Re: [Sugar-devel] A network question

[James Cameron]

On Fri, Nov 11, 2011 at 04:05:49PM +1000, Peter Hewitt wrote:
> And I assume it should be
> sudo ifconfig eth0 10.0.0.1
> not
> sudo iwconfig eth0 10.0.0.1

Indeed, yes.

> Seeing I had an established ad-hoc network I just configured the
> "bad" computer and made sure the "inet addr" line agreed with the
> other two machines on the ad hoc network. I assigned "169.254.7.200"
> to B. Pinging to or from it produced

I've not tested that configuration, sorry.

-- 
James Cameron
http://quozl.linux.org.au/
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[Peter Hewitt]

Not exactly as per your instructions but I think equivalent.

And I assume it should be

sudo ifconfig eth0 10.0.0.1

not

sudo iwconfig eth0 10.0.0.1

Seeing I had an established ad-hoc network I just configured the "bad" 
computer and made sure the "inet addr" line agreed with the other two 
machines on the ad hoc network. I assigned "169.254.7.200" to B. Pinging to 
or from it produced


Destination Host Unreachable.





At 03:39 PM 11/11/2011, James Cameron wrote:

On Fri, Nov 11, 2011 at 03:23:25PM +1000, Peter Hewitt wrote:
> "Ad-hoc Network 1" only appears on G and then only if I have an XO
> 1.5 switched on.

Okay, so hardware is working, firmware is working, some of the software
is working.

I suggest next you gain manual control over networking and see if you
can configure the units to join a new ad-hoc network.  By doing this you
will exclude Sugar and NetworkManager.

On each unit, temporarily turn off NetworkManager ... this will be
turned back on when the unit reboots.

sudo /etc/init.d/NetworkManager stop

Now, raise the interface:

sudo ifconfig eth0 up

Now, join an ad-hoc network:

sudo iwconfig eth0 mode ad-hoc essid xo channel 1

Now, set the addresses of each interface, to a unique number:

sudo iwconfig eth0 10.0.0.1

(increment the last digit for each unit),

Then, test that you can ping each unit from each other unit, e.g. on the
unit that has the address 10.0.0.1, ping the unit that has the address
10.0.0.2:

sudo ping 10.0.0.2

When you get packets back, you know it is working:

$ ping -n 10.0.0.144
PING 10.0.0.144 (10.0.0.144) 56(84) bytes of data.
64 bytes from 10.0.0.144: icmp_req=1 ttl=64 time=13.7 ms
64 bytes from 10.0.0.144: icmp_req=2 ttl=64 time=3.03 ms
64 bytes from 10.0.0.144: icmp_req=3 ttl=64 time=2.84 ms
^C
--- 10.0.0.144 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2002ms
rtt min/avg/max/mdev = 2.843/6.539/13.737/5.090 ms

Once you have done this test, I can say the problem is some sort of
corruption of configuration of the installed operating system, and you
can fix it by reinstalling.

--
James Cameron
http://quozl.linux.org.au/


___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[James Cameron]

On Fri, Nov 11, 2011 at 03:23:25PM +1000, Peter Hewitt wrote:
> "Ad-hoc Network 1" only appears on G and then only if I have an XO
> 1.5 switched on.

Okay, so hardware is working, firmware is working, some of the software
is working.

I suggest next you gain manual control over networking and see if you
can configure the units to join a new ad-hoc network.  By doing this you
will exclude Sugar and NetworkManager.

On each unit, temporarily turn off NetworkManager ... this will be
turned back on when the unit reboots.

sudo /etc/init.d/NetworkManager stop

Now, raise the interface:

sudo ifconfig eth0 up

Now, join an ad-hoc network:

sudo iwconfig eth0 mode ad-hoc essid xo channel 1

Now, set the addresses of each interface, to a unique number:

sudo iwconfig eth0 10.0.0.1

(increment the last digit for each unit),

Then, test that you can ping each unit from each other unit, e.g. on the
unit that has the address 10.0.0.1, ping the unit that has the address
10.0.0.2:

sudo ping 10.0.0.2

When you get packets back, you know it is working:

$ ping -n 10.0.0.144
PING 10.0.0.144 (10.0.0.144) 56(84) bytes of data.
64 bytes from 10.0.0.144: icmp_req=1 ttl=64 time=13.7 ms
64 bytes from 10.0.0.144: icmp_req=2 ttl=64 time=3.03 ms
64 bytes from 10.0.0.144: icmp_req=3 ttl=64 time=2.84 ms
^C
--- 10.0.0.144 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2002ms
rtt min/avg/max/mdev = 2.843/6.539/13.737/5.090 ms

Once you have done this test, I can say the problem is some sort of
corruption of configuration of the installed operating system, and you
can fix it by reinstalling.

-- 
James Cameron
http://quozl.linux.org.au/
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[Peter Hewitt]

At 03:09 PM 11/11/2011, James Cameron wrote:

On Fri, Nov 11, 2011 at 02:54:33PM +1000, Peter Hewitt wrote:
> One further piece of info before answering your questions:
>
> ifconfig on B has NO "inet addr" line - otherwise appears ok (apart
> from the previously mention lack of an eth0 entry).
>
> At 02:34 PM 11/11/2011, James Cameron wrote:
> >I'm presuming identical result to that seen before, as opposed to
> >identical result between the two XO-1s?  It's important.
>
> I'm not sure what you're asking here. I should emphasize that I have
> no "access point" - just 2 XOs.

