Official Feedback for Hexoquinasa v0.9

1 view
Skip to first unread message

Sebastian Silva

unread,
May 13, 2013, 2:56:11 PM5/13/13
to sugar-...@googlegroups.com, todos
Hello friends and hackers of Sugar Network,

This is to share with you official feedback by the Ministry of Education
of Peru with
regard to Hexoquinasa, a downstream distribution of the Sugar Labs
Platform Team
reference deployment platform with access to the Sugar Network.

The plan as coordinated with DIGETE is to release a new version by Fri
24th of May.
The new version will get testing in the field by a select group of DATs
(Docentes de
Acompa�amiento Tecnol�gico).

It is worth noting:
- Most feedback is about activities
- VNC Launcher activity fails to find system x11vncserver binaries.
- Some feedback is vague and might require further inquiry /
testing ("Turtle Art crashes").
- Some feature requests outside our field of action (i.e. sharing
in paint and labyrinth)
- Some activities fail on close (Physics ends with signal 6).

Additionally to the items listed in the report:
- Mesh (or ad hoc) icons are not present in neighbourhood view.
- Test sharing between XO1 and XO1.5
- Sugar Network cache policy was discussed. They would like "to run
activities from server"
to avoid filling user disk.

Warm regards,
Sebastian
117-Informe Lalo (10-05-13).docx

Walter Bender

unread,
May 13, 2013, 4:48:41 PM5/13/13
to sugar-...@googlegroups.com, todos
On Mon, May 13, 2013 at 2:56 PM, Sebastian Silva <seba...@fuentelibre.org> wrote:
Hello friends and hackers of Sugar Network,

This is to share with you official feedback by the Ministry of Education of Peru with
regard to Hexoquinasa, a downstream distribution of the Sugar Labs Platform Team
reference deployment platform with access to the Sugar Network.

The plan as coordinated with DIGETE is to release a new version by Fri 24th of May.
The new version will get testing in the field by a select group of DATs (Docentes de
Acompañamiento Tecnológico).


It is worth noting:
    - Most feedback is about activities
    - VNC Launcher activity fails to find system x11vncserver binaries.
    - Some feedback is vague and might require further inquiry / testing ("Turtle Art crashes").
    - Some feature requests outside our field of action (i.e. sharing in paint and labyrinth)
    - Some activities fail on close (Physics ends with signal 6).

Additionally to the items listed in the report:
    - Mesh (or ad hoc) icons are not present in neighbourhood view.
    - Test sharing between XO1 and XO1.5
    - Sugar Network cache policy was discussed. They would like "to run activities from server"
      to avoid filling user disk.

Warm regards,
Sebastian


--
You received this message because you are subscribed to the Google Groups "Sugar Network" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sugar-network+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.



Sebastian,

The feedback is helpful. Thanks.

From looking at the document, I see it is Sugar 0.94 they are testing. Is that correct? Do you know what version so the activities they are testing? Any way to get access to some log files for the activities that did not load, e.g., Turtle?

thanks.

--
Walter Bender
Sugar Labs
http://www.sugarlabs.org

Sebastian Silva

unread,
May 13, 2013, 11:53:27 PM5/13/13
to sugar-...@googlegroups.com, Walter Bender, todos
El 13/05/13 15:48, Walter Bender escribió:
Sebastian,

The feedback is helpful. Thanks.

From looking at the document, I see it is Sugar 0.94 they are testing. Is that correct? Do you know what version so the activities they are testing? Any way to get access to some log files for the activities that did not load, e.g., Turtle?

Hi Walter,
Indeed we're basing in Sugar v0.94. Here is a link to the technical notes:
http://pe.sugarlabs.org/go/Hexoquinasa/v0.9

I've added activity versions to that page.

With regard to accessing logs, as part of the Sugar Network shell plugin, when an activity fails, the user is presented with a "report failure" dialog. If he so chooses he may then send the report along with the logs. However we haven't received failure reports from Turtle Art [1]. Here's how that list looks with reports, from the Physics activity [2].

