RE: screensaver and power manager dbus interfaces

The screensaver interface looks good. What is the use case for the
"Poke" method? 

Power manager stuff looks nice as well, but maybe the KDE HW ppl can
comment on that a little bit better.

I think this fits in really nice with the long-term vision that Portland
has with DAPI (Deskop API). I would like to suggest to rename the
interface org.gnome.* to org.freedesktop.* and host them as part of DAPI
in the Portland CVS (See portland.freedesktop.org). And then we should
get some DBUS love for DAPI going I guess.

Waldo Bastian
Linux Client Architect - Client Linux Foundation Technology
Channel Platform Solutions Group
Intel Corporation - http://www.intel.com/go/linux
OSDL DTL Tech Board Chairman

>-----Original Message-----
>From: xdg-bounces <at> lists.freedesktop.org [mailto:xdg-
>bounces <at> lists.freedesktop.org] On Behalf Of Richard Hughes
>Sent: Wednesday, May 31, 2006 3:02 PM
>To: xdg <at> lists.freedesktop.org
>Cc: David Zeuthen
>Subject: screensaver and power manager dbus interfaces
>
>Okay, my first post to this list, so I hope I'm aiming in the right
>direction.
>
>gnome power manager :		org.gnome.PowerManager

dbus test failures on Mac OS X

I presume kde-core-devel is the place to post this stuff...

Most of the tests pass, but it looks like a number of them bomb out:

********* Start testing of tst_Hal *********
Config: Using QTest library 4.1.3, Qt 4.1.3
PASS   : tst_Hal::initTestCase()
FAIL!  : tst_Hal::getDevices() 'con.isConnected()' returned FALSE. ()
   Loc: [/Users/ranger/cvs/kdesupport/qt-dbus/selftests/hal/tst_hal.cpp(45)]
FAIL!  : tst_Hal::lock() 'con.isConnected()' returned FALSE. ()
   Loc: [/Users/ranger/cvs/kdesupport/qt-dbus/selftests/hal/tst_hal.cpp(60)]
PASS   : tst_Hal::cleanupTestCase()
Totals: 2 passed, 2 failed, 0 skipped
********* Finished testing of tst_Hal *********

********* Start testing of tst_QDBusAbstractAdaptor *********
Config: Using QTest library 4.1.3, Qt 4.1.3
...
PASS   : tst_QDBusAbstractAdaptor::typeMatching()
PASS   : tst_QDBusAbstractAdaptor::cleanupTestCase()
Totals: 9 passed, 0 failed, 0 skipped
********* Finished testing of tst_QDBusAbstractAdaptor *********

********* Start testing of tst_QDBusConnection *********
Config: Using QTest library 4.1.3, Qt 4.1.3
PASS   : tst_QDBusConnection::initTestCase()
...
PASS   : tst_QDBusConnection::cleanupTestCase()
Totals: 8 passed, 0 failed, 0 skipped
********* Finished testing of tst_QDBusConnection *********

Re: dbus test failures on Mac OS X

Benjamin Reed wrote:
>********* Start testing of tst_Hal *********
>Config: Using QTest library 4.1.3, Qt 4.1.3
>PASS   : tst_Hal::initTestCase()
>FAIL!  : tst_Hal::getDevices() 'con.isConnected()' returned FALSE. ()
>   Loc:
> [/Users/ranger/cvs/kdesupport/qt-dbus/selftests/hal/tst_hal.cpp(45)]
> FAIL!  : tst_Hal::lock() 'con.isConnected()' returned FALSE. () Loc:
> [/Users/ranger/cvs/kdesupport/qt-dbus/selftests/hal/tst_hal.cpp(60)]
> PASS   : tst_Hal::cleanupTestCase()
>Totals: 2 passed, 2 failed, 0 skipped
>********* Finished testing of tst_Hal *********

No connection to the system bus. Is the system bus daemon running? Is HAL 
running?

I don't think MacOS X uses HAL, so probably not.

>QDEBUG : tst_QDBusMarshall::sendVariant(bool) sending message:
>QDBusMessage(type=MethodCall, service="org.kde.selftest",
>path="/org/kde/selftest", interface="org.kde.selftest", name="ping",
>signature="", contents=(QVariant(bool(false)) ) )
>FAIL!  : tst_QDBusMarshall::sendVariant(bool) Compared values are not
> the same Actual (reply.signature()):
>   Expected (QString("v")): v
>   Loc:
> [/Users/ranger/cvs/kdesupport/qt-dbus/selftests/qdbusmarshall/tst_qdbus
>marshall.cpp(225)]
[snip]

RE: screensaver and power manager dbus interfaces

On Wed, 2006-05-31 at 17:48 -0700, Bastian, Waldo wrote:
> The screensaver interface looks good. What is the use case for the
> "Poke" method? 
> 
I guess it can be used by apps like video players to prevent the
screensaver to fire while showing the movie.
--

-- 
Rodrigo Moya <rodrigo <at> gnome-db.org>

Re: screensaver and power manager dbus interfaces

Rodrigo Moya wrote:
>On Wed, 2006-05-31 at 17:48 -0700, Bastian, Waldo wrote:
>> The screensaver interface looks good. What is the use case for the
>> "Poke" method?
>
>I guess it can be used by apps like video players to prevent the
>screensaver to fire while showing the movie.