In that case the results of "test /wlan" or "scan-wifi" will depend on
whether the other XO is running an ad-hoc network, and on any other
access points in your general vicinity ... 150m built-up area, 300m open
area, 2km if on top of a mountain.

> If I keep repeating the "test /wlan" command on either machine (both
> at the ok prompt), the RSSI number keeps changing and the number of
> results can be 1, 2 or 3.

Now that I know you don't have an access point, the "test /wlan" output
text is not particularly useful, except to prove that both units can
receive broadcasts from either your other unit running ad-hoc, or other
people's access points.

> >ok scan-wifi
> >ok dev /wlan
> >ok .properties
>
> Both XOs produce a sensible mac-address.

Good, that excludes USB problems.

Nothing wrong with the hardware.  This must be a software problem.

Boot to Sugar, press F3, right-click on the centre icon, select My
Settings, then Network, and make sure the Radio checkbox has a tick in
it.

Feel free also to "Discard network history" on that screen.


Done ... many times :o)


Then, press F1, and on both laptops click on "Ad-hoc Network 1", then
wait.  Shortly, icons for the other XO should appear on each XO.


"Ad-hoc Network 1" only appears on G and then only if I have an XO 1.5 
switched on.






___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[James Cameron]

On Fri, Nov 11, 2011 at 02:54:33PM +1000, Peter Hewitt wrote:
> One further piece of info before answering your questions:
> 
> ifconfig on B has NO "inet addr" line - otherwise appears ok (apart
> from the previously mention lack of an eth0 entry).
> 
> At 02:34 PM 11/11/2011, James Cameron wrote:
> >I'm presuming identical result to that seen before, as opposed to
> >identical result between the two XO-1s?  It's important.
> 
> I'm not sure what you're asking here. I should emphasize that I have
> no "access point" - just 2 XOs.

In that case the results of "test /wlan" or "scan-wifi" will depend on
whether the other XO is running an ad-hoc network, and on any other
access points in your general vicinity ... 150m built-up area, 300m open
area, 2km if on top of a mountain.

> If I keep repeating the "test /wlan" command on either machine (both
> at the ok prompt), the RSSI number keeps changing and the number of
> results can be 1, 2 or 3.

Now that I know you don't have an access point, the "test /wlan" output
text is not particularly useful, except to prove that both units can
receive broadcasts from either your other unit running ad-hoc, or other
people's access points.

> >ok scan-wifi
> >ok dev /wlan
> >ok .properties
> 
> Both XOs produce a sensible mac-address.

Good, that excludes USB problems.

Nothing wrong with the hardware.  This must be a software problem.

Boot to Sugar, press F3, right-click on the centre icon, select My
Settings, then Network, and make sure the Radio checkbox has a tick in
it.

Feel free also to "Discard network history" on that screen.

Then, press F1, and on both laptops click on "Ad-hoc Network 1", then
wait.  Shortly, icons for the other XO should appear on each XO.

-- 
James Cameron
http://quozl.linux.org.au/
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[Peter Hewitt]

One further piece of info before answering your questions:

ifconfig on B has NO "inet addr" line - otherwise appears ok (apart from 
the previously mention lack of an eth0 entry).


At 02:34 PM 11/11/2011, James Cameron wrote:

On Fri, Nov 11, 2011 at 02:22:53PM +1000, Peter Hewitt wrote:
> At 01:55 PM 11/11/2011, James Cameron wrote:
> >On Fri, Nov 11, 2011 at 01:47:10PM +1000, Peter Hewitt wrote:
> >> ok test net
> >>
> >> nothing on B
> >
> >Does the wireless LED blink during this test?
>
> Yes
>
> >Does the test take about two seconds?
>
> Yes
>
> And ... I have now discovered the command test /wlan
> which produces an identical result on both XO-1s but with different
> RSSI numbers.

I'm presuming identical result to that seen before, as opposed to
identical result between the two XO-1s?  It's important.


I'm not sure what you're asking here. I should emphasize that I have no 
"access point" - just 2 XOs.


If I keep repeating the "test /wlan" command on either machine (both at the 
ok prompt), the RSSI number keeps changing and the number of results can be 
1, 2 or 3.



ok scan-wifi
ok dev /wlan
ok .properties


Both XOs produce a sensible mac-address.

___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[James Cameron]

On Fri, Nov 11, 2011 at 02:22:53PM +1000, Peter Hewitt wrote:
> At 01:55 PM 11/11/2011, James Cameron wrote:
> >On Fri, Nov 11, 2011 at 01:47:10PM +1000, Peter Hewitt wrote:
> >> ok test net
> >>
> >> nothing on B
> >
> >Does the wireless LED blink during this test?
> 
> Yes
> 
> >Does the test take about two seconds?
> 
> Yes
> 
> And ... I have now discovered the command test /wlan
> which produces an identical result on both XO-1s but with different
> RSSI numbers.

I'm presuming identical result to that seen before, as opposed to
identical result between the two XO-1s?  It's important.

> there is no sign of a MAC address on either machine. The output from
> this command is the same on both machines.