[1] http://network.sugarlabs.org/hub/?activities&org.laptop.TurtleArtActivity&reports
[2] http://network.sugarlabs.org/hub/?activities&org.laptop.physics&reports

This is from alsroot's Contributor's Hub.

It is our intention that the Sugar Network project will help users with support and developers with direct feedback from the field.

Regards,
Sebastian

Aleksey Lim

unread,
May 14, 2013, 2:45:01 AM5/14/13
to somos...@lists.sugarlabs.org, sugar-...@googlegroups.com, todos, Walter Bender
On Mon, May 13, 2013 at 10:53:27PM -0500, Sebastian Silva wrote:
> El 13/05/13 15:48, Walter Bender escribi�:
> >Sebastian,
> >
> >The feedback is helpful. Thanks.
> >
> >From looking at the document, I see it is Sugar 0.94 they are
> >testing. Is that correct? Do you know what version so the
> >activities they are testing? Any way to get access to some log
> >files for the activities that did not load, e.g., Turtle?
> >
>
There is an issue, activity fail reports are being visible only
for authenticated users (which is possible only from local
Sugar Network client). Will fix this issue to make reports visible from
the Web as well.

Also, will make it possible (for Web anonymous users) to browse
activity versions that are being used in Hexokinase (Sugar Network
supports layers for its content, it is being used here as a QA
instrument to provide for deployments only particular activities
versions).

--
Aleksey

Walter Bender

unread,
May 14, 2013, 8:22:46 AM5/14/13
to Aleksey Lim, Lista de correo del equipo Somos-Azúcar, sugar-...@googlegroups.com, todos
On Tue, May 14, 2013 at 2:45 AM, Aleksey Lim <als...@sugarlabs.org> wrote:
On Mon, May 13, 2013 at 10:53:27PM -0500, Sebastian Silva wrote:
> El 13/05/13 15:48, Walter Bender escribió:

Access to the log files will be very useful for remote debugging. In the meantime, I'll try to get my hands on a machine running 0.94. Looking at my change logs for Turtle Art, I can hazard a guess that one problem night be a missing dependency, which I removed in a later version. In any case, I recommend that you try using the latest version if possible -- many bug fixes since v160 -- but at least v169, since I fixed a memory leak in that version that would have impacted performance on XO 1.

regards.

-walter

Walter Bender

unread,
May 14, 2013, 2:16:40 PM5/14/13
to Aleksey Lim, Lista de correo del equipo Somos-Azúcar, sugar-...@googlegroups.com, todos
With help from Reuben, I got a machine (1.75) running the 0.94.1 image for Amazonas.
It came with TA v138. Loaded v160 and it launched fine. Loaded v176 and it launched fine.

Not sure what else I can do to debug at this point.

regards.

-walter

Aleksey Lim

unread,
May 16, 2013, 12:38:54 AM5/16/13
to sugar-...@googlegroups.com, Lista de correo del equipo Somos-Azúcar, todos
On Tue, May 14, 2013 at 08:22:46AM -0400, Walter Bender wrote:
> On Tue, May 14, 2013 at 2:45 AM, Aleksey Lim <als...@sugarlabs.org> wrote:
>
> > On Mon, May 13, 2013 at 10:53:27PM -0500, Sebastian Silva wrote:
> > > El 13/05/13 15:48, Walter Bender escribi�:
I've moved Report link to the front activity page,

http://network.sugarlabs.org/hub/?activities&org.laptop.TurtleArtActivity&reports

but there are no reports for TA.
icarito: could you check if it is uplaoding issue or people just
didn't press OK in Upload Report dialog.

At the same time, there are bunch of reports for Physics activity:

http://network.sugarlabs.org/hub/?activities&org.laptop.physics&reports

Also, I've added filtering SN content by layers to global menu
(gear icon in the top-right corner).