It should be enough to disable the screensaver and re-enable it when done.

Disable/enable should be ref-counted so that if two applications ask for 
the screensaver to be disabled, both have to ask for it to be turned on 
again before it is.

--

-- 
Thiago Macieira  -  thiago (AT) macieira.info - thiago (AT) kde.org
  thiago.macieira (AT) trolltech.com     Trolltech AS
    GPG: 0x6EF45358                   |  Sandakerveien 116,
    E067 918B B660 DBD1 105C          |  NO-0402
    966C 33F5 F005 6EF4 5358          |  Oslo, Norway

_______________________________________________
xdg mailing list
xdg <at> lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xdg

Re: screensaver and power manager dbus interfaces

On Thursday 01 June 2006 6:50 am, Thiago Macieira wrote:
> Disable/enable should be ref-counted so that if two applications ask for
> the screensaver to be disabled, both have to ask for it to be turned on
> again before it is.

(Lurking)

That seems a bit odd to me.  It raises all kinds of questions - what about 
calls coming from schell scripts via DBUS? Or, what if a calling app that 
disables the screensaver is killed off or dies before re-enabling it?

A periodic 'Poke' by any app wanting to disable it seems safer. Then if the 
app is hung or dies the 'Poke' stops happening, so the screen-saver is 
re-activated. 

RE: screensaver and power manager dbus interfaces

On Wed, 2006-05-31 at 17:48 -0700, Bastian, Waldo wrote:
> The screensaver interface looks good. What is the use case for the
> "Poke" method? 

If the screensaver is running then it simulates user activity and shows
the unlock box. We call this in gnome-power-manager to show the unlock
box automatically after resume.

Maybe "Poke" is a rubbish name....

Richard.

Re: dbus test failures on Mac OS X

On 6/1/06, Thiago Macieira <thiago <at> kde.org> wrote:

> No connection to the system bus. Is the system bus daemon running? Is HAL
> running?
>
> I don't think MacOS X uses HAL, so probably not.

Ah, yeah, no system bus, I'd only started a session bus.  In theory,
HAL could be ported and send events just as easily as on linux, but I
doubt anyone's done so yet.

That at least explains it.  :)

> The interesting thing is that there is no "got message" for the reply. I
> wonder if this is something I did wrong in the debugging or if no reply
> actually happened.
>
> Please try running "qpong/qpong &" manually before tst_qdbusmarshall so
> that the debugging output for the ping-pong is shown too.
>
> It looks like the qpong program died or something, because it stopped
> replying completely.

That did it.  Now everything but one test (and hal) work:

QDEBUG : tst_QDBusMarshall::sendStringMapOfMap(bytearray) got message:
QDBusMessage(type=MethodReturn, service=":1.15", path="",
interface="", name="", signature="a{sa{say}}",
contents=(QVariantMap(("foo", QVariantMap(("bar", QByteArray("foo"))
("foo", QByteArray("foo")) )) ) ) )

Re: screensaver and power manager dbus interfaces

On 6/1/06, Jason Keirstead <jason <at> keirstead.org> wrote:
> A periodic 'Poke' by any app wanting to disable it seems safer. Then if the
> app is hung or dies the 'Poke' stops happening, so the screen-saver is
> re-activated.
>

For this to work the app calling poke also needs to know the
inactivity time out to ensure it pokes often enough. This seems to be
missing from the interface.

Rich.

Re: screensaver and power manager dbus interfaces

On Thu, 2006-06-01 at 10:07 -0400, William Jon McCann wrote:
> Hi Waldo,
> 
> Bastian, Waldo wrote:
> > The screensaver interface looks good. What is the use case for the
> > "Poke" method? 
> 
> The Poke method is a way to simulate user input.  It is the programmatic 
> equivalent to moving the mouse back and forth or hitting the Shift key 
> or something.  If the screensaver is activated and the lock is enabled 
> then it will prompt for authentication.  If the screensaver is activated 
> and the lock is disabled then it will deactivate the screensaver.  If 
> the screensaver is not active (ie. blanked) then it will reset the idle 
> timers.

Notably this is used by gnome-power-manager when resuming from a suspend
state, yes? 

For example, the user opens the laptop lid (and is careful not to hit
the keyboard or mouse), g-p-m gets notified and pokes the screensaver
and either the session comes back (if lock is disabled) or the unlock
dialog is shown. Without the poke this probably wouldn't happen.

    David