My mistake.

ok scan-wifi
ok dev /wlan
ok .properties

-- 
James Cameron
http://quozl.linux.org.au/
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[Peter Hewitt]

At 01:55 PM 11/11/2011, James Cameron wrote:

On Fri, Nov 11, 2011 at 01:47:10PM +1000, Peter Hewitt wrote:
> ok test net
>
> nothing on B

Does the wireless LED blink during this test?


Yes


Does the test take about two seconds?


Yes

And ... I have now discovered the command test /wlan
which produces an identical result on both XO-1s but with different RSSI 
numbers.



If yes to both questions, perhaps an antenna cable has come lose or
broken.  Try it right next to your access point.


XOs side by side - no change to original symptoms - ie no Mesh icon pulsing 
on B



Lastly, the following test proves communication with the wireless device
itself:

ok dev /wlan
ok .properties
If you see a mac-address (e.g. 00 17 c4 .. .. ..) then some level of
communication was possible.


In case some other newbie is reading this ... there is a period before 
"properties" ... but ...


there is no sign of a MAC address on either machine. The output from this 
command is the same on both machines.


On my XO 1.5 it does list "fullmac" but still no sign of a MAC address.


James Cameron
http://quozl.linux.org.au/


___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[James Cameron]

On Fri, Nov 11, 2011 at 01:47:10PM +1000, Peter Hewitt wrote:
> ok test net
> 
> nothing on B

Does the wireless LED blink during this test?
Does the test take about two seconds?

If yes to both questions, perhaps an antenna cable has come lose or
broken.  Try it right next to your access point.

If the wireless LED does not blink, then check that all USB devices are
removed, since the wireless device shares the USB on XO-1.

Lastly, the following test proves communication with the wireless device
itself:

ok dev /wlan
ok .properties

If you see a mac-address (e.g. 00 17 c4 .. .. ..) then some level of
communication was possible.

-- 
James Cameron
http://quozl.linux.org.au/
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[Peter Hewitt]

ok test net

nothing on B

on G:

RSSI: 0  SSID:  Channel: 1
RSSI: 49  SSID:  Channel: 1


At 01:17 PM 11/11/2011, Jerry Vonau wrote:

On Fri, 2011-11-11 at 12:08 +1000, Peter Hewitt wrote:
> I'm finally starting to investigate networking on my XOs.
>
> I have two XO 1s - Firmware Q2E45 Wireless 5.110.22.p23 Sugar .84.16 
Build 852

>
> Let's call them G (good) & B (bad)
>
> The neighbourhood show three mesh symbols on both.
>
> On G one pulses but on B ... no pulsing.
>
> G finally says "Connected" on Mesh Network 1"
>
> When I click "Connect" on Mesh Network 1 on B nothing changes.
>
> ifconfig shows
>
> eth0, lo & msh0 on G but just lo & msh0 on B
>
> I'm sure they both connected fine when they first arrived about 6 
months ago.

>
> The only difference is that G has been worked very hard but B has mostly
> been resting.
>
> Do I assume a hardware problem? Is there any further testing I can do?
>
> Thanks ... Peter

I'd check "my computer" to see if the firmware version is being
displayed. You can do the OFW hardware tests.

Jerry


___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A network question

[Jerry Vonau]

On Fri, 2011-11-11 at 12:08 +1000, Peter Hewitt wrote:
> I'm finally starting to investigate networking on my XOs.
> 
> I have two XO 1s - Firmware Q2E45 Wireless 5.110.22.p23 Sugar .84.16 Build 852
> 
> Let's call them G (good) & B (bad)
> 
> The neighbourhood show three mesh symbols on both.
> 
> On G one pulses but on B ... no pulsing.
> 
> G finally says "Connected" on Mesh Network 1"
> 
> When I click "Connect" on Mesh Network 1 on B nothing changes.
> 
> ifconfig shows
> 
> eth0, lo & msh0 on G but just lo & msh0 on B
> 
> I'm sure they both connected fine when they first arrived about 6 months ago.
> 
> The only difference is that G has been worked very hard but B has mostly 
> been resting.
> 
> Do I assume a hardware problem? Is there any further testing I can do?
> 
> Thanks ... Peter

I'd check "my computer" to see if the firmware version is being
displayed. You can do the OFW hardware tests.  

Jerry



___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

[Sugar-devel] A network question

[Peter Hewitt]

I'm finally starting to investigate networking on my XOs.

I have two XO 1s - Firmware Q2E45 Wireless 5.110.22.p23 Sugar .84.16 Build 852

Let's call them G (good) & B (bad)

The neighbourhood show three mesh symbols on both.

On G one pulses but on B ... no pulsing.

G finally says "Connected" on Mesh Network 1"

When I click "Connect" on Mesh Network 1 on B nothing changes.

ifconfig shows

eth0, lo & msh0 on G but just lo & msh0 on B

I'm sure they both connected fine when they first arrived about 6 months ago.

The only difference is that G has been worked very hard but B has mostly 
been resting.


Do I assume a hardware problem? Is there any further testing I can do?