--
Aleksey

Walter Bender

unread,
May 16, 2013, 8:08:48 AM5/16/13
to sugar-...@googlegroups.com, Lista de correo del equipo Somos-Azúcar, todos
On Thu, May 16, 2013 at 12:38 AM, Aleksey Lim <als...@sugarlabs.org> wrote:
On Tue, May 14, 2013 at 08:22:46AM -0400, Walter Bender wrote:
> On Tue, May 14, 2013 at 2:45 AM, Aleksey Lim <als...@sugarlabs.org> wrote:
>
> > On Mon, May 13, 2013 at 10:53:27PM -0500, Sebastian Silva wrote:
> > > El 13/05/13 15:48, Walter Bender escribió:

Re Physics, maybe you could test v11.4 [1] which I am working to upstream with Gonzalo. At least on recent OLPC builds, it addresses the problem with activity exit you seem to be encountering. But I haven't tested it on Sugar 0.94.

[1] http://people.sugarlabs.org/walter/Physics-11.4.xo

regards.

-walter

Also, I've added filtering SN content by layers to global menu
(gear icon in the top-right corner).

--
Aleksey
--
You received this message because you are subscribed to the Google Groups "Sugar Network" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sugar-networ...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.


Alan Jhonn Aguiar Schwyn

unread,
May 16, 2013, 4:20:49 PM5/16/13
to somos...@lists.sugarlabs.org, sugar-...@googlegroups.com, todos
One question:

The lunch time is the first time? In a second time is the same?

Python in the first launch it's a bit slower because uses the JIT compile and generates
the .pyo (or .pyc).



Date: Thu, 16 May 2013 08:08:48 -0400
From: walter...@gmail.com
To: sugar-...@googlegroups.com
CC: somos...@lists.sugarlabs.org; to...@somosazucar.org
Subject: Re: [somos-azucar] Official Feedback for Hexoquinasa v0.9
_______________________________________________ somosazucar mailing list somos...@lists.sugarlabs.org http://lists.sugarlabs.org/listinfo/somosazucar

Sebastian Silva

unread,
May 16, 2013, 7:04:00 PM5/16/13
to Lista de correo del equipo Somos-Azúcar, sugar-...@googlegroups.com, todos
It is likely that they hit X (cancel) instead of √ (ok). 

I will ask if they can reproduce the issue and instruct them to press (ok) in connected state for uploading logs, hopefully they can do it before next release.
In any case I've tested latest Turtle Blocks and found no issues so I marked it for inclusion in the next image.


Sebastian Silva
http://somosazucar.org/



2013/5/15 Aleksey Lim <als...@sugarlabs.org>
On Tue, May 14, 2013 at 08:22:46AM -0400, Walter Bender wrote:
> On Tue, May 14, 2013 at 2:45 AM, Aleksey Lim <als...@sugarlabs.org> wrote:
>
> > On Mon, May 13, 2013 at 10:53:27PM -0500, Sebastian Silva wrote:
> > > El 13/05/13 15:48, Walter Bender escribió:

--
Aleksey

Sebastian Silva

unread,
May 16, 2013, 7:05:30 PM5/16/13
to sugar-...@googlegroups.com, Lista de correo del equipo Somos-Azúcar, todos
Great, I will try it asap. This failure was mysterious to me.



2013/5/16 Walter Bender <walter...@gmail.com>

Sebastian Silva

unread,
May 16, 2013, 7:07:11 PM5/16/13
to sugar-...@googlegroups.com, somos...@lists.sugarlabs.org, todos
They did say that they tried it several times. I'm not sure .pyc generation causes a delay because I *think* there is a stage during image building that all python files get precompiled. Still some activities (turtle art) have extra steps on first start (png generation i think).

Thanks for all ideas!



