille http://langille.org/ signature.asc Description: Message signed with OpenPGP using GPGMail
Re: postgresql-server depends on client. Why?
On Oct 17, 2014, at 10:44 AM, Palle Girgensohn gir...@pingpong.net wrote: Thing is, we install header files and man pages in the client only, if memory serves me right. Also, tools like createdb are in the client package. So, apart from the obvious I want to upgrade the client separately (before) the server, there are no arguments for not doing it the way we do it now. But maybe that argument is enough to motivate a change? “want” is not part of the issue. Upgrading the client before upgrading the server is standard procedure as documented by the PostgreSQL project. Palle 17 okt 2014 kl. 16:40 skrev Olli Hauer oha...@gmx.de: In case the dependency on thr client will be removed all ports depending on the server alone at the moment should be checked if they also require the client part. Additional clients should not define a conflict with server versions client version -- Sent from my Android phone with GMX Mail. Please excuse my brevity. Chris Rees cr...@physics.org wrote: Hi Dan, http://svnweb.freebsd.org/ports/head/databases/postgresql90-server/Makefile?revision=286930view=markup#l64 It was always supposed to be the case, however I agree that it's probably not necessary. Anyone mind if it doesn't depend? Chris On 17 October 2014 15:12:51 BST, Dan Langille d...@langille.org wrote: Why is postgresql-server dependent upon postgresql-client? This wasn’t the case in the past and seems to be the case not. Not all server installations need the client. In addition, upgrading the client before upgrading the server is standard procedure: http://www.postgresql.org/docs/9.3/static/upgrading.html i.e. upgrade the client, pg_dump using the new client, then upgrade server, pg_restore. Thanks — Dan Langille -- Sent from my Android device with K-9 Mail. Please excuse my brevity. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org — Dan Langille signature.asc Description: Message signed with OpenPGP using GPGMail
postgresql-server depends on client. Why?
Why is postgresql-server dependent upon postgresql-client? This wasn’t the case in the past and seems to be the case not. Not all server installations need the client. In addition, upgrading the client before upgrading the server is standard procedure: http://www.postgresql.org/docs/9.3/static/upgrading.html i.e. upgrade the client, pg_dump using the new client, then upgrade server, pg_restore. Thanks — Dan Langille signature.asc Description: Message signed with OpenPGP using GPGMail
is bsd.database.mk out of sync with Uses/pgsql.mk?
Today I encountered these build messages with poudriere: Invalid PGSQL default version 92; valid versions are 8.4 9.0 9.1 9.2 9.3 9.4 These messages originate in Mk/Uses/pgsql.mk But Mk/bsd.database.mk claims a different format is correct: # DEFAULT_PGSQL_VER # - PostgreSQL default version, currently 90. Do you agree that the comments in Mk/bsd.database.mk need to updated? — Dan Langille signature.asc Description: Message signed with OpenPGP using GPGMail
Re: vendir distfile has no top level directory
On Aug 7, 2014, at 1:08 AM, olli hauer oha...@gmx.de wrote: On 2014-08-07 05:18, Dan Langille wrote: I’m having a problem upgrading www/bacula-web to the latest version (see https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=191583). The problem: the vendor tarball has no top level directory. e.g. [root@testing ~/tmp/distfiles]# tar -xf /var/ports/distfiles/bacula-web-6.0.1.tgz [root@testing ~/tmp/distfiles]# ls DOCS application core joblogs.php pools.php README backupjob-report.phpclient-report.php index.php jobs.phptest.php [root@testing ~/tmp/distfiles]# Compare that to bacula, where I get: [root@testing ~/tmp/distfiles]# tar -xf /var/ports/distfiles/bacula-7.0.4.tar.gz [root@testing ~/tmp/distfiles]# ls bacula-7.0.4 Is there an optimal way to cater for this without pushing the issue upstream? — Hi Dan, you can specify NO_WRKSUBDIR=yes or overwrite the do-extract target. do-extract is what I will use, thank you. PS: I haven't found the 6.0.x source on sf.net only 5.0.x and 7.0.x The project seems to have moved from bacula.org to web-bacula.org and I think they host only there. — Dan Langille signature.asc Description: Message signed with OpenPGP using GPGMail