Thanks ... Peter

___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] TA cairo conversion

[C. Scott Ananian]

Current version on the cscott-gtk3 branch works on GTK3.  Some
functionality might be missing: I've just discovered a bug in the gir
bindings for gtk_pixmap_get_pixels() which makes 'read pixel' break in
gtk3, for instance.  GTK2 operation should have no regressions. (Testing
welcome!)
 --scott

On Thursday, November 10, 2011, C. Scott Ananian  wrote:
> On Thu, Nov 10, 2011 at 3:21 PM, Walter Bender 
wrote:
>> +1
>>
>> I'll do some testing; I plan to make the next release based on this
>> work. May as well get these bits into the field.
>
> I'm mostly just concerned that future patches base off the converted
> source and use the new class capitalization style so we won't continue
> to have patch conflicts.
>
> That said, I should mention that my testing has been entirely on
> desktop GTK2 and GTK3.  I'm not the expert on sugar's transition to
> GTK3, so there's probably work to do on the sugar integration code.
> Hopefully the same dual GTK2/GTK3 compatibility mechanism can be used
> (I view my main contribution as util/gtkcompat.py, which I hope can be
> extended).
>  --scott
>
> --
>   ( http://cscott.net )
>

-- 
  ( http://cscott.net )
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

[Sugar-devel] [ANNOUNCE] New Dextrose-3 development release: Alpha-3

[Anish Mangal]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Hi,

I'm pleased to announce another development release in the dextrose-3
series. This is our third alpha and builds upon sugar-0.94.1 and a new
dextrose patchset.

The targeted release date for the final stable release is 19th December.
Please see the dextrose-3 roadmap[1] for more details.

== Key points ==

* Based on sugar 0.94.1
* Incremental fixes included in the sugar-dx3-patchset [2]
* Includes initial release of sugar-client in place of
ds-backup-client and dextrose-updater and adds more functionality[3].

== Bug Fixes (Refer to sugar 0.94.1 release notes for upstream fixes) ==

* sl#2822 au#383 Only pre-select already installed activities
* sl#3217 Add en_AU in My Settings - Language
* Don't choke on non-integer activities

== Get it here ==

* XO-1, with gnome,
** http://download.sugarlabs.org/dextrose/testing/dx3/xo1/dx3g055.img
** http://download.sugarlabs.org/dextrose/testing/dx3/xo1/dx3g055.crc

* XO-1.5, with gnome,
** http://download.sugarlabs.org/dextrose/testing/dx3/xo1.5/dx3g055.zd

== Report problems ==

Though this is still a development release, it is pretty stable for a
alpha. Bug reports should be opened on bugs.sugarlabs.org and assigned
the dx3 keyword in order to enable us to track them effectively.

== Activity version updates ==

* Abacus-28
* Chat-73
* FlipSticks-10
* ImageViewer-19
* Ruler-14
* TamTamMini-61
* TurtleArt-122

== New Packages ==

* sugar-client-1.0-3.1.noarch

== Updated Packages ==

* bootfw-q3b22-1.unsigned.noarch
* freetype-2.4.2-6.fc14.i686
* ImageMagick-6.6.4.1-16.fc14.i686
* ImageMagick-c++-6.6.4.1-16.fc14.i686
* kernel-2.6.35.13_xo1.5-20111026.1209.olpc.5d76efb.i586
* krb5-libs-1.8.4-3.fc14.i686
* libdrm-2.4.22-2.fc14.i686
* NetworkManager-0.8.5.93-1.fc14.i686
* NetworkManager-glib-0.8.5.93-1.fc14.i686
* NetworkManager-gnome-0.8.5.93-1.fc14.i686
* nss-softokn-3.12.10-5.fc14.i686
* nss-softokn-freebl-3.12.10-5.fc14.i686
* olpc-powerd-37-1.fc14.i686
* olpc-powerd-dbus-37-1.fc14.i686
* olpc-utils-1.3.7-1.fc14.i686
* orc-0.4.16-3.fc14.i686
* parted-2.3-5.fc14.i686
* perl-5.12.4-148.fc14.i686
* perl-libs-5.12.4-148.fc14.i686
* perl-Module-Pluggable-3.90-148.fc14.noarch
* perl-Pod-Escapes-1.04-148.fc14.noarch
* perl-Pod-Simple-3.13-148.fc14.noarch
* perl-threads-shared-1.32-148.fc14.i686
* python-setuptools-0.6.24-1.fc14.noarch
* rpm-4.8.1-7.fc14.i686
* rpm-libs-4.8.1-7.fc14.i686
* rpm-python-4.8.1-7.fc14.i686
* rsync-3.0.8-2.fc14.i686
* sugar-0.94.1-20111025.dx3.fc14.noarch
* xorg-x11-server-common-1.9.5-2.fc14.i686
* xorg-x11-server-Xorg-1.9.5-2.fc14.i686
* xorg-x11-server-Xvfb-1.9.5-2.fc14.i686

== Removed packages ==

* cronie-1.4.8-2.fc14.i686
* cronie-anacron-1.4.8-2.fc14.i686
* crontabs-1.11-1.20101115git.fc14.noarch
* ds-backup-client-0.11.1.g71d2f16-1.olpc3.noarch
* rsyslog-4.6.3-3.fc14.i686

[1] http://wiki.sugarlabs.org/go/Dextrose/3/Roadmap
[2]
http://people.sugarlabs.org/~silbe/dextrose/patchsets/sugar-dx3-20111025.tar.gz
[3] http://wiki.sugarlabs.org/go/Sugar_Server_Kit/sugar-client