2013/5/16 Alan Jhonn Aguiar Schwyn <ala...@hotmail.com>

Sebastian Silva

unread,
May 21, 2013, 12:31:01 PM5/21/13
to sugar-...@googlegroups.com, walter bender, Lista de correo del equipo Somos-Azúcar, todos

2013/5/16 Walter Bender <walter...@gmail.com>



Re Physics, maybe you could test v11.4 [1] which I am working to upstream with Gonzalo. At least on recent OLPC builds, it addresses the problem with activity exit you seem to be encountering. But I haven't tested it on Sugar 0.94.

[1] http://people.sugarlabs.org/walter/Physics-11.4.xo


I tried this version. Subjectively it seems to run faster. It also does not trigger the failure report on exit.

However, it doesn't save to the Journal, with errors trying to exit: all changes will be lost.

There is a traceback in the logs, please find them attached.

Thanks for helping out,

Sebastian


org.laptop.physics-1.log

Walter Bender

unread,
May 21, 2013, 9:26:02 PM5/21/13
to Sebastian Silva, sugar-...@googlegroups.com, Lista de correo del equipo Somos-Azúcar, todos
I rolled another version [1] today. But I don't think it will fix this
issue. Will try to look tomorrow on a 0.94 machine.

[1] http://people.sugarlabs.org/walter/Physics-11.5.xo


-walter

Sebastian Silva

unread,
May 23, 2013, 3:09:29 AM5/23/13
to Walter Bender, sugar-...@googlegroups.com, Lista de correo del equipo Somos-Azúcar
Hi!
I managed to take a look myself and fixed the issue with Journal preview
which was
preventing to save on exit. I also made it work even for when closing
from the Sugar
frame. Apply with ( you probably knew this but I didn't :-) ):

patch activity.py < physics.patch

Do you think it might be possible to release this new version soon?
It's a big improvement and we build images from ASLO/SN released
activities only.

Regards,
Sebastian

El 21/05/13 20:26, Walter Bender escribi�:
physics.patch

Walter Bender

unread,
May 23, 2013, 8:52:29 AM5/23/13
to sugar-...@googlegroups.com, Lista de correo del equipo Somos-Azúcar
Thanks for digging into this. Using can_close is the correct approach.
Hopefully [1] will work on 0.94. There is allegedly a problem with
rotation I am trying to track down, so there will probably be a v11.7
coming out soon, but probably not an issue for you, since there are no
XO 4 touch machines in
Peru.

regards.

-walter

[1] http://people.sugarlabs.org/walter/Physics-11.6.xo

On Thu, May 23, 2013 at 3:09 AM, Sebastian Silva
<seba...@fuentelibre.org> wrote:
> Hi!
> I managed to take a look myself and fixed the issue with Journal preview
> which was
> preventing to save on exit. I also made it work even for when closing from
> the Sugar
> frame. Apply with ( you probably knew this but I didn't :-) ):
>
> patch activity.py < physics.patch
>
> Do you think it might be possible to release this new version soon?
> It's a big improvement and we build images from ASLO/SN released activities
> only.
>
> Regards,
> Sebastian
>
> El 21/05/13 20:26, Walter Bender escribió:
>
>> I rolled another version [1] today. But I don't think it will fix this
>> issue. Will try to look tomorrow on a 0.94 machine.
>>
>> [1] http://people.sugarlabs.org/walter/Physics-11.5.xo
>>
>>
>> -walter
>>
>> On Tue, May 21, 2013 at 12:31 PM, Sebastian Silva
>> <seba...@fuentelibre.org> wrote:
>>>
>>> 2013/5/16 Walter Bender <walter...@gmail.com>
>>>>
>>>>
>>>>
>>>> Re Physics, maybe you could test v11.4 [1] which I am working to
>>>> upstream
>>>> with Gonzalo. At least on recent OLPC builds, it addresses the problem
>>>> with
>>>> activity exit you seem to be encountering. But I haven't tested it on
>>>> Sugar
>>>> 0.94.
>>>>
>>>> [1] http://people.sugarlabs.org/walter/Physics-11.4.xo
>>>>
>>> I tried this version. Subjectively it seems to run faster. It also does
>>> not
>>> trigger the failure report on exit.
>>>
>>> However, it doesn't save to the Journal, with errors trying to exit: all
>>> changes will be lost.
>>>
>>> There is a traceback in the logs, please find them attached.
>>>
>>> Thanks for helping out,
>>>
>>> Sebastian
>>>
>>>
>>
>>
>
> --
> You received this message because you are subscribed to the Google Groups
> "Sugar Network" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sugar-networ...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>