vendir distfile has no top level directory
I’m having a problem upgrading www/bacula-web to the latest version (see https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=191583). The problem: the vendor tarball has no top level directory. e.g. [root@testing ~/tmp/distfiles]# tar -xf /var/ports/distfiles/bacula-web-6.0.1.tgz [root@testing ~/tmp/distfiles]# ls DOCSapplication core joblogs.php pools.php README backupjob-report.phpclient-report.php index.php jobs.phptest.php [root@testing ~/tmp/distfiles]# Compare that to bacula, where I get: [root@testing ~/tmp/distfiles]# tar -xf /var/ports/distfiles/bacula-7.0.4.tar.gz [root@testing ~/tmp/distfiles]# ls bacula-7.0.4 Is there an optimal way to cater for this without pushing the issue upstream? — Dan Langille ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
libraries with version numbers - sysutils/bacula
I’m having trouble updating the sysutils/bacula-server port to 7.0.4. This is a packaging issue, not a ‘does it run’ issue. Regression testing is fine (see comment below). I have https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=191311 nearly finished. The blocking issue arises with src/cats/install-default-backend.in which later stops the installation of the right symlink files. short version: After bacula-server installs, we have this: [dan@testing /usr/local/lib]$ ls -l libbaccats* -rwxr-xr-x 1 root wheel 1010 Aug 3 18:20 libbaccats-postgresql.la lrwxr-xr-x 1 root wheel 26 Aug 3 18:20 libbaccats-postgresql.so - libbaccats-postgresql.so.7 -rwxr-xr-x 1 root wheel 42622 Aug 3 18:20 libbaccats-postgresql.so.7 -rwxr-xr-x 1 root wheel944 Aug 3 18:20 libbaccats.la lrwxr-xr-x 1 root wheel 15 Aug 3 18:20 libbaccats.so - libbaccats.so.7 -rwxr-xr-x 1 root wheel 6376 Aug 3 18:20 libbaccats.so.7 libbaccats.so.7 should be a symlink to libbaccats-postgresql.so — this change is done by the install-default-backend postgresql script (mentioned soon below). longer version: After installation, a grep gives this: # cd /var/ports/usr/ports/sysutils/bacula-server/work/bacula-7.0.4/src # grep -r install-default-backend * cats/Makefile: ./install-default-backend postgresql 5 $(DESTDIR)$(libdir) cats/Makefile.in: ./install-default-backend @DEFAULT_DB_TYPE@ 5 $(DESTDIR)$(libdir) cats/Makefile.in.orig: ./install-default-backend @DEFAULT_DB_TYPE@ $(LIBBACCATS_LT_RELEASE) $(DESTDIR)$(libdir) See that 5? I think it should be a 7. I have no idea where the 5 is coming from. It should be the value from LIBBACCATS_LT_RELEASE A grep for that gives (not all lines are included here): # grep -r LIBBACCATS_LT_RELEASE * cats/Makefile:LIBBACCATS_LT_RELEASE = 7.0.4 So I get confused and can’t see why. Spending more hours on this seems fruitless. I am also trying to understand the version numbers used in the Bacula port. Bacula 5.2.12_6 installs these libraries: [dan@bacula:/usr/local/lib] $ ls -l libbac* -rwxr-xr-x 1 root wheel 900 Jul 5 22:12 libbac.la lrwxr-xr-x 1 root wheel 15 Jul 5 22:12 libbac.so - libbac.so.5.0.2 lrwxr-xr-x 1 root wheel 15 Jul 5 22:12 libbac.so.5 - libbac.so.5.0.2 -rwxr-xr-x 1 root wheel 434017 Jul 5 22:12 libbac.so.5.0.2 -rwxr-xr-x 1 root wheel 984 Jul 5 22:15 libbaccats-postgresql.la lrwxr-xr-x 1 root wheel 30 Jul 5 22:15 libbaccats-postgresql.so - libbaccats-postgresql.so.5.0.2 lrwxr-xr-x 1 root wheel 30 Jul 5 22:15 libbaccats-postgresql.so.5 - libbaccats-postgresql.so.5.0.2 -rwxr-xr-x 1 root wheel 42270 Jul 5 22:15 libbaccats-postgresql.so.5.0.2 -rwxr-xr-x 1 root wheel 924 Jul 5 22:15 libbaccats.la lrwxr-xr-x 1 root wheel 19 Jul 5 22:15 libbaccats.so - libbaccats.so.5.0.2 lrwxr-xr-x 1 root wheel 24 Jul 5 22:15 libbaccats.so.5 - libbaccats-postgresql.so -rwxr-xr-x 1 root wheel6576 Jul 