- -- 
Anish Mangal
Dextrose Project Manager
Activity Central
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iQEcBAEBAgAGBQJOvEcxAAoJEBoxUdDHDZVpxcgH/0eUUmSoXF9wRoEluqgucvpE
+euC/TCW5VtLseK0jJxp27tqA1fxYr4oexIRkIi1JPjZ/oGiG51okvZd0LXH2hhX
8xEtdEoiThQrvY3oVO6tdxAW0mR8ieXo3G1TBt6rYvvSWzMu7p6hta/Y9gfaUv7J
mEPTbJZm9+cDfum+YUlf+mFolSb9QvUNUompymCW6z0otPmJvsswgTIbvHr8tYkT
NlMAV0TaNmWTgL+A8dE6dP9zq98/mg209wn42B4OO3a1gPP5mXhaq/ib31yQT6So
RJzXY6OuLR5BImg2JkYadf5MeBaZrRsQ370k2AlhQV3VV+2/4z49S2M3QoOc3JU=
=Qxlk
-END PGP SIGNATURE-
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [SoaS] SoaS v7?

[Walter Bender]

On Wed, Nov 9, 2011 at 7:18 PM, Peter Robinson  wrote:
> Hi All,
>
> So there's no rest for the wicked its time to think about what you
> would like to develop for SoaS v7. I know upstream is all very busy
> already the sugar team are full pelt into the conversion of sugar
> to gtk3 and PyGI and there's all sorts of fun stuff going into
> Fedora... so what do you want?

Is there still the issue with instability of persistent data? Maybe we
can look at the one for v7?

regards.

-walter
>
> Peter
> ___
> SoaS mailing list
> s...@lists.sugarlabs.org
> http://lists.sugarlabs.org/listinfo/soas
>



-- 
Walter Bender
Sugar Labs
http://www.sugarlabs.org
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] TA cairo conversion

[C. Scott Ananian]

On Thu, Nov 10, 2011 at 3:21 PM, Walter Bender  wrote:
> +1
>
> I'll do some testing; I plan to make the next release based on this
> work. May as well get these bits into the field.

I'm mostly just concerned that future patches base off the converted
source and use the new class capitalization style so we won't continue
to have patch conflicts.

That said, I should mention that my testing has been entirely on
desktop GTK2 and GTK3.  I'm not the expert on sugar's transition to
GTK3, so there's probably work to do on the sugar integration code.
Hopefully the same dual GTK2/GTK3 compatibility mechanism can be used
(I view my main contribution as util/gtkcompat.py, which I hope can be
extended).
  --scott

-- 
      ( http://cscott.net )
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] TA cairo conversion

[Walter Bender]

On Thu, Nov 10, 2011 at 3:17 PM, C. Scott Ananian  wrote:
> I'm rebased on top of mainline now, staged at the usual place.  I'll
> continue working on getting the GTK3 functional while leaving GTK2
> operation unchanged.

+1

I'll do some testing; I plan to make the next release based on this
work. May as well get these bits into the field.

-walter

>  --scott
>
> --
>       ( http://cscott.net )
>



-- 
Walter Bender
Sugar Labs
http://www.sugarlabs.org
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] TA cairo conversion

[C. Scott Ananian]

I'm rebased on top of mainline now, staged at the usual place.  I'll
continue working on getting the GTK3 functional while leaving GTK2
operation unchanged.
 --scott

-- 
      ( http://cscott.net )
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [SoaS] SoaS v7?

[Art Hunkins]

Yes, SoaS v6 is indeed wonderful. It seems speedier to load the OS, load 
Activities, and shut down than previous versions.

Kudos to Peter Robinson especially.

Art Hunkins
  - Original Message - 
  From: Thomas C Gilliard 
  To: s...@lists.sugarlabs.org 
  Cc: IAEP SugarLabs ; Sugar Devel 
  Sent: Thursday, November 10, 2011 1:17 PM
  Subject: Re: [Sugar-devel] [SoaS] SoaS v7?


  This is an example of a dd writable SoaS-v2.img file which produces a fully 
functional Conventional install on a USB. This is equivalent to a liveinst 
install to USB.[1] 
It was first suggested by bernie [2] in early 2010 as a better alternative 
to the currently used live USB SoaS filesystems with their fragile and limited 
sized overlay, 
a.. This functionality is expected to be included in fedora 17 liveinst 
(Anaconda) 
b.. It would be even nicer to be able to create and install to USB, the 
Soas-v7.img file directly from the first sugar boot screen as a third option 
line, rather from a obscure liveinst command in terminal.


  [1] 
http://wiki.sugarlabs.org/go/Sugar_Creation_Kit#A-Non-Live_.28real_file_structured.29_USB
 (a working 4GB dd Soas-v2.img file and description. As proof of concept)
  [2] http://people.sugarlabs.org/Tgillard/Direct_Soas.txt   (Original 
Announcement)

  SoaS-v6 is great...thanks

  Tom Gilliard
  satellit_

  On 11/09/2011 04:18 PM, Peter Robinson wrote: 
Hi All,

So there's no rest for the wicked its time to think about what you
would like to develop for SoaS v7. I know upstream is all very busy
already the sugar team are full pelt into the conversion of sugar
to gtk3 and PyGI and there's all sorts of fun stuff going into
Fedora... so what do you want?

Peter
___
SoaS mailing list
s...@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/soas



--


  ___
  Sugar-devel mailing list
  Sugar-devel@lists.sugarlabs.org
  http://lists.sugarlabs.org/listinfo/sugar-devel
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A "dismiss all" functionality for notifications?