Sebastian Silva

unread,
May 27, 2013, 3:15:55 PM5/27/13
to sugar-...@googlegroups.com, Walter Bender, Lista de correo del equipo Somos-Azúcar
Hi,
We've been testing latest Physics pre-release using Sugar 0.94, but
the activity was not closing at all. Upon closer look, it seems you missed
a bit of my patch (attached). A preview is pre-fetched into self.preview
because by the time get_preview is called when closing, pygame is already
stopped. With patch, If get_preview finds pre-fetched preview, it will not
try to get from pygame. Now it works properly.

With regard to releasing Hexoquinasa with pre-release Physics, I managed
to shoehorn unzipping pre-release Physics 11.6 into buillding image, but this
goes against the spirit of Sugar Network's goal of providing feedback for
released activities. Hopefully we'll see a release soon. :-)

The matter of using pre-release or customized Implementations of activities in
Sugar Network is not clear, maybe it will happen extending current "layer" [1]
functionality or making available different "channels" as different Contexts.

Probably Alsroot has a vision about how to handle these cases,

Regards,
Sebastian
[1] http://wiki.sugarlabs.org/go/Platform_Team/Sugar_Network/Objects_model

El 23/05/13 07:52, Walter Bender escribió:
canclose.patch

Walter Bender

unread,
May 27, 2013, 3:22:51 PM5/27/13
to Sebastian Silva, sugar-...@googlegroups.com, Lista de correo del equipo Somos-Azúcar
Sorry about that.

Could you please generate your patch using:

git diff > blah.patch

or

diff -rup [old] [new]

or

git format-patch HEAD^


I cannot apply the patch as it is currently formatted (or the previous
one). Thx.

Sebastian Silva

unread,
May 27, 2013, 3:54:07 PM5/27/13
to somos...@lists.sugarlabs.org, Walter Bender, sugar-...@googlegroups.com
Here's the reformatted patch, great to know that.

Thank you for improving Physics activity!

Regards,
Sebastian

El 27/05/13 14:22, Walter Bender escribi�:
>> El 23/05/13 07:52, Walter Bender escribi�:
canclose.patch

Walter Bender

unread,
May 27, 2013, 4:05:33 PM5/27/13
to sugar-...@googlegroups.com, Lista de correo del equipo Somos-Azúcar
Please test http://people.sugarlabs.org/walter/Physics-11.7.xo

-walter

On Mon, May 27, 2013 at 3:54 PM, Sebastian Silva
<seba...@somosazucar.org> wrote:
> Here's the reformatted patch, great to know that.
>
> Thank you for improving Physics activity!
>
> Regards,
> Sebastian
>
> El 27/05/13 14:22, Walter Bender escribió:
>>> El 23/05/13 07:52, Walter Bender escribió:
>>>
>>> Thanks for digging into this. Using can_close is the correct approach.
>>> Hopefully [1] will work on 0.94. There is allegedly a problem with
>>> rotation I am trying to track down, so there will probably be a v11.7
>>> coming out soon, but probably not an issue for you, since there are no
>>> XO 4 touch machines in
>>> Peru.
>>>
>>> regards.
>>>
>>> -walter
>>>
>>> [1] http://people.sugarlabs.org/walter/Physics-11.6.xo
>>>
>>>
>>
>>
>
> --
> You received this message because you are subscribed to the Google Groups
> "Sugar Network" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sugar-networ...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>



Reply all
Reply to author
Forward
0 new messages