5 22:15 libbaccats.so.5.0.2 … etc I think 5.0.2 is derived from 5.2.16 by this patch in files/patch-src__cats__Makefile.in: - $(LIBTOOL_LINK) $(CXX) $(DEFS) $(DEBUG) $(LDFLAGS) -o $@ $(LIBBACSQL_LOBJS) -export-dynamic -rpath $(libdir) -release $(LIBBACSQL_LT_RELEASE) $(DB_LIBS) + $(LIBTOOL_LINK) $(CXX) $(DEFS) $(DEBUG) $(LDFLAGS) -o $@ $(LIBBACSQL_LOBJS) -export-dynamic -rpath $(libdir) -version-info $(LIBBACSQL_LT_RELEASE:S/./:/g:S/12/0/g) I didn’t create that substitution and I do not understand why it is useful. Can help me understand please? I suspect this approach was taken so minor upgrades do not require reinstalling other apps. Thanks. — Dan Langille signature.asc Description: Message signed with OpenPGP using GPGMail
Re: libraries with version numbers - sysutils/bacula
On Aug 3, 2014, at 2:26 PM, Dan Langille d...@langille.org wrote: See that 5? I think it should be a 7. I have no idea where the 5 is coming from. It should be the value from LIBBACCATS_LT_RELEASE A grep for that gives (not all lines are included here): # grep -r LIBBACCATS_LT_RELEASE * cats/Makefile:LIBBACCATS_LT_RELEASE = 7.0.4 With help from madpilot, I found that five, hardcoded in a patch file. Fixed. Thanks. — Dan Langille signature.asc Description: Message signed with OpenPGP using GPGMail
Help wanted with sysutils/bacula-server 7.0.4
I need help please with upgrading sysutils/bacula-server to 7.0.4 The latest patch, incomplete, is in the PR: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=191311#c15 The current problems: * the shared libraries installed by bacula-client are not correct: # ls -l /usr/local/lib/libbacfind* -rwxr-xr-x 1 root wheel 950 Jul 31 19:57 /usr/local/lib/libbacfind.la lrwxr-xr-x 1 root wheel 19 Jul 31 19:57 /usr/local/lib/libbacfind.so - libbacfind.so.7.0.0 lrwxr-xr-x 1 root wheel 19 Jul 31 19:57 /usr/local/lib/libbacfind.so.7 - libbacfind.so.7.0.0 I expected there to be just .7, not 7.0.0 They are built: [root@testing /var/ports/usr/ports.SVN/sysutils/bacula-client/work/bacula-7.0.4/src/lib/.libs]# ls -l libbac.so* lrwxr-xr-x 1 root wheel 15 Jul 31 20:20 libbac.so - libbac.so.7.0.0 lrwxr-xr-x 1 root wheel 15 Jul 31 20:20 libbac.so.7 - libbac.so.7.0.0 -rwxr-xr-x 1 root wheel 446132 Jul 31 20:20 libbac.so.7.0.0 I can’t figure out why they don’t get installed. More eyes welcomed. — Dan Langille signature.asc Description: Message signed with OpenPGP using GPGMail
poudriere and local patches to ports
I have a local patch (https://dan.langille.org/wp-content/uploads/2014/04/nagios-default-comment-cmd-cgi.c.diff_.txt) which I wish to apply to net-mgmt/nagios before it is compiled by poudriere). I have looked at ports-mgmt/portshaker but that seems to be related to merging rather than patching. Any suggestions? — Dan Langille signature.asc Description: Message signed with OpenPGP using GPGMail
databases/postgresql92-server does not honor uid/gid
FYI: I upgraded databases/postgresql92-server a system today. Afterwards, I encountered this situation: $ sudo /usr/local/etc/rc.d/postgresql initdb The files belonging to this database system will be owned by user pgsql. This user must also own the server process. The database cluster will be initialized with locale C. The default database encoding has accordingly been set to SQL_ASCII. The default text search configuration will be set to english. creating directory /usr/local/pgsql/data ... initdb: could not create directory /usr/local/pgsql/data: Permission denied Oh… u… $ ls -ld /usr/local/pgsql drwxr-xr-x 2 70 70 6 Jun 9 21:03 /usr/local/pgsql eh? $ id pgsql uid=10837(pgsql) gid=10837(pgsql) groups=10837(pgsql) On this system, a jail. the uid/gid for pgsql are not standard. I solved the problem with: $ sudo chown pgsql:pgsql /usr/local/pgsql/ — Dan Langille signature.asc Description: Message signed with OpenPGP using GPGMail
Ports moved but without CATEGORIES update