[Christoph Derndorfer]

Am 10.11.2011 17:33 schrieb "Gary Martin" :
>
> Hi Christoph,
>
>
> On 9 Nov 2011, at 23:25, Christoph Derndorfer <
christoph.derndor...@gmail.com> wrote:
>
>> Hi all,
>>
>> while playing around with 11.3.0 today I spent some time sending
Memorize files back and forth between two XOs. This quickly led to the
frame being filled with icons of these transfers so now I'm wondering
whether it might be useful to consider adding a "dismiss all" functionality
somewhere?
>
>
> Interesting you should raise this. While at Paris Sugar Camp, Simon and I
worked on some UI sketches to present all transfer notifications in a
single frame icon and palette so that they can be easily managed all in one
place.

Great, this sounds pretty much like what I had in mind.

The target case was that of a teacher trying to transfer material to
multiple children in a class (the groups feature also needs improvement to
allow single shot transfer to multiple destinations).

Nice. In terms of enabling a very basic level of collaboration, or rather
easy artifact sharing, such a functionality is certainly one of the keys
pieces which is currently missing.

> I'll try to tidy up and get the sketches on the wiki for review.

Looking forward to the sketches!

Thanks,
Christoph

> --Gary
>
>> Cheers,
>> Christoph
>>
>> --
>> Christoph Derndorfer
>>
>> editor, OLPC News [www.olpcnews.com]
>> volunteer, OLPC (Austria) [www.olpc.at]
>>
>> e-mail: christ...@derndorfer.eu
>>
>>
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [SoaS] SoaS v7?

[Thomas C Gilliard]

This is an example of a dd writable SoaS-v2.img file which produces a 
fully functional Conventional install on a USB. This is equivalent to a 
liveinst install to USB.[1]


   It was first suggested by bernie [2] in early 2010 as a better
   alternative to the currently used live USB SoaS filesystems with
   their fragile and limited sized overlay, 


   * This functionality is expected to be included in fedora 17
 liveinst (Anaconda)
   * It would be even nicer to be able to create and install to USB,
 the Soas-v7.img file directly from the first sugar boot screen as
 a third option line, rather from a obscure liveinst command in
 terminal.


[1] 
http://wiki.sugarlabs.org/go/Sugar_Creation_Kit#A-Non-Live_.28real_file_structured.29_USB 
(a working 4GB dd Soas-v2.img file and description. As proof of concept)
[2] http://people.sugarlabs.org/Tgillard/Direct_Soas.txt   (Original 
Announcement)


SoaS-v6 is great...thanks

Tom Gilliard
satellit_

On 11/09/2011 04:18 PM, Peter Robinson wrote:

Hi All,

So there's no rest for the wicked its time to think about what you
would like to develop for SoaS v7. I know upstream is all very busy
already the sugar team are full pelt into the conversion of sugar
to gtk3 and PyGI and there's all sorts of fun stuff going into
Fedora... so what do you want?

Peter
___
SoaS mailing list
s...@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/soas

___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [SPAM?]: Re: [PATCH] Start filtered activity with keyboard

[Simon Schampijer]

On 11/10/2011 06:16 PM, Gary Martin wrote:

Hi Simon,

On 10 Nov 2011, at 08:09, Simon Schampijer  wrote:


On 11/07/2011 01:06 PM, godi...@sugarlabs.org wrote:

From: Gonzalo Odiard

This patch complement the filter functionality in the home view.
If only one activity is selected, Return key start the activity.

Signed-off-by: Gonzalo Odiard


Thanks for the patch.

Design question: your approach is that the 'start with return' functionality is 
only available if there is a clear match. The other option would be to choose 
one of the matches to be the default one (this is the behavior that the GNOME 
shell is using, go to hot corner and type 'fir' for example to see it). They 
highlight the default match, we could do similar with drawing the corner around 
the icon (like when you hover over it). Might be a nice idea to do this in 
general, actually.


Yes that seems like in might be a nice extension to Gonzalo's patch. How would 
you take the decision on picking the default match, by most recent use, most 
frequent use, recent install?


I did not think about it yet. I thought about looking at what the GNOME 
shell is doing. Most recent use, might be a good fit though.



When you say 'nice idea to do this in general' were you thinking of the Journal 
search, and activity list view?


Ahh, I meant, that showing the rectangle (highlight) the icon is a good 
way of feedback. Just for the home view.


Regards,
   Simon

___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [PATCH] Start filtered activity with keyboard

[Gonzalo Odiard]

Thanks for the patch.

>
> Design question: your approach is that the 'start with return'
> functionality is only available if there is a clear match. The other option
> would be to choose one of the matches to be the default one (this is the
> behavior that the GNOME shell is using, go to hot corner and type 'fir' for
> example to see it). They highlight the default match, we could do similar
> with drawing the corner around the icon (like when you hover over it).
> Might be a nice idea to do this in general, actually.
>
> Regards,
>   Simon
>
>
Yes. I did it in this way, because is the simplest from a point of view of
the implementation.
I also think, would be good display in any way to the user, the fact he is
enabled to start the activity,
maybe putting the complete name in the filter, or highlighting the icon.
But is not obvious what is the best, we already have the highlight with the
mouse onhover,
and completing the name in the filter may be is not ok.

