John Watlington | 1 Sep 01:20 2011

Re: Light sensor on XO-1.75


On Aug 29, 2011, at 7:52 PM, James Cameron wrote:

> On Mon, Aug 29, 2011 at 07:38:20PM -0400, Richard Smith wrote:
>> On Mon, Aug 29, 2011 at 1:06 PM, Bert Freudenberg <bert <at> freudenbergs.de> wrote:
>>> And it seems to sit "behind" the blinking left-most LED? And the
>>> blink rate is proportional to the light sensed? Why?
>> 
>> Its proportional because that's how  the sensor works.  The amount of
>> light changes the time to bleed off the charge stored in the diode
>> when you reverse bias it.
> 
> Hmm, we could schedule the samples to start at a constant time relative
> to the previous start, and thereby only change the amount of off time of
> the wireless LED, but this would suck because the sample rate would be
> much slower in bright light.  We would not notice encroaching darkess as
> quickly.

I am moving the LED and blanking circuitry to the storage LED tonight for
the C-build and beyond.   I tried a yellow LED for sensing, and it was less
sensitive to interference from the WiFi LED than the green LED, but not
enough to matter.   The diffusing nature of the lightpipe ensures lots of
back reflection.

Cheers,
wad
John Watlington | 1 Sep 01:21 2011

Graphics for indicating that a firmware update is taking place


Any suggestions for graphics to include in the boot screen when
the firmware is being automatically updated ?   Otherwise it looks
like a hang and a user might try pulling power to reboot...
http://dev.laptop.org/ticket/11205

Cheers,
wad
manuel quiñones | 1 Sep 01:46 2011

Re: Lost "activity failed to launch" smarts in Sugar launch splash

2011/8/31 Gonzalo Odiard <gonzalo <at> laptop.org>:
>
>
> On Tue, Aug 30, 2011 at 9:06 PM, Martin Langhoff <martin <at> laptop.org> wrote:
>>
>> Hi Gonzalo,
>>
>> on os41, if you launch Maze, it fails to start (that's logged
>> already)... and the launcher fails to register the failure, which I
>> don't think has been discussed.
>>
>> shell.log says "ERROR root: Cannot send NotifyLaunchFailure to the
>> shell" -- could this be related to the new activity startup splash? Is
>> this otherwise known?
>>
>
> No, is not related.
> Maze is really broken, but one of the problems (already solved in git)
> is the configuration in the activity.info file and sugar can start the
> activity.
> (the line exec should be "exec = sugar-activity activity.MazeActivity)
> If you want try, must restart sugar after editing the file because are
> cached.

The fix is very simple, a change on a string (the string needs to
match a module that contains the game loop in a function called main).
 I wonder why someone would change the code and commit before even
checking the activity starts!  I will send the fix to sugarlabs ml.

diff --git a/activity.py b/activity.py
(Continue reading)

James Cameron | 1 Sep 04:04 2011

Re: Graphics for indicating that a firmware update is taking place

Fixed it yesterday, please give it a try.
http://dev.laptop.org/ticket/11117

The graphics were designed with the following goals:

0.  not do any graphics operations during the actual flashing, e.g. no
progress bar, because this makes it harder to test any interactions,

1.  not add another icon to the rom: list, just re-use what is there,

2.  lead the user's eye toward the storage LED, which will glow during
the flashing,

3.  suggest that the laptop itself is doing something,

4.  use an ellipsis to indicate a temporary "falling short", or
"omission" of progress,

5.  be interesting enough to cause the user to focus and wonder, by
which time it will all be over.  ;-)  The EC reflash time is tiny.

--

-- 
James Cameron
http://quozl.linux.org.au/
Martin Langhoff | 1 Sep 20:17 2011
Picon

Re: Graphics for indicating that a firmware update is taking place

On Wed, Aug 31, 2011 at 10:04 PM, James Cameron <quozl <at> laptop.org> wrote:
> Fixed it yesterday, please give it a try.
> http://dev.laptop.org/ticket/11117

Interesting.

My preference on this -- following earlier discussions -- would be to
be able to bundle graphics in the bootfw.zip bundle, so that
deployments can provide localized graphics (even in small / limited
format).

This is of course a harder trick when updating the EC. Is there any
practical way of making this possible?

m
--

-- 
 martin.langhoff <at> gmail.com
 martin <at> laptop.org -- Software Architect - OLPC
 - ask interesting questions
 - don't get distracted with shiny stuff  - working code first
 - http://wiki.laptop.org/go/User:Martinlanghoff
James Cameron | 2 Sep 02:06 2011

Re: Graphics for indicating that a firmware update is taking place

On Thu, Sep 01, 2011 at 02:17:12PM -0400, Martin Langhoff wrote:
> On Wed, Aug 31, 2011 at 10:04 PM, James Cameron <quozl <at> laptop.org> wrote:
> > Fixed it yesterday, please give it a try.
> > http://dev.laptop.org/ticket/11117
> 
> Interesting.
> 
> My preference on this -- following earlier discussions -- would be to
> be able to bundle graphics in the bootfw.zip bundle, so that
> deployments can provide localized graphics (even in small / limited
> format).
> 
> This is of course a harder trick when updating the EC. Is there any
> practical way of making this possible?

Yes, but it makes something that is already risky even more dependent on
other factors, such as the access time of the eMMC or SD.  I would not
be willing to go there.  You might ask Mitch.

We probably have room for one global icon for this.  I shall await a
design.

I think time would be better spent on ensuring that the power button is
ignored during OpenFirmware reflash.  It blinks if it is pressed.

It doesn't blink if pressed during EC reflash.

--

-- 
James Cameron
http://quozl.linux.org.au/
(Continue reading)

Kevin Gordon | 2 Sep 07:21 2011
Picon

Wireless Security WPA2

Folks:

I have a site where I sometimes work on my little XO's that has now implemented WPA2 Enterprise wireless security.  On the Gnome side I can edit a Wireless Network Connection to correctly enter the parameters and get onto the network.  However, I cannot seem to get a connection on the Sugar side.(Build11-2/874).  Is there some way I can get Sugar to use the configuration from the Gnome side, or is there a terminal command I can execute in Sugar there that would allow me to make a connection to a WPA2 Enterprise infrastructure?  Even with WPA2 Personal with pre-shared keys, it doesn't seem to be sharing the connection information between Gnome and Sugar, but it in the end, WPA2 Personal it does work just fine on both sides.

Thanks,

KG

_______________________________________________
Devel mailing list
Devel <at> lists.laptop.org
http://lists.laptop.org/listinfo/devel
Jerry Vonau | 2 Sep 07:39 2011
Picon

Re: Wireless Security WPA2

On Fri, 2011-09-02 at 01:21 -0400, Kevin Gordon wrote:
> Folks:
> 
> I have a site where I sometimes work on my little XO's that has now
> implemented WPA2 Enterprise wireless security.  On the Gnome side I
> can edit a Wireless Network Connection to correctly enter the
> parameters and get onto the network.  However, I cannot seem to get a
> connection on the Sugar side.(Build11-2/874).  Is there some way I can
> get Sugar to use the configuration from the Gnome side, or is there a

When in the editor tick the box "Available to all users", sets up a
system connection with the config saved in 
/etc/sysconfig/network-scripts/ifcfg-<conn name> 

>  terminal command I can execute in Sugar there that would allow me to
>  make a connection to a WPA2 Enterprise infrastructure?  Even with
> WPA2 Personal with pre-shared keys, it doesn't seem to be sharing the
> connection information between Gnome and Sugar, but it in the end,
> WPA2 Personal it does work just fine on both sides.
> 
> Thanks,

Should auto-connect on boot too.

Jerry
Simon Schampijer | 2 Sep 10:09 2011
Picon

Gtk3 Hackfest, Praha 2011, news updates

Hi,

two major changes have recently occurred in Sugar's underlying 
technologies. Firstly, GTK+ 2 has been obsoleted by GTK+ 3, and GNOME is 
now based on GTK+ 3. Secondly, PyGTK, the underlying Python library that 
Sugar uses to call into GTK+, has been deprecated in favor of PyGObject 
Introspection (hereafter "PyGI"). More background info can be found at 
Features/GTK3 [1]. Goal of this hackfest is to remove the biggest 
blockers before we can start the porting and potentially start porting 
over.

== Sponsoring regional trips ==
This Hackfest is sponsored by the OLPC Foundation. If you need help for 
funding your travel, please send an email to simon AT laptop DOT org 
before the 8th of October 2011. OLPCF won't be able to refund more than 
150 USD per trip, and we will fund regional (european) travels in 
priority [2].

Regards,
    Simon

[1] http://wiki.sugarlabs.org/go/Features/GTK3
[2] http://wiki.sugarlabs.org/go/Marketing_Team/Events/Gtk3_Hackfest_2011
[3] http://brmlab.cz/|
Sascha Silbe | 2 Sep 14:52 2011

Re: Light sensor on XO-1.75

Excerpts from Paul Fox's message of Tue Aug 30 16:38:13 +0200 2011:

> yes, it's a nice trick.  i think it was discussed on this list last year.
> it's not very sensitive, but it can tell us the difference between full
> sun and not, which is really when we'd want to turn off the backlight.

Given that turning off the backlight usually isn't time critical, how
about triggering a measurement cycle only when user space actually reads
from the sysfs file? That would disturb regular operation of the nearby
LED only occasionally.

Sascha

--

-- 
http://sascha.silbe.org/
http://www.infra-silbe.de/
_______________________________________________
Devel mailing list
Devel <at> lists.laptop.org
http://lists.laptop.org/listinfo/devel

Gmane