FreshPorts told me about this situation. A few ports did not have their CATEGORIES value updated. Thus, they are not in their CATEGORY. The ports are: id | category_id | name | category ---+-+--+ 32308 | 91 | kde4-base-artwork| x11-themes 32311 | 52 | dragon | multimedia 32313 | 10 | p5-perlkde | devel 32315 | 25 | libkcompactdisc | audio 32318 | 10 | smokekde | devel 32320 | 25 | libkcddb | audio 32321 | 25 | kscd | audio 32322 | 3 | pairs| games 32323 | 10 | ruby-korundum| devel 32326 | 8 | krdc | net 32327 | 95 | kopete-kde4 | net-im 32332 | 20 | nepomuk-core | sysutils 32309 | 25 | kmix | audio 32310 | 10 | p5-perlqt| devel 32312 | 52 | kdemultimedia4-mplayerthumbs | multimedia 32314 | 10 | py-pykde4| devel 32316 | 10 | smokeqt | devel 32317 | 10 | ruby-krossruby | devel 32319 | 25 | kio-audiocd | audio 32324 | 10 | ruby-qtruby | devel 32325 | 10 | py-pykdeuic4 | devel 32328 | 52 | kdemultimedia4-ffmpegthumbs | multimedia 32329 | 8 | krfb | net 32330 | 10 | smokegen | devel 32331 | 10 | py-krosspython | devel (25 rows) Sorry that this list is not in a more useful format. -- Dan Langille - http://langille.org ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Ports moved but without CATEGORIES update
Hmm, checking manually, it appears that at least ONE of these is a false positive. x11-themes/kde4-base-artwork is set correctly. Sorry. On Feb 3, 2013, at 8:26 PM, Dan Langille wrote: FreshPorts told me about this situation. A few ports did not have their CATEGORIES value updated. Thus, they are not in their CATEGORY. The ports are: id | category_id | name | category ---+-+--+ 32308 | 91 | kde4-base-artwork| x11-themes 32311 | 52 | dragon | multimedia 32313 | 10 | p5-perlkde | devel 32315 | 25 | libkcompactdisc | audio 32318 | 10 | smokekde | devel 32320 | 25 | libkcddb | audio 32321 | 25 | kscd | audio 32322 | 3 | pairs| games 32323 | 10 | ruby-korundum| devel 32326 | 8 | krdc | net 32327 | 95 | kopete-kde4 | net-im 32332 | 20 | nepomuk-core | sysutils 32309 | 25 | kmix | audio 32310 | 10 | p5-perlqt| devel 32312 | 52 | kdemultimedia4-mplayerthumbs | multimedia 32314 | 10 | py-pykde4| devel 32316 | 10 | smokeqt | devel 32317 | 10 | ruby-krossruby | devel 32319 | 25 | kio-audiocd | audio 32324 | 10 | ruby-qtruby | devel 32325 | 10 | py-pykdeuic4 | devel 32328 | 52 | kdemultimedia4-ffmpegthumbs | multimedia 32329 | 8 | krfb | net 32330 | 10 | smokegen | devel 32331 | 10 | py-krosspython | devel (25 rows) Sorry that this list is not in a more useful format. -- Dan Langille - http://langille.org -- Dan Langille - http://langille.org ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: pkg is segmentation fault with of the cross-compile port N2
On Jul 4, 2012, at 10:09 AM, Ivan Klymenko fi...@ukr.net wrote: В Wed, 4 Jul 2012 10:08:26 + Baptiste Daroussin b...@freebsd.org пишет: On Wed, Jul 04, 2012 at 01:01:44PM +0300, Ivan Klymenko wrote: В Wed, 4 Jul 2012 09:50:35 + Baptiste Daroussin b...@freebsd.org пишет: Fixed in git, i'll backport it to the port asap. Thanks. btw this bug only concerne i386 I unfortunately can not check it at the native i386 system. I have this error when I do build i386 packages in the chroot in the parent amd64 system... When I build the packages in a chroot amd64 inside the parent adm64 - then all is well... Fixed. (Can you confirm?) regards, Bapt Problem no solved :( http://privatepaste.com/5e066b87f5 Thanks all. Please commit whatever fixes. -- Dan Langille http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
graphics/ImageMagick fails to build
revision 2 or lat ruby18-iconv-1.8.7.352,1 An iconv wrapper class for Ruby sudo-1.8.3_1Allow others to run commands as root xcb-proto-1.6 The X protocol C-language Binding (XCB) protocol xcmiscproto-1.2.1 XCMisc extension headers xextproto-7.2.0 XExt extension headers xf86bigfontproto-1.2.0 XFree86-Bigfont extension headers xorg-macros-1.15.0 X.Org development aclocal macros xproto-7.0.22 X11 protocol headers xtail-2.1 Watches the growth of files or directories xtrans-1.2.6Abstract network code for X -- Dan Langille - http://langille.org ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: What ports require this port?
On Mon, February 7, 2011 11:06 am, Dominic Fandrey wrote: On 07/02/2011 03:01, b. f. wrote: Over the weekend, a new feature was added to FreshPorts BETA. A port now lists all ports that require it. This code will eventually move to production after it's been running on on the beta site for a while. Notes: * The code uses the output of make -V RUN_DEPENDS -V LIB_DEPENDS -V BUILD_DEPENDS If you're going to include BUILD_DEPENDS, then you should also use -V PATCH_DEPENDS -V FETCH_DEPENDS. Maybe it should be split into The package requires only listing the INSTALL_DEPENDS and The port requires listing all dependencies. Your post, combined with new information, has raised a question. Consider 'make build-depends-list' and 'make run-depends-list' for devel/glib20: $ make build-depends-list /usr/ports/devel/gettext /usr/ports/devel/gmake /usr/ports/devel/libtool /usr/ports/devel/pcre /usr/ports/devel/pkg-config /usr/ports/lang/perl5.8 /usr/ports/lang/python26 $ make run-depends-list /usr/ports/devel/gettext /usr/ports/devel/pcre /usr/ports/devel/pkg-config /usr/ports/lang/perl5.8 /usr/ports/lang/python26 The run-depends-list contains all of the information from LIB_DEPENDS and RUN_DEPENDS. The build-depends-list includes all of BUILD_DEPENDS, PATCH_DEPENDS, FETCH_DEPENDS, and EXTRACT_DEPENDS. I am tempted to alter FreshPorts and abandon all *_DEPENDS and use just run-depends-list and build-depends-list. How much value is there is having all of this information separated into individual _DEPENDS? -- Dan Langille -- http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
MOVED entry needed for www/p5-Apache2-Scorebord
According to http://www.freshports.org/www/p5-Apache2-Scorebord, this port was repo copied to http://www.freshports.org/www/p5-Apache2-Scoreboard on 17 Jun 2005 12:10:08 (change of name from Scorebord to Scoreboard) There is no such entry in MOVED. This entry should do it: www/p5-Apache2-Scorebord|www/p5-Apache2-Scoreboard|2005-06-12|repo copy to correct name -- Dan Langille - http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
What ports require this port?
Over the weekend, a new feature was added to FreshPorts BETA. A port now lists all ports that require it. This code will eventually move to production after it's been running on on the beta site for a while. Notes: * The code uses the output of make -V RUN_DEPENDS -V LIB_DEPENDS -V BUILD_DEPENDS * Only the default build is queried * This is not recursive. For example, at http://beta.freshports.org/graphics/png/ You will find: Required by: for Build graphics/iiview graphics/optar graphics/tumbler for Libraries astro/astrometry astro/celestia astro/celestia-gnome [long list snipped] More information: * The original request: http://www.freshports.org/phorum/read.php?f=1i=1280t=1280 * Some background on the work: http://news.freshports.org/2011/02/05/what-ports-are-dependant-upon-this-port/ or http://tinyurl.com/6fezxpq -- Dan Langille - http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: What ports require this port?
On 2/6/2011 9:01 PM, b. f. wrote: Over the weekend, a new feature was added to FreshPorts BETA. A port now lists all ports that require it. This code will eventually move to production after it's been running on on the beta site for a while. Notes: * The code uses the output of make -V RUN_DEPENDS -V LIB_DEPENDS -V BUILD_DEPENDS If you're going to include BUILD_DEPENDS, then you should also use -V PATCH_DEPENDS -V FETCH_DEPENDS. I guess that's a job for future work. Another weekend. :) Thanks. -- Dan Langille - http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: What ports require this port?
On 2/6/2011 9:03 PM, b. f. wrote: On 2/6/11, b. f.bf1...@googlemail.com wrote: Over the weekend, a new feature was added to FreshPorts BETA. A port now lists all ports that require it. This code will eventually move to production after it's been running on on the beta site for a while. Notes: * The code uses the output of make -V RUN_DEPENDS -V LIB_DEPENDS -V BUILD_DEPENDS If you're going to include BUILD_DEPENDS, then you should also use -V PATCH_DEPENDS -V FETCH_DEPENDS. And, of course, -V EXTRACT_DEPENDS. While we're here... is there anything special about PATCH_DEPENDS, FETCH_DEPENDS, and EXTRACT_DEPENDS that might be different than the depends I'm already processing? They seem to be the same format: [dan@ngaio:/usr/ports/graphics/tifmerge] $ make -V EXTRACT_DEPENDS tuc:/usr/ports/converters/tuc /usr/local/bin/unzip:/usr/ports/archivers/unzip [dan@ngaio:/usr/ports/www/apache13-modssl] $ make -V PATCH_DEPENDS /usr/local/bin/perl5.8.9:/usr/ports/lang/perl5.8 /usr/local/bin/perl5.8.9:/usr/ports/lang/perl5.8 However, my manual searches to find a port which outputs a value for 'make -V FETCH_DEPENDS' have failed so far. But I'm sure it will turn up during my first few test runs. -- Dan Langille - http://langille.org/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: What ports require this port?
On 2/6/2011 9:50 PM, b. f. wrote:
On 2/6/11, Dan Langilled...@langille.org wrote:
On 2/6/2011 9:03 PM, b. f. wrote:
On 2/6/11, b. f.bf1...@googlemail.com wrote:
While we're here... is there anything special about PATCH_DEPENDS,
FETCH_DEPENDS, and EXTRACT_DEPENDS that might be different than the
depends I'm already processing? They seem to be the same format:
[dan@ngaio:/usr/ports/graphics/tifmerge] $ make -V EXTRACT_DEPENDS
tuc:/usr/ports/converters/tuc
/usr/local/bin/unzip:/usr/ports/archivers/unzip
[dan@ngaio:/usr/ports/www/apache13-modssl] $ make -V PATCH_DEPENDS
/usr/local/bin/perl5.8.9:/usr/ports/lang/perl5.8
/usr/local/bin/perl5.8.9:/usr/ports/lang/perl5.8
IIRC, they are the same ordered triple format. By the way, just
because a port is in the *_DEPENDS of another port, doesn't
necessarily mean that the first port must be installed to build the
second port: sometimes the (usually absent) third element of the
triplet can be used to override the default install. So some
*_DEPENDS, for example, just need to be fetched rather than installed.
Others may just need to execute some simple target defined in the
Makefile of the dependency. I don't know whether you want to include
that level of detail in your Freshports pages.
I think you are referring to suffixes such as :patch, :fetch, :config.
FreshPorts ignores those entirely. It just lists the ports regardless
of those suffixes.
However, my manual searches to find a port which outputs a value for
'make -V FETCH_DEPENDS' have failed so far. But I'm sure it will turn
up during my first few test runs.
fgrep -e FETCH_DEPENDS -nH -r $PORTSDIR --include='*Makefile':
ports/cad/kicad-devel/Makefile:83:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/cad/kicad-devel/Makefile:84:FETCH_DEPENDS+=
bzr:${PORTSDIR}/devel/bazaar-ng
ports/comms/wsjt/Makefile:66:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/devel/compiler-rt/Makefile:62:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/devel/llvm-devel/Makefile:89:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/graphics/seom/Makefile:47:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/graphics/yukon/Makefile:44:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/lang/clang-devel/Makefile:75:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/net/xorp-devel/Makefile:45:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/print/bsd_lpr/Makefile:33:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/www/trac-discussion/Makefile:31:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/www/trac-fivestarvote/Makefile:30:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/www/trac-math/Makefile:37:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/www/trac-ticketimport/Makefile:31:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
ports/www/trac-vote/Makefile:30:FETCH_DEPENDS+=
svn:${PORTSDIR}/devel/subversion
Find a Makefile containing FETCH_DEPENDS is one thing. Getting a value
from the default config settings is another. :) I wasn't explicit about
that, but I chose the words carefully.
--
Dan Langille - http://langille.org/
___
freebsd-ports@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