If we implement what you say, we need highlight the default activity, and
implement keyboard navigation
to the non default options(up/down in gnome-shell). Actually, we do not
have keyboard navigation implemented in the home,
may be will be easy to do when we remove hippo.

Then, I think we should go with the simplest solution now, and improve
later.
This implementation is similar in use no to gnome-shell, but to
tab-completion in bash.

Gonzalo



> __**_
> Sugar-devel mailing list
> Sugar-devel@lists.sugarlabs.**org 
> http://lists.sugarlabs.org/**listinfo/sugar-devel
>
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [PATCH] Start filtered activity with keyboard

[Frederick Grose]

On Thu, Nov 10, 2011 at 12:16 PM, Gary Martin wrote:

> Hi Simon,
>
> On 10 Nov 2011, at 08:09, Simon Schampijer  wrote:
>
> > On 11/07/2011 01:06 PM, godi...@sugarlabs.org wrote:
> >> From: Gonzalo Odiard
> >>
> >> This patch complement the filter functionality in the home view.
> >> If only one activity is selected, Return key start the activity.
> >>
> >> Signed-off-by: Gonzalo Odiard
> >
> > Thanks for the patch.
> >
> > Design question: your approach is that the 'start with return'
> functionality is only available if there is a clear match. The other option
> would be to choose one of the matches to be the default one (this is the
> behavior that the GNOME shell is using, go to hot corner and type 'fir' for
> example to see it). They highlight the default match, we could do similar
> with drawing the corner around the icon (like when you hover over it).
> Might be a nice idea to do this in general, actually.
>
> Yes that seems like in might be a nice extension to Gonzalo's patch. How
> would you take the decision on picking the default match, by most recent
> use, most frequent use, recent install?
>

First in the currently selected sort would be a good default.

   --Fred

>
> When you say 'nice idea to do this in general' were you thinking of the
> Journal search, and activity list view?
>
> --Gary
>
> > Regards,
> >   Simon
> ___
> Sugar-devel mailing list
> Sugar-devel@lists.sugarlabs.org
> http://lists.sugarlabs.org/listinfo/sugar-devel
>
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [PATCH] Start filtered activity with keyboard

[Gary Martin]

Hi Simon,

On 10 Nov 2011, at 08:09, Simon Schampijer  wrote:

> On 11/07/2011 01:06 PM, godi...@sugarlabs.org wrote:
>> From: Gonzalo Odiard
>> 
>> This patch complement the filter functionality in the home view.
>> If only one activity is selected, Return key start the activity.
>> 
>> Signed-off-by: Gonzalo Odiard
> 
> Thanks for the patch.
> 
> Design question: your approach is that the 'start with return' functionality 
> is only available if there is a clear match. The other option would be to 
> choose one of the matches to be the default one (this is the behavior that 
> the GNOME shell is using, go to hot corner and type 'fir' for example to see 
> it). They highlight the default match, we could do similar with drawing the 
> corner around the icon (like when you hover over it). Might be a nice idea to 
> do this in general, actually.

Yes that seems like in might be a nice extension to Gonzalo's patch. How would 
you take the decision on picking the default match, by most recent use, most 
frequent use, recent install? 

When you say 'nice idea to do this in general' were you thinking of the Journal 
search, and activity list view?

--Gary

> Regards,
>   Simon
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] A "dismiss all" functionality for notifications?

[Gary Martin]

Hi Christoph,

On 9 Nov 2011, at 23:25, Christoph Derndorfer  
wrote:

> Hi all,
> 
> while playing around with 11.3.0 today I spent some time sending Memorize 
> files back and forth between two XOs. This quickly led to the frame being 
> filled with icons of these transfers so now I'm wondering whether it might be 
> useful to consider adding a "dismiss all" functionality somewhere?

Interesting you should raise this. While at Paris Sugar Camp, Simon and I 
worked on some UI sketches to present all transfer notifications in a single 
frame icon and palette so that they can be easily managed all in one place. The 
target case was that of a teacher trying to transfer material to multiple 
children in a class (the groups feature also needs improvement to allow single 
shot transfer to multiple destinations).

I'll try to tidy up and get the sketches on the wiki for review.

--Gary

> Cheers,
> Christoph
> 
> -- 
> Christoph Derndorfer
> 
> editor, OLPC News [www.olpcnews.com]
> volunteer, OLPC (Austria) [www.olpc.at]
> 
> e-mail: christ...@derndorfer.eu
> 
> 
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [DESIGN] Toolbars review

[Gonzalo Odiard]

Thanks Gary and Manuel!
Great Team :)

Gonzalo
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

[Sugar-devel] [PATCH] Bundlebuilder: don't fail to package if git is not installed, OLPC #11341

[from = Simon Schampijer]

From: Simon Schampijer 

When git is installed, it will return non-zero if it gets asked to list
the files in a non-git-repository. The subprocess.Popen instantiation is
successful in this case and the returncode attribute will contain the
error code from git. The current code handles this fine and does fall
back to our own source file listing facility.

If git isn't installed, however, trying to instantiate subprocess.Popen
will fail with OSError. We need to catch this and fall back to our own
source file listing facility like we do for the non-repository case.

