Bug#920248: pg_upgradecluster fails with inadequate error message

[Daniel Pocock]

Package: postgresql-common
Version: 181+deb9u2
Severity: important



I was upgrading a few PostgreSQL servers.  Most were successful, one of
them would fail with this error:

# pg_upgradecluster 9.4 main
Error: could not get cluster locales



I checked locale settings on the host (locale ; locale -a), in the
database (show all) and in postgresql.conf.  Nothing obviously wrong.  I
used
 dpkg-reconfigure locales
to install all possible locales too but this didn't resolve the issue.

Web searches find many examples of this error.  Example:

https://serverfault.com/questions/909624/pg-upgradecluster-fails-with-error-could-not-get-cluster-locales?rq=1
"sometimes, in Ubuntu systems, if ssh login user's locale is different"

but that was also a red herring.

https://www.postgresql.org/message-id/CABnTomKNDj%2BorMgyPaPtFkNoBFCuSWnzSAd0X0%2BX4LHSgUbEDQ%40mail.gmail.com

also suggests running pg_upgrade manually.

Another page suggested some users made it work if they manually stop the
service with systemctl before running pg_upgradecluster, that didn't
help me either.

I decided to trace the scripts:

  apt install libdevel-trace-perl

  cp /usr/bin/pg_upgradecluster /tmp

  perl -d:Trace -T /tmp/pg_upgradecluster 9.4 main

The trace output helped me identify where the script was really failing
and add some debugging / print statements.

In my case, I found that it was failing to find the port number.  In
postgresql.conf, the "port" parameter was commented on this host, so it
would use the default, 5432

This leads me to two conclusions:

- it needs to handle these specific cases (e.g. port number) automatically

- it needs to give more accurate errors - in my case, the error "could
not get cluster locales" was misleading for me

Regards,

Daniel

Bug#783346: Bug #783346: patch for booting Jessie domU with wheezy dom0

[Daniel Pocock]

On 20/01/2019 19:01, Hans van Kranenburg wrote:
> reassign 783346 src:xen 4.1.4-3+deb7u4
> severity 783346 normal
> thanks
> 
> Hi Daniel,
> 
> This is about the bug report that you filed before about pygrub in the
> Debian bug tracker against the Xen packages.
> 
> Your bug report was targeted at a Xen package in a Debian distribution
> older than the current stable (Stretch), and I'm trying to clean up our
> open bug report list a bit.
> 
> In the bug history, I see that already was suggested to use the grub2
> based way of booting the virtual machines instead of pygrub. This is
> still recommended.
> 
> Addtionally, the grub-xen-* packages in Debian Buster will now also be
> able to boot a Xen domU using the new PVH virtualization mode, if you
> use the Debian Linux 4.19 kernel package inside.
> 
> Am I right to understand that we can close this bug report now?
> 

If you are still distributing the script with the bug, then I would
recommend:

a) check if the patch is now included upstream, if yes, close the bug, or

b) include the patch in the Debian package, or

c) stop installing the pygrub script in the package so that users have
to use pygrub2 or something else.

However, I migrated from Xen to KVM so I'm not using the package
personally any more.

Regards,

Daniel

Bug#899058: groups

[Daniel Pocock]

For the Zigate and maybe some other devices, the domoticz user needs to
be added to the group dialout

Should the postinst script for the relevant plugins do this, or should
it be done by domoticz.postinst?

Bug#899058: Python issue

[Daniel Pocock]

On another installation today, couldn't find the Zigate (Python-based)
plugin in the Add Hardware pulldown


I looked at /var/log/daemon.log

domoticz[55195]: Status: EventSystem - Python: Failed dynamic library
load, install the latest libpython3.x library that is available for your
platform.



Resolved by

$ sudo apt install libpython3-dev

which installed libpython3.5-dev.


A -dev package probably shouldn't be a runtime dependency though, is it
needed for plugins, or can this be satisfied in another way?

Bug#916972: ITP: domoticz-plugin-zigate -- Domoticz plugin for the Zigate.fr Zigbee transceiver

[Daniel Pocock]

Package: wnpp
Severity: wishlist
X-Debbugs-CC: patr...@pichon.me

The Domoticz home automation software is currently being prepared for
Debian:

https://bugs.debian.org/899058

The Zigate.fr USB stick is an inexpensive transceiver that allows
Domoticz to communicate directly with Zigbee devices (lamps, switches,
smart sockets, etc) without using a proprietary hub or bridge.

This plugin allows Domoticz to use the USB stick.

Upstream are coordinating fixes related to packaging in this issue:

https://github.com/sasu-drooz/Domoticz-Zigate/issues/299

License: GPL-3

https://github.com/sasu-drooz/Domoticz-Zigate/blob/stable/LICENSE.txt

I've tested the Domoticz package and this plugin on a Raspberry Pi 3
running Raspbian stretch.

I'll start maintaining a debian/sid packaging branch in this repository
shortly and post updates in the ITP as it becomes ready to use

https://gitlab.com/dpocock/domoticz-zigate

Regards,

Daniel

--
Debian Developer
https://danielpocock.com

Bug#899058: building for Raspbian on a Debian stretch host

[Daniel Pocock]

Here are the steps I followed to build Domoticz for Raspbian stretch
using a Debian workstation

Trying to compile on the Raspberry Pi fails due to insufficient RAM. 
The workstation has more RAM and cores for a much faster compile.




sudo apt install sbuild ubuntu-dev-tools qemu-user-static binfmt-support
aufs-dkms


If the aufs-dkms module fails to build with dkms, obtain it from Git. 
For example, if running a backports kernel 4.17:


mkdir -p ~/ws/aufs && cd ~/ws/aufs
git clone https://salsa.debian.org/janluca-guest/aufs-debian
git checkout debian/4.17+20180827-1
dpkg-buildpackage -rfakeroot -i.git -j24 -b --no-sign
sudo dpkg -i ../aufs-dkms_4.17+20180827-1_amd64.deb


Now the tools are ready:


mk-sbuild --arch=armhf --debootstrap-no-check-gpg
--debootstrap-mirror=http://archive.raspbian.org/raspbian stretch
su - $USER
mk-sbuild --arch=armhf --debootstrap-no-check-gpg
--debootstrap-mirror=http://archive.raspbian.org/raspbian stretch


mkdir -p ~/ws/domoticz
cd ~/ws/domoticz
wget https://archive.raspbian.org/raspbian.public.key
wget https://github.com/domoticz/domoticz/archive/4.9700.tar.gz
ln -s 4.9700.tar.gz domoticz_4.9700.orig.tar.gz
git clone https://gitlab.com/dpocock/domoticz-debian-raspbian/
cd domoticz-debian-raspbian
git checkout debian/backports/stretch
sbuild -A -d stretch-armhf --host=armhf --build=armhf
--extra-repository-key=`readlink -f ../raspbian.public.key`
--no-apt-update --no-apt-distupgrade -j24 .



To install the package on the Pi:

scp ../domoticz_4.9700-1_armhf.deb raspberrypi:/root
ssh pi@raspberrypi
sudo dpkg -i /root/domoticz_4.9700-1_armhf.deb




To install the Zigate plugin on the Pi:


sudo rm /usr/sbin/plugins
sudo ln -s /usr/lib/domoticz/plugins /usr/sbin/plugins   (Bug
workaround, or use the patch mentioned above)
cd /usr/lib/domoticz/plugins
git clone https://github.com/sasu-drooz/Domoticz-Zigate
cd Domoticz-Zigate
git checkout stable


Regards,

Daniel


--
Debian Developer
https://danielpocock.com

Bug#899058: incomplete plugins path

[Daniel Pocock]

The binary appears to be compiled with the relative path "plugins/" for
plugins

The working directory is /usr/sbin and it creates an empty directory
/usr/sbin/plugins at startup

It doesn't use /usr/lib/domoticz/plugins so it doesn't discover any
plugins installed there.

As a workround, I created a symlink before starting the process,

    ln -s /usr/lib/domoticz/plugins /usr/sbin/plugins

but that is only a hack

Bug#899058: disable upstream version popup

[Daniel Pocock]

There is a popup in the web interface announcing new upstream versions,
the text is "A new version of Domoticz is Available!"

This should probably be disabled for the package, as updates are managed
through Debian.

Is there a config option to disable the update check or does it require
a patch?

>From a privacy perspective it is also not a good idea for a package to
make un-necessary connections to upstream

Regards,

Daniel

Bug#914340: [DAViCal-devel] Bug#914340: system details

[Daniel Pocock]

On 04/12/2018 08:41, Florian Schlichting wrote:

> Do you want to try with openssl 1.1.0j-1~deb9u1 (make sure to restart
> apache2 so all the children use the new libraries) and see if those
> segfaults disappear?
>


Thanks for looking into that

I installed the updated openssl

Previously Apache / DAViCal would get into this state approximately
once a week, so if it runs smoothly for 2-3 weeks then hopefully that is
the fix for it.

That said, even if a segfault appears in openssl and there is a fix for
that, there are sometimes cases where another library mentioned in the
stack has used the openssl API incorrectly or passed a bad argument and
ideally that code would be examined too.  I haven't personally had time
to look closely enough at the openssl fix to see if it is such a case.

Regards,

Daniel

Bug#549655: ganglia-webfrontend - new version

[Daniel Pocock]

Hi,

Thanks for feedback on these bugs

There have been many new upstream releases, fixing a lot of bugs, but
they haven't been packaged yet because nobody has had time to package a
couple of new Javascript dependencies.

It was also suggested that the FTP masters might accept the new
Javascript as long as the ganglia-webfrontend source package includes
the unminified version of each.

If anybody wants to volunteer to help with that it would be very welcome.

Regards,

Daniel

Bug#914340: system details

[Daniel Pocock]

Here are some of the package versions that are installed:

ii  apache22.4.25-3+deb9u6

ii  davical1.1.7-1~bpo9+1

ii  dbconfig-pgsql 2.0.8

ii  libapache2-mod-php 1:7.0+49

ii  libapache2-mod-php7.0  7.0.30-0+deb9u1

ii  openssl1.1.0f-3+deb9u2

ii  libssl1.1:amd641.1.0f-3+deb9u2

ii  php1:7.0+49

ii  php-pgsql  1:7.0+49

ii  php7.0 7.0.30-0+deb9u1

ii  php7.0-pgsql   7.0.30-0+deb9u1


The pdo_pgsql module comes from package php7.0-pgsql:


# grep pdo_pgsql /var/lib/dpkg/info/*.list
/var/lib/dpkg/info/davical-doc.list:/usr/share/doc/davical-doc/api/function-check_pdo_pgsql.html
/var/lib/dpkg/info/davical-doc.list:/usr/share/doc/davical-doc/api/source-function-check_pdo_pgsql.html
/var/lib/dpkg/info/libphp-adodb.list:/usr/share/php/adodb/drivers/adodb-pdo_pgsql.inc.php
/var/lib/dpkg/info/php7.0-pgsql.list:/usr/lib/php/20151012/pdo_pgsql.so
/var/lib/dpkg/info/php7.0-pgsql.list:/usr/share/php7.0-pgsql/pgsql/pdo_pgsql.ini

Bug#914340: [DAViCal-devel] Bug#914340: webserver sometimes needs to be restarted

[Daniel Pocock]

On 22/11/2018 23:25, Florian Schlichting wrote:
> Hi Daniel,
> 
> On Thu, Nov 22, 2018 at 12:35:59PM +0100, Daniel Pocock wrote:
>> Warning: CalDAV: No response status doing webdav sync for calendar foo
>> Warning: CalDAV: Error doing webdav sync: undefined
>> Warning: There has been an error reading data for calendar: foo.
>> However, this error is believed to be minor, so the program will attempt
>> to continue. Error code: DAV_REPORT_ERROR. Description: There has been
>> an error reading data for calendar:
>> https://foo.example.org/davical/caldav.php/foo/home/. It has been
>> disabled until it is safe to use it.
>> Warning: There has been an error reading data for calendar: foo.
>> However, this error is believed to be minor, so the program will attempt
>> to continue. Error code: READ_FAILED. Description:
> 
> I don't remember ever seeing this from Thunderbird for calendars on a
> DAViCal server.
> 


On DAVdroid, this is the error:

12-03 10:36:59.599  2613   471 W davdroid: [syncadapter.SyncManager] I/O
exception during sync, trying again later
12-03 10:36:59.599  2613   471 W davdroid: EXCEPTION
java.io.IOException: unexpected end of stream on
Connection{dav.example.org:443, proxy=DIRECT
hostAddress=dav.example.org/10.1.2.3:443
cipherSuite=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 protocol=http/1.1}




>> Simply logging into the web server and doing
>>
>>   systemctl restart apache2
>>
>> resolves the issue and all the clients work again for another week or so.
>>
>> Has anybody else observed this?
>>
>> Where should I look for additional logging or clues next time this
>> happens?  Should I enable any extra logging options in DAViCal, Apache
>> or PostgreSQL?
>
> Well I'd start with the Apache logs: when TB says "no response status",
> what does Apache think happened with the request, which status did it
> log, anything in the error log?
>
> A default install of the davical package pulls in mod_php, which AFAIK
> doesn't keep state between requests; do you use a different PHP
> interpreter, or something like a Zend server or memcached? Nothing in
> the way davical processes or responds to a request should change from a
> webserver restart, really...
>
> If you still think davical may be at fault and want to dig deeper, have
> a look at https://wiki.davical.org/index.php/Debugging - I'd recommend
> setting $c->dbg["ALL"] = 1; and perhaps limiting things to a single
> remote IP. The result is fairly verbose but should show you what the
> requests and responses look like, and will provide enough detail to pin
> down where things go south on the server side.




There is nothing in Apache access.log or error.log for this particular
virtual host.

I press sync once in DAVdroid and this appears 3 times in the main
error.log:

[Mon Dec 03 09:42:19.538805 2018] [core:notice] [pid 13779] AH00052:
child pid 5106 exit signal Segmentation fault (11)
[Mon Dec 03 09:42:19.538919 2018] [core:notice] [pid 13779] AH00052:
child pid 5113 exit signal Segmentation fault (11)
[Mon Dec 03 09:42:19.538945 2018] [core:notice] [pid 13779] AH00052:
child pid 5115 exit signal Segmentation fault (11)

Nothing from journalctl

I made a capture with tcpdump as well, I notice the server has sent "New
Session Ticket" and the client has sent three "Application Data" packets
and then the server closes the connection

I did

  apt install systemd-coredump

and used prlimit to set core size unlimited for the running process but
I can't find any core dumps yet, maybe the apache2 needs to be
restarted, but then the problem will not happen again for another week.

I also tried connecting to the main process with

   gdb /usr/sbin/apache2 13779

but that doesn't show any output when the seg fault happens.


Then I tried guessing which child process would crash next and connected
to it with gdb, looks like an issue in pdo_pgsql.so with PHP 7, stack is
below.

Should the bug be reassigned to the php-pgsql package?  As we have a
reproducible stack trace I feel it is now an RC issue too.

Is there any other data I should gather before restarting the Apache
process again?  I can leave it like this for a couple of hours maximum.


Thread 1 "apache2" received signal SIGSEGV, Segmentation fault.
0x7fa662662ed4 in ERR_clear_error () from
target:/usr/lib/x86_64-linux-gnu/libcrypto.so.1.1
(gdb) bt
#0  0x7fa662662ed4 in ERR_clear_error () from
target:/usr/lib/x86_64-linux-gnu/libcrypto.so.1.1
#1  0x7fa66267dfe9 in ?? () from
target:/usr/lib/x86_64-linux-gnu/libcrypto.so.1.1
#2  0x7fa66bdf5739 in __pthread_once_slow
(once_control=0x7fa6629a72d4, init_routine=0x7fa66267dfe0) at
pthread_once.c:116
#3  0x7fa6626d3f59 in CRYPTO_THREAD_run_once () from
target:/usr/lib/x86_64-linu

Bug#912143: fixed upstream

[Daniel Pocock]

https://github.com/resiprocate/resiprocate/commit/9a3abc0118e40805dd93a9ab2e6852ea640a0f64


Need to add a macro (upstream) to support backport builds on Debian
systems running stretch / asio 1.10

Bug#897841: [Pkg-middleware-maintainers] Bug#897841: Patch upstream doesn't fix

[Daniel Pocock]

On 20/07/18 11:38, Thomas Goirand wrote:
> Hi,
> 
> I tried applying the patch from upstream, unfortunately, the build still
> fails at the same moment. Maybe we should just disable this -Werror thing?
> 

Thanks for testing it

I notice that the fix didn't change every file with this problem, it
needed to be changed on 5 more lines to get a successful build.

Regards,

Daniel

Bug#914340: webserver sometimes needs to be restarted

[Daniel Pocock]

Package: davical
Version: 1.1.5-1
Severity: important

>From time to time, the clients fail to sync with DAViCal.

Many of the clients don't indicate there is an error.  For example, the
Lightning plugin for Thunderbird logs errors like this to the console
where ordinary users would never see them:


Warning: CalDAV: No response status doing webdav sync for calendar foo
Warning: CalDAV: Error doing webdav sync: undefined
Warning: There has been an error reading data for calendar: foo.
However, this error is believed to be minor, so the program will attempt
to continue. Error code: DAV_REPORT_ERROR. Description: There has been
an error reading data for calendar:
https://foo.example.org/davical/caldav.php/foo/home/. It has been
disabled until it is safe to use it.
Warning: There has been an error reading data for calendar: foo.
However, this error is believed to be minor, so the program will attempt
to continue. Error code: READ_FAILED. Description:




but it doesn't show any error in the GUI until somebody tries to do
something like dismissing a reminder.  Other clients, like DAVdroid,
just fail to sync without showing any feedback.

Simply logging into the web server and doing

  systemctl restart apache2

resolves the issue and all the clients work again for another week or so.

Has anybody else observed this?

Where should I look for additional logging or clues next time this
happens?  Should I enable any extra logging options in DAViCal, Apache
or PostgreSQL?

Bug#899058: Raspbian build?

[Daniel Pocock]

Has anybody tried to build the package for Raspbian on the Raspberry Pi already?

Are all the dependencies already available there?

Bug#901890: Bug#901889: displays "Page not found" error when user not logged in

[Daniel Pocock]

On 21/06/18 08:29, Daniel Lange wrote:
> Am 20.06.2018 um 23:05 schrieb Daniel Pocock:
>> What is the package / software being used for the current DebConf?
> 
> A Django application called wafer:
> https://github.com/CTPUG/wafer
> 


Thanks

#901889 appears to be a generic application issue so I opened it in that
Github project.

The other bugs appear to be relevant to any DebConf, not just DebConf18

Is there a generic issue tracker for DebConf's implementation of wafer,
for issues that are not DebConf18 specific?

Or would it be reasonable to put one or both of those bugs in wafer's
bug tracker too?

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=901890
- ask people if they require assistance buying tickets

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=901891
- ask people for visa letter details

In the case of the latter, it looks like it was already opened in the
DebConf18 issue tracker:

https://salsa.debian.org/debconf-team/public/websites/dc18/issues/43

but not implemented hence the last minute email sent around asking
people to send information needed by the visa team.

Bug#901889: displays "Page not found" error when user not logged in

[Daniel Pocock]

On 20/06/18 23:00, Julien Cristau wrote:
> Summit was used for past debconfs so is irrelevant for future ones.


What is the package / software being used for the current DebConf?

Bug#901889: displays "Page not found" error when user not logged in

[Daniel Pocock]

On 20/06/18 21:17, Julien Cristau wrote:
> On Tue, Jun 19, 2018 at 22:37:59 +0200, Daniel Pocock wrote:
> 
>> Package: summit.debconf.org
>>
> This is not the right place to record issues about the debconf18 site.
> Please use
> https://salsa.debian.org/debconf-team/public/websites/dc18/issues
> 



These are issues for any future DebConf, they are not specific to
DebConf18, so why should they go into the DebConf18 issue tracker?
Won't they just get lost after DebConf18 and we'll have the same
problems again next year?

Bug#901890: ask people if they require assistance buying tickets

[Daniel Pocock]

Package: summit.debconf.org
Severity: wishlist

Some attendees (for example, GSoC/Outreachy interns) don't always have
credit cards or don't have a credit limit high enough to purchase a flight.

On the page asking for the amount required, it would be useful to add an
extra menu with a couple of options:

* I can pay for flights using my own credit card (default)
* I don't have a credit card or the available limit may be insufficient

Bug#901891: ask people for visa letter details

[Daniel Pocock]

Package: summit.debconf.org
Severity: wishlist

During the registration process, ask the user if they require a visa

Offer a link (customizable for each year/event) to the official site for
checking visa requirements.

If they select the option "visa required", ask for any details required
by the visa team, example[1], so that extra requests for information
don't need to be made later.

Extra care may be needed to ensure sensitive information is not stored
unless necessary.



1. https://lists.debian.org/debconf-announce/2018/06/msg1.html

Bug#901889: displays "Page not found" error when user not logged in

[Daniel Pocock]

Package: summit.debconf.org

If a user is not logged in and they click the link at the bottom of the
bursary status email, e.g.


 https://debconf18.debconf.org/users/some_user


the "Page not found" error appears with the text "The page you requested
cannot be found"

In this case it should probably show a prompt for the user to log in.

Bug#892321: attempts to suspend failing after crash

[Daniel Pocock]

On 18/06/18 09:51, Michael Biebl wrote:
> Am 18.06.2018 um 09:26 schrieb Daniel Pocock:
> 
>> The problem: a corrupted swap partition
>>
>> Before deciding how systemd should detect that, it may first be
>> necessary to decide if it is the job of systemd or some other process to
>> detect the problem
> 
> How is a swap partition relevant for suspend?
> 


I don't know, but suspend would not work until I manually resolved the
corruption issue

Bug#892321: attempts to suspend failing after crash

[Daniel Pocock]

On 17/03/18 03:07, Michael Biebl wrote:
> On Thu, 8 Mar 2018 10:02:32 +0000 Daniel Pocock  wrote:
>> Package: systemd
>> Version: 232-25+deb9u1
>> Severity: important
>>
>>
>> Laptop was connected to a Thunderbolt dock, lid closed, using desktop
>> monitor and keyboard, everything connected through the dock.  It is a
>> laptop with encrypted LVM and swap on LVM.
>>
>> Laptop's thunderbolt cable was bumped very slightly without coming out
>> completely and then pushed back in.  As the lid was down, it appears the
>> laptop tried to suspend
>>
>> It didn't really suspend though, it became frozen with a blank screen.
>>
>> After a hard reboot, every subsequent attempt to close the lid or undock
>> also failed, the screen would go blank and it wouldn't come back.
>>
>> journalctl and /var/log/messages provided no clues.
>>
>> Guessing it was a problem with the swap partition, tried the following:
>>
>>  swapoff /dev/mapper/vg00--vg-swap_1
>>  mkswap /dev/mapper/vg00--vg-swap_1
>> mkswap: /dev/mapper/vg00--vg-swap_1: warning: wiping old swap signature.
>> Setting up swapspace version 1, size = 23.8 GiB (25513947136 bytes)
>> no label, UUID=
>> swapon /dev/mapper/vg00--vg-swap_1
>>
>> In this case, the swap partition is named in /etc/fstab, not the UUID. 
>> For somebody with the UUID in fstab it would be necessary to update
>> there too and also run update-initramfs
>>
>> This is not hard for an experienced user to resolve but very annoying as
>> you only discover you have a problem after the second attempt to sleep.
>>
>> Can systemd or whatever else is involved in the suspend/sleep mechanism
>> be improved to detect problems like this, give some kind of error to the
>> user and maybe even help them automatically correct it?
> 
> It's unclear to me what exactly your specific problem is/how and how
> systemd should detect that.
> 

The problem: a corrupted swap partition

Before deciding how systemd should detect that, it may first be
necessary to decide if it is the job of systemd or some other process to
detect the problem

Bug#872867: related bug

[Daniel Pocock]

Some interesting comments in here:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=900849

Bug#900849: allowing alternative country lists on a Debian host/installer

[Daniel Pocock]

On 06/06/18 16:31, Ian Jackson wrote:
> Daniel Pocock writes ("Bug#900849: allowing alternative country lists on a 
> Debian host/installer"):
>> I've been thinking about technical solutions to help the country list
>> bug[1] and the following possibility came to mind:
>>
>> - define a virtual package with a name like "country-codes"
> 
> I have read through the "country list bug" (#872867) and my analysis
> of this problem is as follows:
> 
>  * The purpose of the installer's "country list" is to provide helpful
>defaults for a lot of configuration settings that vary from one
>place to another.
> 
>  * To serve this purpose, it is not necessary or desirable for the

I'm not only talking about the installer though, I opened this bug for
both the installer and post-install.  The analysis you provide is very
good for the installer and I would also be very interested in any
thoughts you have about how it should work after install.

There are various organizations and businesses who use the GENC list
instead of the ISO list.  Mozilla recently chose to use GENC, as noted
in their bug tracker.  Companies that deal with cross-border activities,
e.g. airlines, may also be choosing GENC.


>installer's list to contain only "countries" (whatever that might
>mean).  This is because:
>   - Different people have different views about what counts
> as a country (or what the name of a country might be).
> Debian is ill-equipped to resolve these questions.
>   - Sometimes the configuration might need to be different
> for different parts of the same "country" anyway.
>   - Even if we had an agreed list of countries, their
> official names not provide the most helpful choice for users.
> 
>  * In some cases a generally-recognised country will have so many
>regions that it is not sensible to provide a top-level menu entry
>for them.  Eg, the states of the USA.  In that case a submenu
>will be appropriate.   
> 
>  * If iso country codes are used internally, that is an implementation
>detail.  Where a region which ought to appear in the list does not
>have an iso country code, a different technical solution should be
>adopted.
> 
>  * It is not desirable to try to solve this problem by providing some
>kind of alternative country code list for use by custom installers.
>Of course local derivatives with different installer defaults may
>make sense, but Debian upstream installers should work properly for
>everyone.
> 
> I conclude that:
> 
>  * The description of the list should be changed from "countries" to
>"regions and countries".
> 
>  * An entry should be provided in this list (specifically, a region
>should be broken out from any possible parent(s)) when any of the
>following are true:
>   (i) The region wants to have different configuration or
>   different defaults;
>   (ii) Users may not recognise the parent region name(s) as
>   applicable to them
>This also applies when the region is not strictly speaking bounded
>geographically.  For example, it applies when considering different
>different political alignments (or, perhaps cultural, religious or
>or ethnic groups) of people within a single geogrpahic region.
> 
>  * If there would otherwise be a large number of entries for a
>particular region or country, use of a submenu is appropriate.
> 
>  * Names should be primarily chosen so that users know what entry to
>pick in the menu; but it is reasonable to try to avoid giving
>offence.
> 
> Some worked examples:
> 
>  * Kosovo obviously needs to have its own entry.
> 
>  * Taiwan and the PRC need to be different entries.  The Taiwan one
>should probably be labelled "Taiwan (ROC)", and the PRC one "China
>(PRC)" because those will be recognised by more people and give
>less offence.
> 
>  * In principle, there should be an entry for Uyghurs in Xinjiang who
>wish to use the region's unofficial timezone (which is more suited
>to local solar time than official PRC time), and the Uyghur
>language.  But in practice actually selecting such an option in
>one's computer configuration is likely to be very dangerous[1]
>so we might be doing Uyghur users a disservice by offering them
>this option.  This is a difficult case.
> 
>  * I think it will be necessary to have separate options in Bosnia
>for the Federation of Bosnia and Herzegovina on the one hand,
>and for the Republika Srpska on the other.
> 
>  * Belgium needs two entries.  AIUI the keyboards as well as the
>languages are different.
> 
>  * AIUI it would not be necessary to provide any entries for Kashmir;
>the user can select India or Pakistan as appropriate.
> 
> I could be wrong about the details above.
> 
> Ian.
> 
> [1] see eg
>   
> https://www.economist.com/briefing/2018/05/31/china-has-turned-xinjiang-into-a-police-state-like-no-other
> 

Bug#900849: allowing alternative country lists on a Debian host/installer

[Daniel Pocock]

Package: iso-codes
Severity: wishlist
Version: 3.75-1
X-Debbugs-CC: debian-de...@lists.debian.org


I've been thinking about technical solutions to help the country list
bug[1] and the following possibility came to mind:

- define a virtual package with a name like "country-codes"

- src:iso-codes could be split into multiple binary packages, one would
be called iso-country-codes

- package the GENC[2] list as genc-country-codes

- both the iso-country-codes package and the genc-country-codes package
could provide the virtual country-codes so only one can be installed

Would other packages that depend on iso-codes be comfortable using GENC
if it was installed this way?


As an alternative, maybe both lists (ISO-3166 and GENC) could be
installed concurrently and the alternatives[3] mechanism could be used
to switch between them?


For the Debian Installer[4], maybe the installer can ask the user to
choose "Do you want to use ISO-3166 or GENC or load a custom country
list from a USB stick?".


1. https://bugs.debian.org/872867
2. https://nsgreg.nga.mil/doc/view?i=2500
3. https://wiki.debian.org/DebianAlternatives
4. https://wiki.debian.org/DebianInstaller

Bug#898259: RFP: vscode -- Microsoft Visual Studio Code

[Daniel Pocock]

Package: wnpp
Severity: wishlist
X-Debbugs-CC: debian-de...@lists.debian.org,debian-outre...@lists.debian.org


Visual Studio Code regularly comes up in discussions.  Several GSoC
students have asked about using it for Python projects.

Upstream releases[1] the source code under an MIT license on Github

They also distribute binary versions that may or not be built from an
identical code base and contain other (non-free) license clauses.  The
differences are discussed[2] in Github.  It is suggested that the
difference between the source and binary version is like the difference
between chromium and chrome.

Various people have commented about the product sending[3] telemetry
data to the upstream but it is not clear if that happens if it is built
from the source but presumably that can be patched.

The FAQ also mentions[4] auto-updates, that also appears to be something
that would be unexpected for a package user.

There is some background information[5] about the product on Wikipedia.

Any feedback about the suitability of this software as a package or the
use of this software for free software development in general and
alternatives (whether packaged or not) would be welcome.

Regards,

Daniel


1. https://github.com/Microsoft/vscode/blob/master/LICENSE.txt
2. https://github.com/Microsoft/vscode/issues/60#issuecomment-161792005
3.
https://code.visualstudio.com/Docs/supporting/FAQ#_how-to-disable-telemetry-reporting
4.
https://code.visualstudio.com/Docs/supporting/FAQ#_how-do-i-opt-out-of-vs-code-autoupdates
5. https://en.wikipedia.org/wiki/Visual_Studio_Code

Bug#790814: evaluating Kanboard for mentoring

[Daniel Pocock]

On 05/05/18 15:56, Thomas Levine wrote:
> You did make it sound very appealing in comparison with those I tried
> that are available in Sandstorm. Please install it, and I will at least
> try it out.
> 
> It might save me enough time to be worth the effort to administrate it,
> though I would probably check with other hackers to see if anyone else
> wants to administrate it, and even then, I would probably do that only
> if other Debian GSoC projects were going to use it too.
> 



I had a look at the package[1] and it doesn't appear ready to use yet.
I made a quick attempt to make it work:

- created /etc/kanboard

- copied config.default.php to /etc/kanboard/config.php

- symlinked /usr/share/kanboard/config.php -> /etc/kanboard/config.php

- created /etc/kanboard/apache.conf:


Alias /kanboard /usr/share/kanboard/

Allow from all
Options -MultiViews
Require all granted
DirectoryIndex index.php


for inclusion into a VirtualHost

- in the config.php, I set DATA_DIR = /var/lib/kanboard

- created the directory:
  mkdir -p /var/lib/kanboard/data
  chown www-data.www-data /var/lib/kanboard/data

It still doesn't work, this appears in the log:

Fatal error:  require(): Failed opening required
'/usr/share/kanboard/app/../vendor/autoload.php'
(include_path='.:/usr/share/php:/usr/share/pear') in
/usr/share/kanboard/app/common.php on line 3


Most of the above steps should be done by the package itself.  If
anybody has time to help finish the package I'd be happy to install it
on the outreach-lab.debian.net server.  The package doesn't need to be
uploaded into Debian, just as long as the necessary changes are in a
clone of the Git repository so I can build the package myself.
Alternatively, maybe somebody can check with other teams using Kanboard
to see if they have packaged it in another repository or something.

Regards,

Daniel


1. https://salsa.debian.org/debian/kanboard

Bug#896943: bug still appears unfixed in diagram after closing with Version header

[Daniel Pocock]

On 26/04/18 11:20, Adam D. Barratt wrote:
> On 2018-04-26 09:20, Daniel Pocock wrote:
>> I closed this bug by sending an email to the -done address and include a
>> Version pseudo-header:
>>
>> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=724305;msg=70
>>
>> The diagram displayed at the top right corner of the page still suggests
>> the bug is open in versions after the fix:
>>
>> https://bugs.debian.org/cgi-bin/version.cgi?collapse=1;package=postbooks;info=1;fixed=4.5.0;found=postbooks%2F4.0.2-3;absolute=0
>>
> 
> There are no Debian package versions "after" 4.5.0, because there never
> has been such a version. (See below rather than repeating why.)
> 
>> A subsequent message sent to the control server (using a different
>> version number) has fixed the diagram though.  Why didn't the Version
>> pseudo-header work?
> 
> Because there has never been an upload of postbooks to Debian that used
> the version number "4.5.0" in a changelog stanza.
> 
> Versions aren't arbitrary strings in BTS version tracking. If you want
> to use them in fixed / found lists, they need to match actual Debian
> version numbers that appear in the package's debian/changelog as
> uploaded to the archive.
> 


Could the BTS send a response explaining that when somebody tries to use
a Version it won't handle?

The precise upstream version with the fix is an accurate piece of
information, is there a strong reason not to accept that information
from a maintainer and then use the version comparison algorithm to
automatically decide which changelog versions are fixed?

Regards,

Daniel

Bug#896943: bug still appears unfixed in diagram after closing with Version header

[Daniel Pocock]

package: bugs.debian.org


Hi,

I closed this bug by sending an email to the -done address and include a
Version pseudo-header:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=724305;msg=70

The diagram displayed at the top right corner of the page still suggests
the bug is open in versions after the fix:

https://bugs.debian.org/cgi-bin/version.cgi?collapse=1;package=postbooks;info=1;fixed=4.5.0;found=postbooks%2F4.0.2-3;absolute=0


A subsequent message sent to the control server (using a different
version number) has fixed the diagram though.  Why didn't the Version
pseudo-header work?

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=724305;msg=74

Regards,

Daniel

Bug#896760: DDPO: add Salsa icon to VCS column

[Daniel Pocock]

On 24/04/18 13:20, James McCoy wrote:
> On Tue, Apr 24, 2018 at 10:27:47AM +0200, Daniel Pocock wrote:
>> When looking at a long list of packages it is hard to see if any have
>> been missed.  The VCS URLs don't make it obvious because sometimes the
>> VCS has moved but there hasn't been a new upload of the package.
> 
> How would one detect that the VCS has moved if there hasn't been a new
> upload?
> 


It may involve a couple of scripts:

- on alioth, a job that runs periodically and detects whether
.git/description mentions salsa

- on salsa, a job that checks the debian/changelog or debian/control in
each repository to identify the name of the source package

This might not be perfect but it might give some hints

Regards,

Daniel

Bug#896760: DDPO: add Salsa icon to VCS column

[Daniel Pocock]

Package: qa.debian.org
Severity: wishlist

In the VCS column of the DDPO table, it would be useful to have an icon
to indicate whether a package has been moved to salsa.

When looking at a long list of packages it is hard to see if any have
been missed.  The VCS URLs don't make it obvious because sometimes the
VCS has moved but there hasn't been a new upload of the package.

Regards,

Daniel

Bug#895850: duplicate contact issue

[Daniel Pocock]

Package: evolution
Version: 3.22.6-1+deb9u1
Severity: important


I was looking for a particular contact and I typed the name in the
search box.

The contact appeared twice in the search results.

I'm running a standard Debian stretch installation.  The address book
comes from a DAViCal server on Debian stretch.

I looked in the PostgreSQL backend for DAViCal and found the contact
only exists once.  I searched by both name and email address to make
sure.  I also searched on other applications linked to the same DAViCal
(e.g. CardBook for Thunderbird) and they only find the contact once.

I looked at Evolution on another system (laptop) using the same DAViCal
account and it also had the same contact twice.

Looking in the cache file:
~/.cache/evolution/addressbook/foo/cache.xml

there are two entries.

In the GUI, I right clicked each of them and exported them to vcf files.

Here are some differences I observed:

First file Second file


FN = full name FN is the first name only
N = full name  N = full name
UID not in SQL UID matches the value in SQL
latest data: nolatest data: yes,
  includes an extra email address

REV:2015   REV:20180330...
  (when I added extra email addr)


I right clicked the address book and clicked "Refresh", the duplicate
was still there.

I looked at the Evolution preferences window to see if there was a way
to clear the cache, I couldn't find it.

I closed the GUI, killed all the evolution-addressobk* processes,
deleted the cache.xml file and started the GUI again.  The duplicate
contact (left column above) was not there any more.

Why does the cache not realize that one of the contacts doesn't exist on
the server?

I also notice other people have complained about duplicates appearing in
the calendar:

https://ask.fedoraproject.org/en/question/53400/duplicate-calendar-entries-in-evolution/

Bug#892321: attempts to suspend failing after crash

[Daniel Pocock]

Package: systemd
Version: 232-25+deb9u1
Severity: important


Laptop was connected to a Thunderbolt dock, lid closed, using desktop
monitor and keyboard, everything connected through the dock.  It is a
laptop with encrypted LVM and swap on LVM.

Laptop's thunderbolt cable was bumped very slightly without coming out
completely and then pushed back in.  As the lid was down, it appears the
laptop tried to suspend

It didn't really suspend though, it became frozen with a blank screen.

After a hard reboot, every subsequent attempt to close the lid or undock
also failed, the screen would go blank and it wouldn't come back.

journalctl and /var/log/messages provided no clues.

Guessing it was a problem with the swap partition, tried the following:

 swapoff /dev/mapper/vg00--vg-swap_1
 mkswap /dev/mapper/vg00--vg-swap_1
mkswap: /dev/mapper/vg00--vg-swap_1: warning: wiping old swap signature.
Setting up swapspace version 1, size = 23.8 GiB (25513947136 bytes)
no label, UUID=
swapon /dev/mapper/vg00--vg-swap_1

In this case, the swap partition is named in /etc/fstab, not the UUID. 
For somebody with the UUID in fstab it would be necessary to update
there too and also run update-initramfs

This is not hard for an experienced user to resolve but very annoying as
you only discover you have a problem after the second attempt to sleep.

Can systemd or whatever else is involved in the suspend/sleep mechanism
be improved to detect problems like this, give some kind of error to the
user and maybe even help them automatically correct it?

Bug#850898: queried upstream

[Daniel Pocock]

I've sent an email to query the upstream plans for OpenSSL 1.1

Bug#889783: sleep, disconnect Thunderbolt display, open laptop lid, X thinks it is dual-headed

[Daniel Pocock]

package: xserver-xorg-core
version: 2:1.19.2-1+deb9u2
severity: important


The laptop was being used with a Thunderbolt 3 dock.  The lid was down
while it was being used and the display attached to the dock was the
only active display.

The sleep button was pressed on the dock and the laptop went to sleep.

The Thunderbolt 3 cable was removed from the laptop.

The lid was opened.

The unlock screen appeared - but the place to type the password was not
visible.  The mouse cursor could be moved beyond the left edge of the
screen, as if another screen was present.  It appears the password field
was being rendered on the other screen.

Typing the password unlocked the laptop but it continued to think it was
dual-headed.

Closing and opening the lid again it started behaving normally.

Bug#888896: RFP: redmine_agile -- Kanban / Agile board plugin for Redmine

[Daniel Pocock]

Package: wnpp
Severity: wishlist
X-Debbugs-CC: pkg-ruby-extras-maintain...@lists.alioth.debian.org

Hi all,

I tested the redmine_agile plugin on a Debian stretch system.

I put notes about my experience with it here:

https://github.com/dpocock/redmine_agile/blob/master/Debian_Install.txt

and if anybody can give feedback about the process I used that would be
really helpful to anybody who tries to make a package.

Notice that the plugin includes a few hundred lines of non-free
JavaScript and until that is resolved it can't go into main.

The Ruby part of the plugin is GPLv2 code.

It also depends on one library, redmine_crm, which is also GPLv2

Regards,

Daniel

Bug#815234: lintian: Detect Rdata files with embedded code

[Daniel Pocock]

On 30/01/18 09:03, Chris Lamb wrote:
> tags 815234 + moreinfo
> thanks
> 
>> determine if any code is embedded in the file 
> 
> So, how does one do this? :)
> 

Asking the GSoC applicants?  Do we have a tag for bugs that provide an
interesting challenge like this?

Bug#887127: RFP: leansdr -- Lightweight, portable software-defined radio with leandvb for DVB

[Daniel Pocock]

Package: wnpp
Severity: wishlist
X-Debbugs-CC: debian-h...@lists.debian.org,p...@pabr.org


https://github.com/pabr/leansdr


License: GPL-3


I saw this demonstrated by Michel HB9DUG at RHL'18, it is also discussed
in Swiss ATV

http://home.swissatv.ch/2017/01/leandvb-quesako/

Bug#886999: libreoffice crashes

[Daniel Pocock]

On 12/01/18 13:13, Rene Engelhard wrote:
> On Fri, Jan 12, 2018 at 12:51:00PM +0100, Daniel Pocock wrote:
>> Package: libreoffice
>> Version: 1:5.2.7-1
>^
>>
>> On one particular host running jessie, LibreOffice starts very slowly
>  ^^^
>  yeah, right, and this is a bug then why?
> 
>  ^^^
>really? above version is stretch.
> 
>> and after a few minutes, LibreOffice crashes
>>
>> journalctl shows an error like this at each crash:
>>
>> Window manager warning: Invalid WM_TRANSIENT_FOR window 0x3400053
>> specified for 0x340040b
>>
>>
>> It was working fine before, problem only started today.
> 
> And what makes you think this is a LO bug then?


The recovery window appears immediately announcing it can get the
document back

If it is able to save correctly and show that window, it would be useful
if it gave some clue about why the crash happened, e.g. did a malloc
call fail, is there a stack trace?  The window doesn't give the user any
hints to troubleshoot further.

I'm guessing it may be an issue with low physical RAM on a host that
doesn't have dedicated video RAM but the errors don't give me enough
detail to be certain.

Regards,

Daniel

Bug#886999: libreoffice crashes

[Daniel Pocock]

Package: libreoffice
Version: 1:5.2.7-1

On one particular host running jessie, LibreOffice starts very slowly
and after a few minutes, LibreOffice crashes

journalctl shows an error like this at each crash:

Window manager warning: Invalid WM_TRANSIENT_FOR window 0x3400053
specified for 0x340040b


It was working fine before, problem only started today.

Bug#881379: feedback, setup

[Daniel Pocock]

Hi,

I tried it against DAViCal on Debian jessie and it appears to work.

One thing I notice is that it is not integrated with the standard
Address Book window the way SOGo Connector is.  So you can't go to the
Address Book window and click New to add a CardDAV address book.

You have to go to the Tools menu and look for the CardBook option, it
appears below Address Book on my Thunderbird.  Then it has its own
Address Book window and its own wizard for adding (remote) address books.

Regards,

Daniel

Bug#886339: gnome-shell fails to sleep and resume

[Daniel Pocock]

Package: gnome-shell
Version: 3.22.3-3
Severity: important

Sometimes when the laptop lid is closed it doesn't really go to sleep.
Sometimes after opening it the desktop doesn't come back, just a black
screen or flickering screen.

In one instance, after opening it again, it was observed that the X
screen (tty2) was black and flickering every few seconds.  Pressing
CTRL-ALT-F1 showed the X screen (tty1) for a new user login but pressing
CTRL-ALT-F2 again went back to the flickering screen.

I was able to log in on tty3 and run journalctl, it found this:

Error setting property 'PowerSaveMode' on interface
org.gnome.Mutter.DisplayConfig: Timeout was reached (g-io-error-quark, 24)

I searched for that in a search engine and found this:

https://bugzilla.redhat.com/show_bug.cgi?id=1401382

Checking the memory usage of the gnome-shell process I found it was at
400MB.  I tried to kill it and it wouldn't stop.  Then I used kill -9
and it stopped and a few moments later a new gnome-shell process was
started and I was able to use the system again, all the desktop windows
were still present as they were before.

I've saved a copy of the full journalctl output from the day when this
happened.  At the time when I closed the log includes a few errors:


org.gnome.Shell.desktop[1797]: Window manager warning: Failed to set
power save mode for output eDP-1: Permission denied


17:14:06 gnome-shell[1797]: JS ERROR: TypeError: item is undefined


17:14:06
NMConnectionDevice<._activeConnectionChanged@resource:///org/gnome/shell/ui/status/network.js:388

wrapper@resource:///org/gnome/gjs/modules/lang.js:178


17:14:09 systemd-sleep[16824]: Suspending system...

and about 20 seconds later more messages about, like this:

17:14:29 kernel: PM: Preparing system for sleep (mem)


but I think those started appearing after I opened the lid again because
it wasn't really sleeping.

Regards,

Daniel

Bug#885766: X desktop crash when opening a new tab

[Daniel Pocock]

Package: firefox
Version: 57.0.1-1

I pressed Ctrl-T to open a new tab and the whole Xorg session restarted.

Looking at journalctl -n 1 I found memory issues (log below)

Looking at Ganglia (chart attached), I found that the system was using
10GB out of 16GB, so there was 6GB free RAM.  After logging in again, I
only start Firefox and Thunderbird and its using 4GB again, as
demonstrated in the chart.



19:02:45 host1 evolution.desktop[12511]: Memory pressure relief: Total:
res = 12349440/12349440/0, res+swap = 0/0/0
19:03:15 host1 evolution.desktop[12511]: Memory pressure relief: Total:
res = 12349440/12349440/0, res+swap = 0/0/0
19:03:45 host1 evolution.desktop[12511]: Memory pressure relief: Total:
res = 12349440/12349440/0, res+swap = 0/0/0
19:04:15 host1 evolution.desktop[12511]: Memory pressure relief: Total:
res = 12349440/12349440/0, res+swap = 0/0/0
19:04:45 host1 evolution.desktop[12511]: Memory pressure relief: Total:
res = 12349440/12349440/0, res+swap = 0/0/0
19:04:45 host1 systemd[1]: Started Run anacron jobs.
19:04:45 host1 anacron[2100]: Anacron 2.3 started on 2017-12-29
19:04:45 host1 anacron[2100]: Normal exit (0 jobs run)
19:04:45 host1 systemd[1]: anacron.timer: Adding 2min 33.276752s random
time.
19:05:01 host1 CRON[2110]: pam_unix(cron:session): session opened for
user root by (uid=0)
19:05:01 host1 CRON[2111]: (root) CMD (command -v debian-sa1 > /dev/null
&& debian-sa1 1 1)
19:05:01 host1 CRON[2110]: pam_unix(cron:session): session closed for
user root
19:05:15 host1 evolution.desktop[12511]: Memory pressure relief: Total:
res = 12349440/12349440/0, res+swap = 0/0/0
19:05:24 host1 /usr/lib/gdm3/gdm-x-session[3173]: intel_do_flush_locked
failed: Cannot allocate memory
19:05:24 host1 gnome-terminal-[7552]: gnome-terminal-server: Fatal IO
error 11 (Resource temporarily unavailable) on X server :1.
19:05:24 host1 systemd[3155]: gnome-terminal-server.service: Main
process exited, code=exited, status=1/FAILURE
19:05:24 host1 org.a11y.atspi.Registry[3257]: XIO:  fatal IO error 11
(Resource temporarily unavailable) on X server ":1"
19:05:24 host1 org.a11y.atspi.Registry[3257]:   after 66481 requests
(66481 known processed) with 0 events remaining.
19:05:24 host1 libreoffice-writer.desktop[17526]: X IO Error
19:05:24 host1 systemd[3155]: gnome-terminal-server.service: Unit
entered failed state.
19:05:24 host1 polkitd(authority=local)[895]: Unregistered
Authentication Agent for unix-session:2 (system bus name :1.168, object
path /org/freedesktop/PolicyKit1/AuthenticationAgent, locale
en_IE.UTF-8) (disconnected from bus)
19:05:24 host1 kerneloops-applet.desktop[3415]: kerneloops-applet: Fatal
IO error 11 (Resource temporarily unavailable) on X server :1.

Bug#884319: me too, more details

[Daniel Pocock]

Hi all,

I'm seeing the same problem

The system is running stretch, it is an NFS client with OVS networking

I also intermittently assist in maintaining Debian's NFS packages but
this problem doesn't look like it relates to the NFS package.

Here is my /etc/network/interfaces





auto lo
iface lo inet loopback

auto ovsbr0
iface ovsbr0 inet dhcp
ovs_type OVSBridge
ovs_ports eth0

iface ovsbr0 inet6 dhcp
accept_ra 1
ovs_type OVSBridge
ovs_ports eth0

allow-ovsbr0 eth0
iface eth0 inet manual
ovs_bridge ovsbr0
ovs_type OVSPort






Regards,

Daniel

Bug#885573: [Pkg-utopia-maintainers] Bug#885573: setting global search domains from installation time

[Daniel Pocock]

On 28/12/17 21:56, Michael Biebl wrote:
> Am 28.12.2017 um 21:55 schrieb Daniel Pocock:
>>
>>
>> On 28/12/17 21:29, Michael Biebl wrote:
>>> Am 28.12.2017 um 12:00 schrieb Michael Biebl:
>>>> Am 28.12.2017 um 09:24 schrieb Daniel Pocock:
>>>>> Package: network-manager
>>>>> Version: 1.6.2-3
>>>>>
>>>>> If a user specifies a domain during system installation, should it
>>>>> become a global search domain in the global-dns section of
>>>>> NetworkManager.conf or should NetworkManager discover it through some
>>>>> other mechanism and put it into resolv.conf?
>>>>
>>>> I don't think so.
>>>
>>> I want to add that network-manager would be the wrong package anyway as
>>> the configuration that is generated during installation is created by d-i.
>>>
>>
>>
>> I understand the configuration is generated by d-i, but should the
>> network-manager package be looking for that configuration data somehow?
> 
> It certainly is.
> 
> 



In my case, the domain I specified in d-i didn't make it into my
resolv.conf as a search domain.



signature.asc
Description: OpenPGP digital signature

Bug#885573: [Pkg-utopia-maintainers] Bug#885573: setting global search domains from installation time

[Daniel Pocock]

On 28/12/17 21:29, Michael Biebl wrote:
> Am 28.12.2017 um 12:00 schrieb Michael Biebl:
>> Am 28.12.2017 um 09:24 schrieb Daniel Pocock:
>>> Package: network-manager
>>> Version: 1.6.2-3
>>>
>>> If a user specifies a domain during system installation, should it
>>> become a global search domain in the global-dns section of
>>> NetworkManager.conf or should NetworkManager discover it through some
>>> other mechanism and put it into resolv.conf?
>>
>> I don't think so.
> 
> I want to add that network-manager would be the wrong package anyway as
> the configuration that is generated during installation is created by d-i.
> 


I understand the configuration is generated by d-i, but should the
network-manager package be looking for that configuration data somehow?




signature.asc
Description: OpenPGP digital signature

Bug#885575: md: potentially incorrect warnings about alignment

[Daniel Pocock]

Package: linux
Version: 4.13.13-1~bpo9+1

Looking at the output of dmesg, I see warnings like this:


md0: Warning: Device sde3 is misaligned


Checking the same disk with parted tells me it is aligned correctly:


# parted /dev/sde
GNU Parted 3.2
Using /dev/sde
Welcome to GNU Parted! Type 'help' to view a list of commands.
(parted) align-check optimal 3   
3 aligned



Checking the disk with "fdisk -l" shows me:


# fdisk -l /dev/sde

Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes



When the same disk is attached through a USB HDD dock, the "fdisk -l"
output gives a different value for optimal I/O size:


# fdisk -l /dev/sdb

Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 33553920 bytes

=> 33553920 / 512 = 65535 sectors of 512 bytes


When I partitioned the disk, I used MiB alignment (multiple of 2048
sectors of 512 bytes each)

Is the warning from the md kernel module misleading or is parted
align-check wrong?  Can the warning be made more verbose to indicate why
it believes the partition is not aligned correctly?

Bug#885572: related

[Daniel Pocock]

There is also a related Debian bug[1] about whether the domain specified
during Debian installation should be in the global-dns section.


1. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=885573

Bug#885573: related

[Daniel Pocock]

There is also a related bug[1] about the missing documentation/examples
for global-dns



1. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=885572

Bug#885574: problems with USB-C DisplayPort adapters

[Daniel Pocock]

Package: xserver-xorg-core
Version: 1.19.2-1+deb9u2

Hardware configuration: Intel HD Graphics 620 (Kabylake GT2)

Sockets on PC: HDMI and USB-C

Debian stretch was working fine with the HDMI output and the
xserver-xorg-video-intel driver (specified manually in xorg.conf) and
default 4.9 kernel

I wanted to try attaching to the same monitor using DisplayPort.  I
tried two devices:

- a basic USB-C to DisplayPort adapter cable ("Ibra" brand)

- a USB-C hub/dock device with multiple outputs ("TNP" brand, Mini
DisplayPort, HDMI, VGA, USB 3, card readers), lsusb mentions USB
vendor:device = 05e3:0610 (Genesys Logic)

Both of them worked immediately for text mode (BIOS screens and Linux
boot process).

For X, neither of them worked with the xserver-xorg-video-intel
package/driver.  I decided to remove that package and try running
without any xorg.conf to make it use the modesetting driver.

After that change, I found that the "Ibra" adapter would work with X but
the hub would not work.

Then I tried updating to the latest kernel from backports,
4.13.13-1~bpo9+1 and now the USB-C hub/DisplayPort is working with X too.

When it wasn't working, I couldn't find any helpful feedback in the
logs.  The only thing I noticed sometimes is:

(EE) open /dev/fb0: Permission denied

Bug#885572: man page for NetworkManager.conf: insufficient detail about global-dns section

[Daniel Pocock]

Package: network-manager
Version: 1.6.2-3

I reported[1] this upstream, maybe it could also be addressed within
README.Debian or the wiki[2]

Given that NetworkManager is meant to be GUI-oriented, it feels strange
that users have to specify search domains through a config file and even
more strange that it is not easy because of documentation.

The man page about NetworkManager.conf mentions "GLOBAL-DNS SECTION" and
"GLOBAL-DNS-DOMAIN SECTIONS" but there is not a lot of detail and no
examples.

Searching online I found examples like this:

[global-dns]
enable=yes
searches=example.org

[global-dns-domain-*]
servers=8.8.8.8,8.8.4.4

[global-dns-domain-example.org]
servers=10.0.100.1,10.0.200.1



The "enable=yes" doesn't appear to be necessary, it works without that
line for me.

It is also not clear what I should do if I only want to have "searches"
set globally but I want the DNS servers set automatically from DHCP or
other sources.

For example, I tried just setting this:


[global-dns]
searches=example.org


and not adding any global-dns-domain- sections and the searches
parameter was ignored.




1. https://bugzilla.gnome.org/show_bug.cgi?id=791849
2. https://wiki.debian.org/NetworkManager

Bug#885573: setting global search domains from installation time

[Daniel Pocock]

Package: network-manager
Version: 1.6.2-3

If a user specifies a domain during system installation, should it
become a global search domain in the global-dns section of
NetworkManager.conf or should NetworkManager discover it through some
other mechanism and put it into resolv.conf?

There is also a related bug about the missing documentation/examples for
global-dns

Bug#870058: xsane failing since stretch upgrade

[Daniel Pocock]

On 29/07/17 15:32, Jörg Frings-Fürst wrote:
> tags 870058 + moreinfo
> thanks
> 
> 
> Hello Daniel,
> 
> thank you for spending your time helping to make Debian better with
> this bug report.
> 
> Please can you check your system as described in[1]?
> 


hp-doctor complains about some dependencies (below), could any of these
be related to the problem?

I notice that for command line scanimage, changing from zeroconf
(zc=something) to direct IP (with ip=192.168.1.xxx) makes it work but
gimp is still trying to use the zc= URI

I tried to modify the URI in .gimp-2.8/pluginrc and that didn't fix it
either.

Other pages with the same SetPml error in the log:
https://answers.launchpad.net/hplip/+question/153657
https://answers.launchpad.net/hplip/+question/21257


Missing Required Dependencies
-
error: 'cups' package is missing or 'cups' service is not running.
error: 'libcups2' package is missing or 'cups' service is not running.
error: 'python3-pyqt4' package is missing/incompatible
error: 'gtk2-engines-pixbuf' package is missing/incompatible
error: 'libsane-dev' package is missing/incompatible
error: 'libcups2-dev' package is missing or 'cups' service is not running.
error: 'cups-bsd' package is missing or 'cups' service is not running.
error: 'cups-client' package is missing or 'cups' service is not running.
error: 'libcupsimage2' package is missing or 'cups' service is not running.
error: 'libcupsimage2-dev' package is missing or 'cups' service is not
running.
Missing Optional Dependencies
-
error: 'avahi-utils' package is missing/incompatible
error: 'python3-dbus.mainloop.qt' package is missing/incompatible

Bug#883859: debian/control: VCS URLs are wrong

[Daniel Pocock]

Package: thunderbolt-tools
Version: 0.9.2-1

The VCS URLs in debian/control point to the upstream repository:


Vcs-Git: https://github.com/01org/thunderbolt-software-user-space.git
Vcs-Browser: https://github.com/01org/thunderbolt-software-user-space


and there is no Debian packaging branch or debian/* directory in there.

Bug#883857: please backport for stretch-backports

[Daniel Pocock]

Package: thunderbolt-tools
Version: 0.9.2-1
Severity: wishlist

Please provide a backport of this package for stretch-backports or
indicate if you are not going to backport so somebody else can do it.

I've tested it on a system running stretch with the kernel package
4.13.13-1~bpo9+1 and it compiles and appears to work.

Bug#808847: more spam

[Daniel Pocock]

On 06/12/17 02:45, Scott Kitterman wrote:
> 
> 
> On December 5, 2017 10:35:50 AM EST, Daniel Pocock <dan...@pocock.pro> wrote:
>>
>> I received feedback from somebody that my mail server wasn't accepting
>> messages from one of their servers because of this issue.
>>
>> I manually changed my settings to the new value and I notice that
>> additional spam is now getting through.
>>
>> Maybe it is better to be strict and expect non-spammers to use SPF
>> correctly.
> 
> This is fixed in stable, so not sure why you unarchived the bug.
> 
> Personally, I thought the old default was reasonable, but too many people 
> whined.
> 
> What is it you want to have happen?
> 

The old default (in jessie) was stricter and potentially reduces the
amount of spam.

The old default also reduces the number of good messages that are
received, but the sender receives a bounce telling them why their
message couldn't be delivered with a helpful link to help them
understand how to fix their SPF entry in DNS.

If it went back to a stricter default, the people who whined can
manually whitelist senders with bad/missing SPF records or override the
default.

At this stage, I'm not asking to reopen the bug or change the default
again immediately, I think it may be useful to gather more data on the
issue to see if there are other opinions and also compare against other
distributions and public mail services.

Regards,

Daniel

Bug#808847: more spam

[Daniel Pocock]

I received feedback from somebody that my mail server wasn't accepting
messages from one of their servers because of this issue.

I manually changed my settings to the new value and I notice that
additional spam is now getting through.

Maybe it is better to be strict and expect non-spammers to use SPF
correctly.

Bug#864638: add-ons and console output

[Daniel Pocock]

I noticed the problem again today.  Thunderbird had been started on
Saturday and so it has failed relatively quickly this time.

The add-ons:

Lightning
Adblock Plus
Enigmail
LookOut   (for TNEF attachments)
Mail Merge
TBDialOut

Looking at the console I see things like this:



Warning: CalDAV: No response status doing multiget for calendar daniel2
Warning: CalDAV: Get failed: CalDAV: Error: got status undefined
fetching calendar data for daniel2, null
Warning: There has been an error reading data for calendar: daniel2.
However, this error is believed to be minor, so the program will attempt
to continue. Error code: 0x80004005. Description: CalDAV: Error: got
status undefined fetching calendar data for daniel2, null
Warning: There has been an error reading data for calendar: daniel2.
However, this error is believed to be minor, so the program will attempt
to continue. Error code: READ_FAILED. Description:


Is there a way to enable timestamps on the console output?

The CalDAV server is Davical on Debian.  There is nothing in the
server-side error.log file.



Regards,

Daniel

Bug#882128: bulk subscription of a single address to multiple lists

[Daniel Pocock]

On 19/11/17 14:40, Alexander Wirt wrote:
> On Sun, 19 Nov 2017, Daniel Pocock wrote:
> 
>> Package: lists.debian.org
>> Severity: wishlist
>>
>> It would be useful to provide a way for a single person to quickly
>> subscribe to multiple lists.
>>
>> An example of this is for new participants in Outreachy and GSoC.
>>
>> There could be a few different ways that people interact with this:
>>
>> - on the web site, provide check boxes next to the list names, the user
>> can choose multiple check boxes and be subscribed with a single
>> confirmation email
>>
>> - somebody (e.g. a GSoC mentor) can construct a URL that contains the
>> names of multiple lists and send that URL to people through email or by
>> placing the URL in the wiki.  When somebody clicks the URL, they see a
>> list of the mailing lists they are going to join and they are asked to
>> enter the email address they want to subscribe with.
> like: https://www.debian.org/MailingLists/subscribe ?
> 


That appears to do half of what is being suggested

Where is the source for that page?  There is only a generic link to
https://www.debian.org/devel/website/using_cvs

Does it take URL parameters to select some of the lists?

Would the email address be better at the top of the page?

Regards,

Daniel

Bug#882128: bulk subscription of a single address to multiple lists

[Daniel Pocock]

Package: lists.debian.org
Severity: wishlist

It would be useful to provide a way for a single person to quickly
subscribe to multiple lists.

An example of this is for new participants in Outreachy and GSoC.

There could be a few different ways that people interact with this:

- on the web site, provide check boxes next to the list names, the user
can choose multiple check boxes and be subscribed with a single
confirmation email

- somebody (e.g. a GSoC mentor) can construct a URL that contains the
names of multiple lists and send that URL to people through email or by
placing the URL in the wiki.  When somebody clicks the URL, they see a
list of the mailing lists they are going to join and they are asked to
enter the email address they want to subscribe with.

Regards,

Daniel

Bug#881489: playback scaled to 4k fullscreen is choppy

[Daniel Pocock]

On 12/11/17 18:23, Sebastian Ramacher wrote:
> On 2017-11-12 18:06:51, Daniel Pocock wrote:
>>
>> Should the automatic setting use a different default?
> 
> Depends(TM), I guess.
> 

Reported upstream

https://trac.videolan.org/vlc/ticket/19077

Bug#881489: playback scaled to 4k fullscreen is choppy

[Daniel Pocock]

On 12/11/17 17:43, Sebastian Ramacher wrote:
> Control: tags -1 + moreinfo
> 
> On 2017-11-12 11:23:24, Daniel Pocock wrote:
>> Package: vlc
>> Version: 2.2.6-1~deb9u1
>>
>>
>> Trying to play a video file with a resolution 720x526, 25fps, MPEG-1/2
>> (problem also observed with other videos at different resolutions)
>>
>> With the window maximized and the option "Always Fit Window" selected,
>> it appears to zoom to 3:1 and playback appears good.
>>
>> When fullscreen is selected (up to the 4k resolution of the display), I
>> frequently see what appears to be a horizontal tear across the middle of
>> the picture, especially when there is a lot of movement in the video.
>>
>> Using mpv 0.23.0-2+b2 at fullscreen with the same video I don't observe
>> the problem.
>>
>> I don't have libvdpau1 installed due to problems with the frame rate[1]
>>
>> Hardware is Intel i3-7100U CPU @ 2.40GHz and Kaby Lake integrated
>> graphics Intel HD Graphics 620 (Kabylake GT2)
> 
> Please provide the output of vlc -vvv. Did you try selecting different video
> output modules (e.g. the OpenGL one)?
> 

I had a look at the preferences window, video tab.  The "Output" option
was automatic.

Looking at the logs, I notice it was selecting xcb_xv (XVideo)


[7f3a54001268] xcb vout display debug: connected to X11.0 server
[7f3a54001268] xcb vout display debug:  vendor : The X.Org Foundation
[7f3a54001268] xcb vout display debug:  version: 11902000
[7f3a54001268] xcb vout display debug: using screen 0xf5
[7f3a54001268] xcb_xv vout display debug: using XVideo extension v2.2
[7f3a54001268] xcb_xv vout display debug: using adaptor Intel(R)
Textured Video
[7f3a54001268] xcb_xv vout display debug: using port 74
[7f3a54001268] xcb_xv vout display debug: using image format 0x30323449
[7f3a54001268] xcb_xv vout display debug: using X11 visual ID 0x20
(depth: 24)
[7f3a54001268] xcb_xv vout display debug: using X11 window 0x0320
[7f3a54001268] xcb_xv vout display debug: using X11 graphic context
0x0322
[7f3a54001268] core vout display debug: VoutDisplayEvent 'fullscreen' 0
[7f3a54001268] core vout display debug: VoutDisplayEvent 'resize'
3840x2017 window
[7f3a54001268] core vout display debug: using vout display module
"xcb_xv"






I manually changed it to "OpenGL GLX video output (XCB)" and the problem
goes away.



[7f10c8001268] core vout display debug: using vout display module
"xcb_glx"




Looking at the mpv logs as well, it appears to be using OpenGL by default:



[vo/opengl] Initializing OpenGL backend 'wayland'
[vo/opengl/wayland] failed to connect to a wayland server: check if a
wayland compositor is running
[vo/opengl] Initializing OpenGL backend 'x11probe'
[vo/opengl/x11] X11 opening display: :1
[vo/opengl/x11] X11 running at 3840x2160 (":1" => local display)
[vo/opengl/x11] Detected wm supports NetWM.
[vo/opengl/x11] Detected wm supports FULLSCREEN state.
[vo/opengl/x11] Detected wm supports ABOVE state.
[vo/opengl/x11] Detected wm supports BELOW state.
[vo/opengl/x11] Display 0 (DP1): [0, 0, 3840, 2160] @ 30.00 FPS
...
[vf] Video filter chain:
[vf]   [in] 720x576 [64:45] yuv420p bt.601/bt.601-625/bt.1886/limited
CL=mpeg2/4/h264
[vf]   [out] 720x576 [64:45] yuv420p bt.601/bt.601-625/bt.1886/limited
CL=mpeg2/4/h264
[cplayer] VO: [opengl] 720x576 => 1024x576 yuv420p
[cplayer] VO: Description: Extended OpenGL Renderer


Is there any other detail I should look for in the log?

Should the automatic setting use a different default?

Regards,

Daniel

Bug#881489: playback scaled to 4k fullscreen is choppy

[Daniel Pocock]

Package: vlc
Version: 2.2.6-1~deb9u1


Trying to play a video file with a resolution 720x526, 25fps, MPEG-1/2
(problem also observed with other videos at different resolutions)

With the window maximized and the option "Always Fit Window" selected,
it appears to zoom to 3:1 and playback appears good.

When fullscreen is selected (up to the 4k resolution of the display), I
frequently see what appears to be a horizontal tear across the middle of
the picture, especially when there is a lot of movement in the video.

Using mpv 0.23.0-2+b2 at fullscreen with the same video I don't observe
the problem.

I don't have libvdpau1 installed due to problems with the frame rate[1]

Hardware is Intel i3-7100U CPU @ 2.40GHz and Kaby Lake integrated
graphics Intel HD Graphics 620 (Kabylake GT2)

Regards,

Daniel


1. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=856535

Bug#881166: log files, postrotate example: start-stop-daemon or invoke-rc.d?

[Daniel Pocock]

Package: debian-policy
Version: 4.1.1.1

In the section on log files[1], the example uses start-stop-daemon:

postrotate
start-stop-daemon -K -p /var/run/foo.pid -s HUP -x /usr/sbin/foo -q
endscript

while elsewhere in the policy manual it is suggested to use invoke-rc.d

Inspecting my own system, I noticed most logrotate scripts use
invoke-rc.d, not start-stop-daemon

Imagine the following scenario:

- a package distributes an init script that runs the process with a PID file

- later on, the package maintainer starts distributing a systemd unit
file as well, running the process in foreground without a PID file, but
the init script is still included in the package too

In this case, when the package is installed on a systemd system,
start-stop-daemon would never find a PID file and rotation doesn't work.

Some scripts also appear to make a check if invoke-rc.d really exists
and then fall back to "service foo reload":

postrotate
if which invoke-rc.d >/dev/null 2>&1; then
invoke-rc.d foo rotate > /dev/null
else
if [ -f /var/run/foo.pid ]; then
/etc/init.d/foo rotate > /dev/null
fi
fi
endscript


Should the debian-policy example be changed to invoke-rc.d or to
something more elaborate?

Regards,

Daniel


1. https://www.debian.org/doc/debian-policy/#log-files

Bug#877865: IMAP access to bug reports

[Daniel Pocock]

On 06/10/17 17:35, Don Armstrong wrote:
> On Fri, 06 Oct 2017, Daniel Pocock wrote:
>> Package: bugs.debian.org
>> Severity: wishlist
>>
>> Bug reports can currently be accessed in mbox format
>>
>> Has anybody looked at ways to provide read-only IMAP access to the bug
>> reports or another API that mail clients like Thunderbird could use?
> Hrm; I've not, but that sounds interesting. I'd certainly consider
> patches which enabled such a thing. [Though I'd think that having
> thunderbird open a mailbox on local disk would be pretty easy?]
>
> I did have thoughts about making a more-maildir style storage for mail,
> but it was really low on my priority list.
>

People could rsync a Maildir very easily too

Bug#877865: IMAP access to bug reports

[Daniel Pocock]

Package: bugs.debian.org
Severity: wishlist

Bug reports can currently be accessed in mbox format

Has anybody looked at ways to provide read-only IMAP access to the bug
reports or another API that mail clients like Thunderbird could use?

Bug#877698: create debug packages

[Daniel Pocock]

On 04/10/17 16:02, Sebastian Ramacher wrote:
> On 2017-10-04 15:48:57, Daniel Pocock wrote:
>> Package: src:resiprocate
>> Severity: wishlist
>>
>> Create packages with debug symbols
>>
>> https://wiki.debian.org/DebugPackage
> Automatic dbgsym packages for resiprocate are already in the archive:
>
> $ rmadison repro-dbgsym
> repro-dbgsym | 1:1.10.2-1   | unstable-debug | kfreebsd-amd64, 
> kfreebsd-i386
> repro-dbgsym | 1:1.11.0~beta1-3 | stable-debug   | amd64, arm64, armel, 
> armhf, i386, mips, mips64el, mipsel, ppc64el, s390x
> repro-dbgsym | 1:1.11.0~beta1-3 | unstable-debug | powerpc
> repro-dbgsym | 1:1.11.0~beta5-1 | unstable-debug | amd64, arm64, armel, 
> armhf, i386, mips, mips64el, mipsel, ppc64el, s390x

Thanks for that feedback, does that mean there is nothing that needs to
be done manually for this?

Regards,

Daniel

Bug#877698: create debug packages

[Daniel Pocock]

Package: src:resiprocate
Severity: wishlist

Create packages with debug symbols

https://wiki.debian.org/DebugPackage

Bug#877200: help users identify unknown keyboard layouts

[Daniel Pocock]

Package: debian-installer
Severity: wishlist

The installer currently asks the user to select their keyboard layout
from a list.

Sometimes the user or person doing the installation doesn't know the
name/country for their keyboard layout.

It would be useful to offer some type of keyboard identification wizard
that takes them through a series of questions to identify their keyboard
and then asks them to press some keys (like "z" and "/") to see if they
are correctly recognized.

Bug#245465: related bugs / discussions

[Daniel Pocock]

https://lists.debian.org/debian-devel/2012/05/msg01092.html
https://lists.debian.org/debian-devel/2012/06/msg00311.html (summary of
the above)

https://wiki.debian.org/SSDOptimization#Reduction_of_SSD_write_frequency_via_RAMDISK

and various bugs about systemd doing tmp on tmpfs by default:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=718906
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=807769

and systemd ignoring RAMTMP:
https://github.com/debian-pi/raspbian-ua-netinst/issues/210

Bug#875393: partman-auto: allow user to easily edit mount points and sizes

[Daniel Pocock]

Package: partman-auto
Severity: wishlist

When people select the guided partitioning, the list of possible
partitions is very restricted (all in one, root+home or root+home+var+tmp)

It would be useful to have an extra option, "custom", where the user can
see a list:

/ (root)   100%
/home  0%
/var   0%
/tmp   0%

and edit the percentages or replace them with sizes in GB

If they put 0% or 0, that partition/LV would not be created.

For tmp, it would also be useful to have some way to indicate that tmpfs
/ ramdisk should be used (see also bug #245465)

This screen would also show the amount of space free if they don't use
it all.

This would aim to help people with a single disk (or single RAID volume)
and wouldn't need to try and cover use cases where they want to assign
different mounts to different physical volumes or choose individual
filesystem types.

Bug#875394: new install, gdm3 selected but alternatives/x-session-manager is Xfce

[Daniel Pocock]

Package: gdm3
Version: 3.22.3-3

During a new install of stretch (9.1.0), both GNOME and Xfce were selected

The popup appeared asking to choose a default display manager (gdm3 or
lightdm).  gdm3 was selected.

After installation, gdm3 appears, user logs in and sees the Xfce desktop

ls -l /etc/alternatives/x-session-manager

shows that it is using /usr/bin/startxfce4

Can the choice of display manager and session manager be combined, so if
the user chooses gdm3 they also get /usr/bin/gnome-session?


I changed this using:

$ sudo update-alternatives --config x-session-manager

Bug#875389: sound volume spontaneously decreases

[Daniel Pocock]

Package: gnome-settings-daemon
Version: 3.22.2-2+deb9u2
Severity: important

I've noticed that sometimes the sound volume starts decreasing
spontaneously.  Sometimes it happens when music is playing, sometimes it
happens when no audio playback is in progress.

The square-shaped popup showing the current sound level just appears
over the desktop for a second and the bar moves a little bit to the left
(decrease) and then it goes away.  It keeps happening until the level
drops all the way to 0 / muted.  It even pops up sometimes when it is
already completely muted.

It only happens rarely, not every day, sometimes it happens
intermittently for a period of 5 - 10 minutes and then doesn't happen again.

I've only noticed this problem on one system.

The system has a wireless remote control.  I removed the USB dongle
(wireless receiver) to eliminate the possibility that is related to the
problem.

There is also a USB headset (with microphone) with volume buttons that
can activate the GNOME volume popup.  I removed that too.

They keyboard is a standard USB keyboard with no volume buttons.

The system is an Intel NUC and the spec mentions an infrared sensor on
the front.  There is no active infrared transmitter in the room that I
am aware of.

The monitor has sound and has a different on-screen-display for audio
volume, the volume controls for the monitor do not activate the sound
controls in GNOME.

Has anybody else observed bugs like this?

Is there any way to determine which device is activating the volume
settings?

Is there any debugging that can be enabled to reveal what is doing this?

Could the popup be improved to show which input device triggered the
volume change?

Regards,

Daniel

Bug#875274: UDD/dmd: show submitted bugs with moreinfo tag

[Daniel Pocock]

Package: qa.debian.org
User: qa.debian@packages.debian.org
UserTags: udd
Severity: wishlist

It would be really useful if the HTML page ("To do" section) and
iCalendar feed could show a developer the bugs they submitted against
other packages that have a "moreinfo" tag.

Bug#874503: [Pkg-swan-devel] Bug#874503: warn about unsupported kernel versions

[Daniel Pocock]

On 08/09/17 15:36, Yves-Alexis Perez wrote:
> On Wed, 2017-09-06 at 15:57 +0100, Daniel Pocock wrote:
>> There is some issue with kernel version 4.11
>>
>> strongswan wasn't working, I looked with tcpdump and saw messages like
>> "truncated-udplength 0"
>>
>> I used a search engine to find out about that string and it took me to
>> this page:
>>
>> https://www.linuxquestions.org/questions/linux-networking-3/strongswan-ikev2
>> -clients-can-access-everything-but-strongswan-machine-itself-truncated-udp-
>> length-0-a-4175606751/
>>
>> with the quote "So this appears to be a kernel-related issue; I had
>> updated to 4.11.2 recently. Downgrading back to 4.10.9, without any
>> other changes to my setup, resolves the problem."
>>
>> I downgraded my own system to a 4.9 kernel and it worked again.
>>
>> There was no obvious error from StrongSWAN in the console or in the logs
>> to tell me about this, I can imagine some people losing a lot of time on
>> something like this if they don't come across this page.
> Well, that's a bug in the Linux kernel then, which should be fixed there. I
> don't really see the point in adding documentation for kernel bugs in
> strongSwan, although I sympathize about the lack of information.
>
> When you have identified the relevant kernel versions, commit IDs fixing the
> bug etc., feel free to add them to that bug.


If the kernel gives any indication of this condition to the StrongSWAN
process in userspace, then it would be very desirable for the process to
report that, even if it is just one line in the log with the severity of
warning or error.

Bug#863257: me too, more upstream links

[Daniel Pocock]

Hi,

I've seen this problem on a fresh install of stretch on a Thinkpad T470s
when docked.

I found another upstream bug:

https://bugzilla.gnome.org/show_bug.cgi?id=709062

and a Fedora bug that was incorrectly closed:

https://bugzilla.redhat.com/show_bug.cgi?id=876134

If I dock the laptop, press the wake-up button on the dock, open the lid
and close it again the login screen migrates over to the external monitor.

Regards,

Daniel

Bug#677944: bas-mailx priority has changed

[Daniel Pocock]

I've reopened this bug because of two things:

a) the priority of bsd-mailx was lowered[1] to optional in 2016 (Robert,
can you comment on that?)

b) consequently, mailutils is now being installed by default in stretch
and it doesn't respect /etc/mailname - #253513[2]

Regards,

Daniel


1.
https://anonscm.debian.org/cgit/users/robert/bsd-mailx.git/commit/?id=94629511972c3da0954522828582f581b80615ba
2. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=253513

Bug#874503: warn about unsupported kernel versions

[Daniel Pocock]

Package: strongswan
Version: 5.5.1-4+deb9u1


There is some issue with kernel version 4.11

strongswan wasn't working, I looked with tcpdump and saw messages like
"truncated-udplength 0"

I used a search engine to find out about that string and it took me to
this page:

https://www.linuxquestions.org/questions/linux-networking-3/strongswan-ikev2-clients-can-access-everything-but-strongswan-machine-itself-truncated-udp-length-0-a-4175606751/

with the quote "So this appears to be a kernel-related issue; I had
updated to 4.11.2 recently. Downgrading back to 4.10.9, without any
other changes to my setup, resolves the problem."

I downgraded my own system to a 4.9 kernel and it worked again.

There was no obvious error from StrongSWAN in the console or in the logs
to tell me about this, I can imagine some people losing a lot of time on
something like this if they don't come across this page.

Regards,

Daniel

Bug#874461: avoid enabling kexec reboots by default

[Daniel Pocock]

Package: kdump-tools
Version: 1:1.6.1-1


I installed kdump-tools on a few systems

One of the dependencies is kexec-tools.

When the kexec-tools is installed, it appears to be enabled by default
for kexec-style reboots.  Somebody who only has kexec-tools on their
system to satisfy kdump-tools may not want kexec reboots.  Could this
scenario be avoided without having to prompt the user during the
kexec-tools postinst?

Regards,

Daniel

Bug#253513: this is a bug

[Daniel Pocock]

According to policy

/usr/share/doc/debian-policy/policy.txt.gz

"Non-conformance with
 guidelines denoted by _should_ (or _recommended_) will generally be
 considered a bug, but will not necessarily render a package unsuitable
 for distribution."

and further down:

"If your package needs to know what hostname to use on (for example)
 outgoing news and mail messages which are generated locally, you
 should use the file `/etc/mailname'.  It will contain the portion
 after the username and `@' (at) sign for email addresses of users on
 the machine (followed by a newline)."

Note that policy uses the word "should" for mailname, so wishlist doesn't 
appear to be a suitable severity for this issue.

This bug is for the sender address, the merged bug #509810 is for the recipient 
address, so it is actually a different issue and it remains a wishlist item, 
I've unmerged them.

I've observed that on a freshly installed stretch system, the mailutils 
implementation of the mail command is now the default:
  
   /etc/alternatives/mail -> /usr/bin/mail.mailutils

while on older systems, it was bsd-mailx:

   /etc/alternatives/mail -> /usr/bin/bsd-mailx

and so it impacts every new install.  Therefore I've chosen the severity 
"serious" for this issue.

People can work around this issue with:

   $ sudo apt install bsd-mailx

As a consequence of this issue, when people send outgoing mail from a stretch 
system using the /usr/bin/mail command, the sender domain may not be what they 
expect and their mail may not even be accepted by their relay host.  If all 
their older Debian systems are running bsd-mailx this is an issue that will 
take them by surprise on newly built systems.

Even if upstream doesn't want to support /etc/mailname because they consider it 
a Debian thing, as mentioned in the earlier message in this bug, the Debian 
version should probably contain a patch or it should not be the default 
/usr/bin/mail.

Bug#874441: installing the pbuilder package, complains if sources.list using apt-transport-tor

[Daniel Pocock]

Package: pbuilder
Version: 0.228.7

To reproduce the problem:

1. install apt-transport-tor

  apt update && apt install apt-transport-tor

2. configure sources.list to use apt-transport-tor

cat > /etc/apt/sources.list << EOF
deb  tor+http://vwakviie2ienjx6t.onion/debian 
stretchmain contrib non-free
deb-src  tor+http://vwakviie2ienjx6t.onion/debian 
stretchmain contrib non-free

deb  tor+http://sgvtcaew4bxjd7ln.onion/debian-security
stretch/updatesmain contrib non-free
deb-src  tor+http://sgvtcaew4bxjd7ln.onion/debian-security
stretch/updatesmain contrib non-free

# stretch-updates, previously known as 'volatile'
deb  tor+http://vwakviie2ienjx6t.onion/debian 
stretch-updatesmain contrib non-free
deb-src   tor+http://vwakviie2ienjx6t.onion/debian 
stretch-updatesmain contrib non-free

# stretch-backports, previously on backports.debian.org
deb  tor+http://vwakviie2ienjx6t.onion/debian 
stretch-backportsmain contrib non-free
deb-src   tor+http://vwakviie2ienjx6t.onion/debian 
stretch-backportsmain contrib non-free
EOF


3. try to install pbuilder

  apt update && apt install pbuilder


4. a dialog appears complaining that no valid sources were found, the
user has to type another source URL such as the example URL
http://httpredir.debian.org/debian

Bug#873960: virtualized desktop becomes unresponsive to mouse clicks

[Daniel Pocock]

Package: libspice-server1
Version: 0.12.8-2.1+deb9u1
Severity: important

I'm running Debian stretch on client, KVM server and virtualized desktop.

I've tried virt-manager and spicy on the client side using the current
packages in stretch:


gir1.2-spice-client-glib-2.0 0.33-3.3
gir1.2-spice-client-gtk-3.0 0.33-3.3
libspice-client-glib-2.0-8:amd64 0.33-3.3
libspice-client-gtk-3.0-4:amd64 0.25-1+b1
libspice-client-gtk-3.0-5:amd64 0.33-3.3
libspice-protocol-dev 0.12.12-1
libspice-server1:amd64 0.12.8-2.1+deb9u1
spice-client-glib-usb-acl-helper 0.33-3.3
spice-client-gtk 0.33-3.3
spice-vdagent 0.17.0-1
virt-manager 1:1.4.0-5



Initially, it was all working but with a little latency after mouse
clicks.  Now, when I connect to the virtualized desktop, keyboard input
works but mouse clicks don't have any impact.

The virtualized desktop is GNOME 3.

I've tried connecting with virt-manager (sets up ssh tunnel for me) and
also setting up the ssh tunnel manually and connecting to the tunnelled
port with spicy.

In spicy, I tried various settings like "Toggle mouse mode" (mouse
cursor vanishes when in server mode), "Disable inputs" and "sync
modifiers" and none of them resolve the issue.



/var/log/libvirt/qemu/host.log doesn't have any clues, these are the
debug messages that appear when spicy connects:

main_channel_link: add main channel client
main_channel_handle_parsed: agent start
main_channel_handle_parsed: net test: latency 1.544000 ms, bitrate
15082001 bps (14.383317 Mbps)
red_dispatcher_set_cursor_peer:
inputs_connect: inputs channel client create


On the virtualized system, I ran

$ sudo systemctl restart gdm

and then logged in again and the mouse is working again.

Looking at /var/log/Xorg.0.log.old I found some messages like this:


[   201.369] Build Date: 11 February 2015  12:32:02AM
[   201.369] xorg-server 2:1.16.4-1 (http://www.debian.org/support)
...

[1576555.727] (II) config/udev: removing device spice vdagent tablet
[1576555.757] (II) evdev: spice vdagent tablet: Close
[1576555.757] (II) UnloadModule: "evdev"
[1576555.826] (II) config/udev: Adding input device spice vdagent tablet
(/dev/input/mouse2)
[1576555.826] (II) No input driver specified, ignoring this device.
[1576555.826] (II) This device may have been added with another device file.
[1576555.827] (II) config/udev: Adding input device spice vdagent tablet
(/dev/input/js1)
[1576555.827] (II) No input driver specified, ignoring this device.
[1576555.827] (II) This device may have been added with another device file.
[1576555.896] (II) config/udev: Adding input device spice vdagent tablet
(/dev/input/event5)
[1576555.896] (**) spice vdagent tablet: Applying InputClass "evdev
pointer catchall"
[1576555.896] (II) Using input driver 'evdev' for 'spice vdagent tablet'
[1576555.896] (**) spice vdagent tablet: always reports core events
[1576555.896] (**) evdev: spice vdagent tablet: Device: "/dev/input/event5"
[1576555.896] (--) evdev: spice vdagent tablet: Vendor 0 Product 0
[1576555.896] (--) evdev: spice vdagent tablet: Found 3 mouse buttons
[1576555.896] (--) evdev: spice vdagent tablet: Found scroll wheel(s)
[1576555.896] (--) evdev: spice vdagent tablet: Found relative axes
[1576555.896] (--) evdev: spice vdagent tablet: Found absolute axes
[1576555.896] (--) evdev: spice vdagent tablet: Found x and y absolute axes
[1576555.896] (--) evdev: spice vdagent tablet: Found absolute touchscreen
[1576555.896] (II) evdev: spice vdagent tablet: Configuring as touchscreen
[1576555.896] (II) evdev: spice vdagent tablet: Adding scrollwheel support
[1576555.896] (**) evdev: spice vdagent tablet: YAxisMapping: buttons 4
and 5
[1576555.896] (**) evdev: spice vdagent tablet: EmulateWheelButton: 4,
EmulateWheelInertia: 10, EmulateWheelTimeout: 200
[1576555.896] (**) Option "config_info"
"udev:/sys/devices/virtual/input/input20/event5"
[1576555.896] (II) XINPUT: Adding extended input device "spice vdagent
tablet" (type: TOUCHSCREEN, id 10)
[1576555.896] (WW) evdev: spice vdagent tablet: touchpads, tablets and
touchscreens ignore relative axes.
[1576555.897] (II) evdev: spice vdagent tablet: initialized for absolute
axes.
[1576555.897] (**) spice vdagent tablet: (accel) keeping acceleration
scheme 1
[1576555.897] (**) spice vdagent tablet: (accel) acceleration profile 0
[1576555.897] (**) spice vdagent tablet: (accel) acceleration factor: 2.000
[1576555.897] (**) spice vdagent tablet: (accel) acceleration threshold: 4

Bug#873843: fails with error "found multiple ISRC commands"

[Daniel Pocock]

Package: flac
Version: 1.3.2-1



$ flac ... --cuesheet=disc.cue data.bin

fails with this error:

data.bin: ERROR parsing cuesheet "disc.cue" on line 17:
  found multiple ISRC commands




The CUESHEET was obtained by this series of commands:

cdrdao read-cd --device $CDDEV --driver generic-mmc-raw \
  --read-raw ${TOC_FILE}

cueconvert $TOC_FILE} disc.cue

The TOC contains two ISRC entries for each track:

TRACK AUDIO
NO COPY
NO PRE_EMPHASIS
TWO_CHANNEL_AUDIO
ISRC "GBCLS1234567"
CD_TEXT {
  LANGUAGE 0 {
TITLE "foo"
PERFORMER "foo"
SONGWRITER ""
COMPOSER "foo"
ARRANGER "foo"
DISC_ID ""
ISRC "GB-CLS-12-34567"
  }
}
FILE "data.bin" 0 03:21:23



and the CUESHEET contains entries like this for each track:

ISRC "GB-CLS-12-34567"
ISRC GBCLS1234567


May be related to bug #815587

Bug#872867: is ISO-3166 really the optimal list for our users?

[Daniel Pocock]

On 24/08/17 18:02, Wouter Verhelst wrote:

> Case in point: why would Kosovo *need* to be on the d-i list of
> countries? Is there a difference between Kosovo and Serbia[1] or other
> countries in the former Yugoslavia that d-i would need to be aware of
> (such as locale settings, keyboard layout, time zone, or preferred
> mirror), or is this bug in fact just the result of people with a
> background of Kosovan nationalist ideology complaining about it to
> you[2]? So far I've only seen you argue that "Kosovo needs to be on the


Another consideration is the uncertainty:  for many ordinary users (not
necessarily people with strong political opinions) it is simply not
clear what the technical consequences will be if they select RS or AL
from the list.  Simply abolishing the list and asking people to manually
select a timezone may be one way to eliminate this uncertainty.

About 50% of the population is under 30[1] - many of them presumably
have no memory of being a Serbian or Yugoslavian, they have simply been
brought up as Kosovans.  This is also the age group where people are
likely to try free software.  I haven't observed any Kosovan taking up
arms when they see the country list, rather, it is more sadness,
disappointment and frustration.

Regards,

Daniel


1. http://www.indexmundi.com/kosovo/age_structure.html

Bug#872867: is ISO-3166 really the optimal list for our users?

[Daniel Pocock]

On 24/08/17 18:02, Wouter Verhelst wrote:

> Case in point: why would Kosovo *need* to be on the d-i list of
> countries? Is there a difference between Kosovo and Serbia[1] or other
> countries in the former Yugoslavia that d-i would need to be aware of
> (such as locale settings, keyboard layout, time zone, or preferred
> mirror), or is this bug in fact just the result of people with a

90% of citizens in Kosovo are ethnic Albanian[1] and would probably
prefer the installer to help them select the Albanian language (coming
soon thanks to a translation sprint at FOSSCamp.cc) and mirrors in
Kosovo.  These are probably the two main differences from RS.  Kosovo
has EUR as currency, RS does not.

The locale chooser and the country list for mirrors could potentially
work in different ways too, e.g. maybe subdivisions could be introduced
for the mirror list.

It doesn't impact time zone

I don't have hard data about the keyboard layouts users prefer in the
region, but I've seen US style keyboards there.

Regards,

Daniel



1. https://en.wikipedia.org/wiki/Kosovo_Albanians

Bug#872867: is ISO-3166 really the optimal list for our users?

[Daniel Pocock]

On 24/08/17 08:51, Wouter Verhelst wrote:

> In the case of XYZ, "use a different distribution" isn't going to
> silence such people. Instead, they'll just yell harder. "Debian's making
> a political statement about XYZ, and it's wrong, and I told them that
> it's wrong, and they're ignoring me!"
>

Including and excluding are different things

Including extra "countries" may cause offense to people with certain
political sensitivities, but doesn't cause any technical problem for
people in other countries.

So if Debian has a policy that we favour inclusion over exclusion and
that any country can be listed if at least one DD visits there and
confirms it exists, is that political or would that be a policy that can
be defended?

I wonder how long it would be before somebody proposed California or
Scotland?  If they entered the list at the same time as Kosovo (and Hutt
River) then it becomes a lot less political and the focus is not on a
single region.

> Whatever solution you[1] come up with should avoid that.
> 
> [1] I'm assuming you're planning on submitting a patch, since you
> suggested this change in the first place...
> 

I would not want to waste time coding something before we have some
consensus about what the community is comfortable with.

Bug#872867: is ISO-3166 really the optimal list for our users?

[Daniel Pocock]

On 23/08/17 19:22, Wouter Verhelst wrote:
> On Tue, Aug 22, 2017 at 11:02:27AM +0200, Daniel Pocock wrote:
>> While it is good that we use material from official sources, Debian is
>> independent of any state and may not need to feel constrained by such lists/
>> standards in the same way that a commercial software vendor might be.
> 
> On the other hand, the advantage of having an official standard to point
> to is that we can deflect complaints when they appear.
> 
> There are many areas in dispute in this world, and in some cases
> deciding whether a particular area is or is not a country will result in
> offending one or the other party. When Debian accidentally and

We should be respectful of all users, but I don't think fear of causing
offence should be the primary concern.  Otherwise we never would have
moved to systemd.

I feel the concern should be providing a technically valid installation
process for as many users as possible.


> temporarily updated the representation of the countries in the installer
> so that they would refer to the area sometimes referred to as the
> "Republic of China", otherwise known as "Taiwan", in a particular way,
> this offended one of our developers enough that he decided to leave the
> project.
> 

If a developer puts his political opinions against the needs of users in
 a particular region, then is it possible that developer is failing to
respect the Debian Social Contract and may not be eligible to be a
developer?

In such cases, it is sometimes possible to identify other developers or
users who adopted a project because they favour the decision too.

> While it may be true that the list of countries in ISO-3166 is decided
> upon by a small number of people, presumably these people are aware of
> all the peculiarities and policital sensitivies in deciding what is or
> isn't a country, and what should or should not be allowed in a list of
> countries. As such, deciding to use ISO-3166 as our base to decide on
> which countries to list keeps Debian politically neutral in an area for
> which we really have no expertise and in which we really should not get
> ourselves involved one way or the other.
> 

I agree Debian should not get into the political side of this debate.
This bug is only for the technical side...

> It may make sense to change the list of places so that it also includes
> subdivisions, provided we do so in either a way which makes the
> distinction between "country" and "subdivision" clear, or a way which
> puts both in a singular list but which makes it clear that the list does
> not refer to countries in any form or sort. Both solutions would allow
> areas such as Kosovo to appear on the list of places, without offending
> people who believe Kosovo should not be considered an independent nation
> (yes, there are such people).
> 

If that means a user in Kosovo is more likely to configure their system
correctly, then it is a good technical solution, similar to what I
described (leaving out the country codes for such regions and helping
them choose alternatives).

We could also have a disclaimer, "Not all entries in this list are
officially recognized as countries, some are disputed territories that
have been included for the purpose of helping users in those regions get
the optimal configuration."

More concise: "Select the entry from this list of regions and countries
that most closely matches your geographic location"


> We should not, however, move away from ISO-3166 as our basis for
> deciding what is or isn't a country, unless you can point to another
> list which has the same level of international recognition as ISO-3166.
> I don't think such a list exists, however.
> 
> If we continue to use the ISO-3611 list, then if any error in the list
> of countries in our installer exists it will either be a bug in the
> installer code (which we could obviously fix and apologise for,
> hopefully without offending anyone), or a matter of "the list is
> outdated" (which we would usually fix by rebuilding the installer,
> hopefully also without offending anyone), or an error in the ISO-3611
> list (in which case people might be offended, but the offending elements
> would not be ours and we can tell them to get ISO to update the list
> rather than to complain to us).
> 
> Debian is about Free Software; it is not about International Politics.
> Let's keep it that way.

I agree - but producing Free Software means we should feel free to
innovate in this area as long as we are open and honest with our users
about what we are doing and why.

As systemd demonstrated, Debian can't please everybody all the time.

Regards,

Daniel

Bug#872867: another option: GENC

[Daniel Pocock]

Mozilla decided[1] to use GENC[2], a list produced by the US Government
and including additional countries that are not correctly covered by
ISO3166.

The US Govertment's approach (making their own list based on ISO3166) is
similar in some ways to the approach I have suggested for Debian.


1. https://bugzilla.mozilla.org/show_bug.cgi?id=733417#c20
2. https://nsgreg.nga.mil/doc/view?i=2500

Bug#872867: is ISO-3166 really the optimal list for our users?

[Daniel Pocock]

On 22/08/17 10:52, Bastian Blank wrote:
> On Tue, Aug 22, 2017 at 12:28:18AM +0200, Daniel Pocock wrote:
>> According to this section in the Debian Installer i18n guide[1], the
>> list of countries is based on ISO3166.
> Kosovo is listed in ISO3166 as RS-KM, see
> https://www.iso.org/obp/ui/#iso:code:3166:RS 
>
> However we don't list subdivisions.

I'm not sure adding support for subdivisions would satisfy Debian users
in some of these situations but it could be one part of the solution.

>> According to wikipedia, ISO3166 comes from a group of 10
>> representatives[2] from mostly rich countries.
> Most of them from countries actually reognizing Kosovo as independent,
> so what?

While it is good that we use material from official sources, Debian is
independent of any state and may not need to feel constrained by such
lists/standards in the same way that a commercial software vendor might be.


>> It may be interesting to support micronations like the Principality of
>> Hutt River[5] too.
> This one is not recognized by anyone.
>
> Please let the peolpe of Kosovo speak for themselves.

Well, when somebody in Kosovo set up a new mirror, they listed[1] it
under AL (Albania) rather than RS (Serbia).

Regards,

Daniel

1. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=867255

Bug#872867: is ISO-3166 really the optimal list for our users?

[Daniel Pocock]

Package: localechooser
Version: 2.69
Severity: wishlist

While visiting Kosovo, I have met several people interested in becoming
Debian users.

The first time I helped one of them run the Debian installation, I was
baffled to find that their country was not in the list offered by the
installer.

It is potentially not a good situation for a developer to be in a
country, surround by 1.8 million people who consider themselves to be
citizens of a country that our installer doesn't know about. 
Fortunately Kosovans are really nice people and weren't upset about this
but I can image they must feel quite sad each time something like this
happens.

According to this section in the Debian Installer i18n guide[1], the
list of countries is based on ISO3166.

According to wikipedia, ISO3166 comes from a group of 10
representatives[2] from mostly rich countries.

The Debian Social Contract[3], point 4, requires us to put our users
first.  Can Debian do more to listen to the 1.8 million potential users
in Kosovo and other countries like it?  Are we actually obliged to
respect how our users see themselves over and above the way some
bureaucrats in Geneva see them?

I would propose that we regard ISO3166 as a subset of the list of
countries perceived by our users and that for the next installer
release, the locale chooser offers a list of countries with a disclaimer
that some of them are not in ISO3166 but they are included at the
request of our users.

Maybe the list can show two columns, country and country code.  For
countries where ISO3166 is not competent, the country code column could
be blank.

Implementing this might be tricky but not impossible.  For example, a
crude implementation may simply display the extra countries in the main
list, but if somebody selects one of them, the installer shows a message
apologizing for the fact they are not fully supported and offering to
help them choose from a subset of related country codes.  Maybe their
preferred choice (verbose country name) could be saved somewhere for
later use when their country is fully supported and a future version of
localechooser will help them adapt to their eventual country code during
a future dist-upgrade.

For Kosovo in particular, Wikipedia notes[4] that "The code XK is being
used by the European Commission
,^[21]
 the IMF
, and SWIFT
,^[22]
 CLDR and
other organizations as a temporary country code for Kosovo
.".  Other 2-letter codes are
mentioned elsewhere so the ideal solution may avoid using a 2 letter
code for such countries or maybe it can borrow one of the codes used by
other international organizations who didn't wait for ISO3166.

It may be interesting to support micronations like the Principality of
Hutt River[5] too.

Regards,

Daniel


1. https://d-i.alioth.debian.org/doc/i18n/ch01s05.html#idm45330184240096
2. https://en.wikipedia.org/wiki/ISO_3166#Members
3. https://www.debian.org/social_contract
4. https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2
5. https://en.wikipedia.org/wiki/Principality_of_Hutt_River

Bug#865633: [Pkg-libvirt-maintainers] Bug#865633: can't drag windows or click "Activities" in GNOME 3 guest

[Daniel Pocock]

On 13/08/17 10:17, Petter Reinholdtsen wrote:
> This RC issue has caused virt-viewer to be thrown out of testing.  Anyone
> got any idea how to fix it?
> 

I did a full reboot of the physical host and the guest VM and after that
I was able to connect successfully from virt-viewer on stretch.

This is not ideal though.  Has anybody else seen the issue?

Bug#871903: difficulty changing WPA2 enterprise credentials

[Daniel Pocock]

Package: network-manager-gnome
Version: 1.4.4-1
Severity: important

I was using a network with WPA2 enterprise (PEAP) credentials for a few
days.

I was given a new username and password today.

When the old credentials stopped working, there was no error, it would
just give up and try connecting to another SSID.  At this point, it
should show an error popup advising the credentials were not accepted
(any more)

Clicking on the connection to edit it was unpredictable, the edit window
would not always appear

After putting in the new credentials, it still wouldn't connect.  I
looked in the settings again and found the old credentials had come back.

I typed the new credentials again and tried turning the hardware switch
off and on

It still wouldn't connect.  I checked the settings and this time it
appeared to have remembered the new settings.

Then I tried the "Forget" button on the "Reset" tab and then entered the
new credentials again.  It still couldn't connect.

After that, I tried putting the laptop to sleep and waking it up again. 
After this, I was finally able to connect.

Bug#871415: debian-dug-kosovo mailing list creation

[Daniel Pocock]

Package: lists.debian.org
Severity: wishlist

This is a request for a new mailing list

Name: debian-dug-kosova

Rationale: local users and visitors to Kosovo

Short description: Discussion list for the Kosovo Debian community

Long description:
"Discussing user queries, events and other local topics for Debian in Kosovo"

Language used on this list: English and Shqip

Category: misc / user / dug

Subscription Policy: open

Post Policy: open to list subscribers

Web Archive: yes

Bug#870505: can't sign when subkey expired, insufficient error message

[Daniel Pocock]

Package: enigmail
Version: 1.9.8.1-1~deb9u1

I wanted to send a signed message but Enigmail wouldn't let me.  A popup
was appearing:

"encryption command failed"

After investigating at the command line, I found my subkey had expired. 
This is fine for a knowledgable user but not sufficient for somebody who
relies on the GUI.

This error appears for many other problems too, so even using a search
engine doesn't help a user work out what is wrong.

I used "gpg --edit-key " to see my keys and this made it obvious the
key had expired.

On the machine with my master private key, I ran the following commands
to update the expiry on the subkeys:

gpg --edit-key
key 2
key 3
expire
365
save


and then I sent my key out to the keyservers, exported the whole key to
a file and imported it on the laptop where subkeys are used.

It would be really useful to have some of this automated through
Enigmail too: it could:

a) show a message telling you when the subkey expired

b) remind you to update it on the machine with the master key

c) offer to try and fetch the updated key from a keyserver or import it
from a file

As a bonus, maybe it can even warn the user 30 days before it expires?

Regards,

Daniel

Bug#870058: xsane failing since stretch upgrade

[Daniel Pocock]

Package: xsane
Version: 0.999-5
Severity: serious

Since upgrading to stetch, the system won't scan with xsane.

scanimage works from the command line.

In gimp, I choose my device and the popup appears:


Title: xsane 0.999
Message: scanning for devices



and then after a few seconds:


Title: Error
Message: Failed to open device
`hpaio:/net/HP_LaserJet_?zc=abcd1234': Invalid argument.




Looking at /var/log/syslog I find this:

12:45:26 xsane: io/hpmud/pml.c 348: SetPml failed ret=12
12:45:54 xsane: io/hpmud/hpmud.c 373: device_cleanup: device
uri=hp:/net/HP_LaserJet_?zc=abcd1234
12:45:54 xsane: io/hpmud/hpmud.c 385: device_cleanup: close device dd=1...
12:45:54 xsane: io/hpmud/hpmud.c 387: device_cleanup: done closing
device dd=1

Bug#867548: playback to HDMI monitor stutters / pauses

[Daniel Pocock]

On 07/07/17 18:07, Felipe Sateler wrote:
> Control: tags -1 moreinfo
> 
> On Fri, Jul 7, 2017 at 4:53 AM, Daniel Pocock <dan...@pocock.pro
> <mailto:dan...@pocock.pro>> wrote:
> 
> Package: pulseaudio
> Version: 10.0-1
> Severity: important
> 
> When playing audio through the HDMI monitor, I've noticed it
> stutters/pauses every few seconds, very briefly, less than a second each
> time.  It is very irritating to listen to.
> 
> In the middle of playback, I can open the GNOME settings panel for sound
> and switch the output to another device (e.g. my USB sound card
> connected to an amp with an optical cable) and it works fine.  As soon
> as I switch back to the HDMI output the problem comes back.
> 
> 
> Could you attach a verbose log please?
> 
> https://wiki.ubuntu.com/PulseAudio/Log
>  


After following those instructions, I don't hear any playback at all

I notice the killall command doesn't actually kill every pulseaudio
process, there is a process running as user Debian-gdm.  Do I need to do
something to stop that too before logging will work?

Regards,

Daniel

Bug#867548: playback to HDMI monitor stutters / pauses

[Daniel Pocock]

Package: pulseaudio
Version: 10.0-1
Severity: important

When playing audio through the HDMI monitor, I've noticed it
stutters/pauses every few seconds, very briefly, less than a second each
time.  It is very irritating to listen to.

In the middle of playback, I can open the GNOME settings panel for sound
and switch the output to another device (e.g. my USB sound card
connected to an amp with an optical cable) and it works fine.  As soon
as I switch back to the HDMI output the problem comes back.

Bug#866115: [Pkg-monitoring-maintainers] Bug#866115: dist-upgrade to Stretch broke ganglia-web

[Daniel Pocock]

On 29/06/17 08:49, Alois Schloegl wrote:
> 
> 
> On 2017-06-27 15:15, Daniel Pocock wrote:
>>
>>
>> On 27/06/17 14:58, Alois Schloegl wrote:
>>> Package: ganglia-web
>>>
>>>
>>> The following ganglia packages are installed:
>>>
>>> ganglia:~# dpkg -l|grep ganglia
>>> ii  ganglia-modules-linux 1.3.6-2
>>> amd64Ganglia extra modules for Linux (IO, filesystems,
>>> multicpu)
>>> ii  ganglia-monitor   3.6.0-7+b1
>>> amd64cluster monitoring toolkit - node daemon
>>> ii  ganglia-monitor-python3.6.0-7
>>> all  cluster monitoring toolkit - python modules
>>> ii  ganglia-webfrontend   3.6.1-3
>>> all  cluster monitoring toolkit - web front-end
>>> ii  libganglia1   3.6.0-7+b1
>>> amd64cluster monitoring toolkit - shared libraries
>>>
>>>
>>> After dist-upgrading from Debian 8 to Debian 9, the ganglia web
>>> interface is not working anymore, it just shows this error message.
>>>
>>>
>>> === START-OF-MESSAGE ===
>>>
>>> Sorry, you do not have access to this resource.
>>> "); } try { $dwoo = new Dwoo($conf['dwoo_compiled_dir'],
>>> $conf['dwoo_cache_dir']); } catch (Exception $e) { print "
>>> There was an error initializing the Dwoo PHP Templating Engine: ".
>>> $e->getMessage() . "
>>>
>>> The compile directory should be owned and writable by the apache user.
>>
>>
>> On my system it is working, I see the following owner and permissions:
>>
>> $ ls -la /var/lib/ganglia-web/dwoo/
>> total 16
>> drwxr-xr-x 4 www-data root 4096 Jan 22  2015 .
>> drwxr-xr-x 4 root root 4096 Jan 22  2015 ..
>> drwxr-xr-x 2 www-data root 4096 Apr  4  2014 cache
>> drwxrwxrwx 3 www-data root 4096 Mar 16  2015 compiled
>>
>>>
>>>
>>> Do you have any suggests for debugging this issue ?
>>>
>>
>> Please send the output of this command:
>>
>> $ ls -la /var/lib/ganglia-web/dwoo/
>>
> 
> 
> These permissions are the same at my machine.
> 
> ganglia:~#  ls -la /var/lib/ganglia-web/dwoo/
> insgesamt 16
> drwxr-xr-x 4 www-data root 4096 Jun 27 12:37 .
> drwxr-xr-x 4 root root 4096 Jun 27 12:37 ..
> drwxr-xr-x 2 www-data root 4096 Jul  7  2016 cache
> drwxrwxrwx 3 www-data root 4096 Jul 27  2015 compiled
> 
> 
> I tried further, purging and reinstalling ganglia-web and apache did not
> help.
> What solved the problem was enabling all modules with
> 
>a2enmod
> Your choices are: access_compat actions alias allowmethods asis
> auth_basic auth_digest auth_form auth_kerb authn_anon authn_core
> authn_dbd authn_dbm authn_file authn_socache authnz_fcgi authnz_ldap
> authz_core authz_dbd authz_dbm authz_groupfile authz_host authz_owner
> authz_user autoindex buffer cache cache_disk cache_socache cern_meta cgi
> cgid charset_lite data dav dav_fs dav_lock dbd deflate dialup dir
> dump_io echo env expires ext_filter file_cache filter gnutls headers
> heartbeat heartmonitor http2 ident imagemap include info
> lbmethod_bybusyness lbmethod_byrequests lbmethod_bytraffic
> lbmethod_heartbeat ldap log_debug log_forensic lua macro mime mime_magic
> mpm_event mpm_itk mpm_prefork mpm_worker negotiation perl php7.0 proxy
> proxy_ajp proxy_balancer proxy_connect proxy_express proxy_fcgi
> proxy_fdpass proxy_ftp proxy_hcheck proxy_html proxy_http proxy_http2
> proxy_scgi proxy_wstunnel ratelimit reflector remoteip reqtimeout
> request rewrite sed session session_cookie session_crypto session_dbd
> setenvif slotmem_plain slotmem_shm socache_dbm socache_memcache
> socache_shmcb speling ssl status substitute suexec unique_id userdir
> usertrack vhost_alias xml2enc
> 
> and enabling all of them solved the problem.
> 
> I'm not sure which of them I really need, but suspect that there is a
> missing dependency on either to php7 or gnutls module.
> 


On one of my systems running ganglia-web I see the following enabled:

$ sudo a2query -m
alias (enabled by unknown)
authz_groupfile (enabled by unknown)
authn_file (enabled by unknown)
deflate (enabled by unknown)
setenvif (enabled by unknown)
authz_user (enabled by unknown)
authz_core (enabled by maintainer script)
mpm_prefork (enabled by site administrator)
status (enabled by unknown)
php7.0 (enabled by site administrator)
env (enabled by unknown)
socache_shmcb (enabled by maintainer script)
access_compat (enabled by maintainer script)
cgi (enabled by unknown