te, as Didier says, they are more of a hindrance than a help as
they make ourselves ignore real test failures.
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.n
also close the OSK with a
double-tap above it (in the empty area left of the Send button).
Does anyone know some actual specification for the intended behaviour,
so that we can encode that in the test?
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu D
I guess we can adjust the swipe to
start a tad below the OSK edge so that it'll actually close.
Thanks!
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-ph
, using the
reply_handler/error_handler kwargs:
http://dbus.freedesktop.org/doc/dbus-python/doc/tutorial.html#making-asynchronous-method-calls
I've used this approach in good old Jockey back then for backend
method calls which can take minutes.
Martin
--
Martin Pitt
dbus
However, there's not a lot of existing examples there, it seems few
people ever do this.
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post t
landing-plan task to the existing bug, so that all information is in
one place? If it's a bug-less new feature or it fixes several bugs, a
new landing-plan bug report is of course better.
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Develop
Robert Park [2013-12-17 17:35 -0800]:
> I've collected some useful links. This is the kind of stuff that a
> spreadsheet just can't do.
Nice! Would you mind putting these links into the project description
at https://launchpad.net/landing-plan ? That makes them easier to
find.
M
ore-dev (but not manager)
always have to track down someone in IRC to do this busywork for me.
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post
e.launchpad.net/~pitti/dialer-app/maguro-tests/+merge/193774
. I'll restart it now to see where we stand there.
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
signature.asc
Description: Digita
Martin Pitt [2013-12-20 6:53 +0100]:
> The new stack trace in
> https://jenkins.qa.ubuntu.com/job/trusty-touch-mako-smoke-dialer-app-autopilot/72/artifact/clientlogs/_usr_bin_dialer-app.32011.crash
> is still Qt/Mir related, but looks different, so I take it we'll need
> a new bu
Martin Pitt [2013-12-20 6:53 +0100]:
> For the record, Mir crashes from the dialer-app tests aren't something
> new, I filed a long-standing one in October already:
> https://launchpad.net/bugs/1240400 This was a case where switching
> from SF to Mir caused crashes, but at th
which
does that second-phase information collection and then tag it for
whoopsie upload.
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post to :
Rasmus Eneman [2014-01-31 1:30 +0100]:
> Systemd? Have I missed something important?
systemd as in the project, not as PID 1. We've been using it for a
long time (udev, logind, timedated, etc.)
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu D
y same approach should work for apps as well. Please talk to
me if you want to do this and you run into trouble. There's certainly
lots of stuff that we can simplify or improve (e. g. providing
higher-level macros in the test sensor format, etc.).
Thanks,
Martin
--
Martin Pitt
about that?
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post to : ubuntu-phone@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ubu
kill the "apport" instance, of course.
Do we actually have evidence that core dumps are incomplete due to
killing unity8, or is that rather that we don't get core dumps at all?
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | De
there, then backported to our ofono-rilmodem branch:
https://github.com/rilmodem/ofono/pull/56
Waiting for upstream to comment/land that, then we can do the
correspondinging packaging changes (which are rather easy to do, and
Dimitri already did them anyway in a branch of his').
Marti
ready have a very low landing rate. This is
now the anti-thesis of our original "velocity" goal.
I think the compromise of keeping the development series going, but
simply keep daily images as "-proposed" until that is fixed is the
right way forward here.
Martin
--
Martin
ng all developers in
the world would consistently create back buttons in the UI, or
otherwise create an undiscoverable inconsistency. So from that POV
putting it in the current location on the top left title bar is a good
decision.
Martin
--
Martin Pitt| http://www.piware
ne a lot more during the day than
at night. A white-on-black "night mode" based on the illumination
sensor or just the time of day would be extra-cool, of course :)
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post to : ubuntu-phone@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ubuntu-phone
More help :
n)
Dimitri proposed a fix for that a month ago:
https://code.launchpad.net/~xnox/dialer-app/fix-dial-number-crash/+merge/209677
But that still didn't land.
This is https://bugs.launchpad.net/dialer-app/+bug/1287628, FTR.
Martin
--
Martin Pitt| http://www.
nline vs. .mo translations for a
few strings isn't going to reach a significant size.
So I think which approach to use should be decided based on how we
want the translation lifecycle/rollout to look like only.
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Devel
n why this can't be deciced on a per-package level if necessary,
and as you say there might indeed be phone/click/scope specic reasons
why we'd want one or the other. So I'm not particularly favouring
either approach here as I haven't discussed that in detail with other
people
, or we don't care about
updating translations separately, let's avoid the hassle completely.
In the "update whole system as an image" the reasons for having
langpacks in the first place are much smaller than for the classic
.deb based desktop.
Thanks,
Martin
--
Martin Pitt
alls.
Well, it's going to be two different systems no matter what, unless we
want to abolish the langpacks for desktop as well. But we have lived
with two ways of shipping translations since breezy or so :-)
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer
. But OTOH when we don't
ship translations in langpacks but in the apps themselves, we don't
have the option at all to enable/disable languages individually
because we'd always ship with all of them.
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu De
designed. We've had the gnome/kde/common
split for ages because that was what our main products at that time
(Ubuntu/Kubuntu) needed. These days, Kubuntu has its own langpacks
(kde-l10n-*) so we should indeed reorg our current langpack split more
along the desktop/phone lines.
service (or also
/run/systemd/system/foo.service for one-shot replacements).
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post to : u
update-rc.d in Ubuntu will work for all three
systems (Debian's already does, we need to merge), so IMHO update-rc.d
should be considered the standard interface for disabling/enabling
services with any init system.
Please let's kill default files whose only purpose i
sion. I. e. in
each upstart job/systemd units you have to add ugly error handling
for the case when you disable the service with /etc/default. I. e.
A job/unit starts, but then really doesn't, there isn't a service
around, the init system thinks it's crashed,
rite
access to customizers. There's hardly anything to customize then.
Also, why would someone who wants to build a custom image start from
our default Ubuntu one instead of building one for himself with
whichever customizations?
Thanks,
Martin
--
Martin Pitt| http:/
#x27;d need to restrict
writeability to e. g. /etc/systemd/system/*.wants/ .
But even with the ability to disable arbitrary services you can
already ruin/change the user experience quite a bit, so we still need
to be careful there what we allow there.
Thanks,
Martin
--
Martin Pitt
ngs like autopilot
we can certainly see how to run them without dpkg-installing -- like,
unpack them in /tmp and set $PATH, $PYTHONPATH etc. But it's rather
limiting to say "your tests can't have apache/dbusmock/samba/lcov/etc"
just because they are only available as .deb packa
t be other search queries which are more appropriate of
course. That doesn't cover click packages, but at least the .deb
ones.
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https:
ier to do, you can also drop use-session-dbus from
/etc/X11/Xsession.options.
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
signature.asc
Description: Digital signature
--
Mailing list: https:
Martin Pitt [2014-06-03 7:41 +0200]:
> > dbus-x11 - which only contains /etc/X11/Xsession.d/75dbus_dbus-launch
> > and depends on the dbus-launch package
>
> In case it's easier to do, you can also drop use-session-dbus from
> /etc/X11/Xsession.options.
That'
er way, the bug would be that the process
doesn't have $DBUS_SESSION_BUS_ADDRESS in the first place, and it
would just behave "differently wrongly" in either case, but never
correct.
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer
uild products out of them..), this isn't necessary, as packages in
main are automatically covered by langpacks.
Thanks,
Martin
[1]
https://blueprints.launchpad.net/ubuntu/+spec/development-1406-touch-language-packs
--
Martin Pitt| http://www.piware.de
Ubuntu Develo
stro, I was told that this wouldn't work either.
Would it work to push to lp:project *and* upload? Would the CI train
consider this as a consistent change?
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.o
Hello again,
Martin Pitt [2014-06-13 15:38 +0200]:
> Thus, we need this addition to debian/control's Source: header:
>
> X-Ubuntu-Use-Langpack: yes
As Didier and Robert assured that it's ok to push the change to the
branch and upload, I did that now for dialer-app. I or
Martin Pitt [2014-06-16 9:36 +0200]:
> On Wed, with the new LP export, I'll double-check that everything is
> alright and release the rebuilt dialer-app into utopic. Then we can
> tackle the other dozen projects.
FTR, this happened now. language-pack-touch-* have the dialer-app
tr
the
other ~ 20 projects (see the other thread), but with dialer-app it was
shown that the plumbing of translations works.
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: http
a more simple testing for that.
I'll bug Ted about that next week :)
Indeed with these two pieces we can then run any Debian or click
source test in any of the schroot/QEMU/LXC/ssh/null runners that
autopkgtest provides :)
Martin
--
Martin Pitt| http://www
buntu/+source/gdb/+bug/1233185
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
signature.asc
Description: Digital signature
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post to : ubuntu-
adb (and another one for nova, as a Proof of Concept -- go wild).
> Modifications to autopkg to allow it to read and understand clicks (pull
> depends from them, provision them, etc)
Done.
Feedback appreciated!
Martin
--
Martin Pitt| http://www.piware.de
U
erything including networking,
the setup script could also certainly save the network config, wipe,
and restore the network config.
Thanks for any insights!
Martin
[1]
http://www.piware.de/2014/07/deb-click-schroot-lxc-qemu-phone-cloud-one-autopkgtest-to-rule-them-all/
[2] https://wiki.ubuntu.com/Tou
, or
sending Unity a D-BUS call?
Thanks for any insights!
Martin
[1]
http://www.piware.de/2014/07/deb-click-schroot-lxc-qemu-phone-cloud-one-autopkgtest-to-rule-them-all/
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debi
ly easy to replicate "factory-reset" with a
bunch of adb commands, or should we rather assume ubuntu-device-do as
a dependency in the CI airline if we want to reset phones quickly?
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (w
> what you want.
That seems fine, at least as long as we don't ship the phone by
default to require a password. So I think it actually might be what I
want :-)
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Develop
Martin Pitt [2014-07-02 17:16 +0200]:
> So I think there ought to be a simple way (D-BUS call, flag file,
> etc.) to either disable the greeter (and then have to restart unity),
> or unlock it with a D-BUS call (not a security issue, you are already
> logged into the device anyway).
I
Martin Pitt [2014-07-03 10:11 +0200]:
> I filed https://bugs.launchpad.net/unity8/+bug/1337188 about that, as
> it's a real blocker for running tests on the phone in a sensible and
> robust way.
Actually, that's not true. Starting an app causes the unity lock
screen/greeter
e full root/boot loader access
anyway so this is pretty much the only time when we can configure
this?
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubun
o give you user adb).
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post to : ubuntu-phone@lists.launchpad.net
Unsubscribe : https://launchpa
is out. This is a consequence of fixing LP
#1349579, but the crash happens after the important information for
whoopsie has been collected. Thus it's purely cosmetical.
Fixed in http://bazaar.launchpad.net/~apport-hackers/apport/trunk/revision/2829
Martin
--
Martin Pitt
suppose for both things we need a way to enable sudo when flashing
the device. Will/does ubuntu-device-flash --developer-mode do that? I.
e. we'll set up our test devices with that once, and from then on adb
will stay root even after factory reset?
Thanks,
Martin
--
Martin Pitt
Hello all,
TL;DR: various things will be untranslated on the phone in the next
days.
Martin Pitt [2014-06-18 13:26 +0200]:
> FTR, this happened now. language-pack-touch-* have the dialer-app
> translations, so the next ubuntu-phone image build will have its
> dialertranslations back. So
/
Another question: If your goal is to lock down the device, why do we
start adb by default? It would make much more sense to only enable
that on demand -- when you use your phone, all you (should) need is
MTP, an always running adbd allows you to circumvent the PIN
restriction and access all your da
s become a new phablet-factory-reset tool, or be integrated
into phablet-config, or something else?
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
signature.asc
Description: Digital signature
--
Martin Pitt [2014-08-21 11:02 +0200]:
> > well patches to phablet-tools (or dbus-property-service which is the
> > on-device side of this) will indeed be happily accepted to add more
> > features ... my current focus is to get all of phablet-config,
> > phablet-shell and
re arch: all), or also from
ubuntu to any PPA (i. e. silo).
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post to : ubuntu-phone@li
x27;t just copy that yesterday as there already is a previous
upload which isn't in RTM and which I wasn't sure about. Discussed
with Bill, and he said he's currently re-syncing messaging-app with
trunk/utopic, so that should land soon.
Martin
--
Martin Pitt
/ubuntu-rtm/queue/, unpack the tarball
containing the ddebs, and install those manually.
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
signature.asc
Description: Digital signature
--
Mailing list: https
am, does this need
some particular magic to get enabled? Could we do that for the silo
PPAs?
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
signature.asc
Description: Digital signature
--
Mailing
ones.
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post to : ubuntu-phone@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ubuntu-p
atically unlock Unity. So chicken - egg again.
Since adb doesn't run by default anyway, and one has to explicitly
enable it with "developer mode", what's the thing that we are trying
to prevent here?
Martin
--
Martin Pitt| http://www.piware.de
Ub
x27;s at least a known issue now.
Thanks,
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
--
Mailing list: https://launchpad.net/~ubuntu-phone
Post to : ubuntu-phone@lists.launchpad.net
Unsubscribe : htt
rch 25. I refreshed it a couple of days after that, but forgot to
push it to the keyservers. Langpack updates should be back soon now,
I'll keep an eye on it.
Thanks for pointing out!
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | De
Martin Pitt [2015-04-08 11:40 +0200]:
> Heroldich Robin [2015-04-08 8:13 +0200]:
> > Łukasz Zemczak mentioned in a previous landing mail that the phone
> > translations update at every Tuesdays, but as I see, it only updated
> > at 24th, two weeks ago.
>
> Confirmed. T
th bugs like this and getting an
ever-growing spaghetti heap of file bind mounts which we'll never be
able to revert/change.
Martin
--
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
signature.asc
Descrip
70 matches
Mail list logo