Signed-off-by: Simon Schampijer 
---
 src/sugar/activity/bundlebuilder.py |   28 ++--
 1 files changed, 22 insertions(+), 6 deletions(-)

diff --git a/src/sugar/activity/bundlebuilder.py 
b/src/sugar/activity/bundlebuilder.py
index e7641bb..7405fcf 100644
--- a/src/sugar/activity/bundlebuilder.py
+++ b/src/sugar/activity/bundlebuilder.py
@@ -184,11 +184,19 @@ class XOPackager(Packager):
 bundle_zip.close()
 
 def _get_files_in_git(self):
-git_ls = subprocess.Popen(['git', 'ls-files'], stdout=subprocess.PIPE,
-  cwd=self.config.source_dir)
+try:
+git_ls = subprocess.Popen(['git', 'ls-files'],
+  stdout=subprocess.PIPE,
+  cwd=self.config.source_dir)
+except OSError:
+logging.warn('XOPackager, git is not installed, ' \
+ 'fall back to filtered list')
+return list_files(self.config.source_dir,
+  IGNORE_DIRS, IGNORE_FILES)
 stdout, _ = git_ls.communicate()
 if git_ls.returncode:
-# Fall back to filtered list
+logging.warn('XOPackager, this is not a git repository, ' \
+ 'fall back to filtered list')
 return list_files(self.config.source_dir,
   IGNORE_DIRS, IGNORE_FILES)
 
@@ -204,11 +212,19 @@ class SourcePackager(Packager):
  self.config.tar_name)
 
 def get_files(self):
-git_ls = subprocess.Popen(['git', 'ls-files'], stdout=subprocess.PIPE,
-  cwd=self.config.source_dir)
+try:
+git_ls = subprocess.Popen(['git', 'ls-files'],
+  stdout=subprocess.PIPE,
+  cwd=self.config.source_dir)
+except OSError:
+logging.warn('SourcePackager, git is not installed, ' \
+ 'fall back to filtered list')
+return list_files(self.config.source_dir,
+  IGNORE_DIRS, IGNORE_FILES)
 stdout, _ = git_ls.communicate()
 if git_ls.returncode:
-# Fall back to filtered list
+logging.warn('SourcePackager, this is not a git repository, ' \
+ 'fall back to filtered list')
 return list_files(self.config.source_dir,
   IGNORE_DIRS, IGNORE_FILES)
 
-- 
1.7.7

___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [PATCH sugar-toolkit] Bundlebuilder: don't fail to package if git is not installed, OLPC #11341

[Simon Schampijer]

On 10/21/2011 01:40 PM, Gonzalo Odiard wrote:

I think we should fallback to use a manifest file if present.
In particluar, the wikipedia activity can be builded with different data
files,
but the same codebase, and is using a manifest file to select the included
files.
Anyway, is he only case. May be we can implement it in a custom setup.py in
the activity,
but imho would be better if the toolkit resolve it.

Gonzalo


Well, we just removed all of the MANIFEST support :/ Are those files 
static or dynamically created? Maybe you can work with git branches (you 
would have to cherry-pick patches though)?


Regards,
   Simon
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [PATCH sugar-toolkit] Bundlebuilder: don't fail to package if git is not installed, OLPC #11341

[Simon Schampijer]

On 10/24/2011 02:43 PM, Sascha Silbe wrote:

Excerpts from Simon Schampijer's message of 2011-10-21 12:27:02 +0200:


We need to make sure that we do not fail if git is not
installed and use the same fallback code as if we are
in a directory where no git information is present.


How about:

"""
When git is installed, it will return non-zero if it gets asked to list
the files in a non-repository. The subprocess.Popen instantiation is
successful in this case and the returncode attribute will contain the
error code from git. The current code handles this fine.

If git isn't installed, however, trying to instantiate subprocess.Popen
will fail with OSError. We need to catch this and fall back to our own
source file listing facility like we do for the non-repository case.
"""


Sounds good to me.


[sugar.activity.bundlebuilder.XOPackager]

  def _get_files_in_git(self):
-git_ls = subprocess.Popen(['git', 'ls-files'], stdout=subprocess.PIPE,
-  cwd=self.config.source_dir)
+try:
+git_ls = subprocess.Popen(['git', 'ls-files'],
+  stdout=subprocess.PIPE,
+  cwd=self.config.source_dir)
+except OSError:
+# Git not installed. Fall back to filtered list
+return list_files(self.config.source_dir,
+  IGNORE_DIRS, IGNORE_FILES)


Since our own listing code can generate different results than git
ls-files does, we should log a warning (same applies to
SourcePackager.get_files()).

Sascha


Yes, I can add the warning.

Regards,
   Simon
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Re: [Sugar-devel] [PATCH] Start filtered activity with keyboard

[Simon Schampijer]

On 11/07/2011 01:06 PM, godi...@sugarlabs.org wrote:

From: Gonzalo Odiard

This patch complement the filter functionality in the home view.
If only one activity is selected, Return key start the activity.

Signed-off-by: Gonzalo Odiard


Thanks for the patch.

Design question: your approach is that the 'start with return' 
functionality is only available if there is a clear match. The other 
option would be to choose one of the matches to be the default one (this 
is the behavior that the GNOME shell is using, go to hot corner and type 
'fir' for example to see it). They highlight the default match, we could 
do similar with drawing the corner around the icon (like when you hover 
over it). Might be a nice idea to do this in general, actually.


Regards,
   Simon
___
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel