Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-13 Thread Jonas Smedegaard
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Wed, Apr 08, 2009 at 01:40:57PM +0100, Peter Robinson wrote:
 There are two separate issues I encountered with TA (and Measure). 
 One is the need for patches to the alsa audio to accommodate the 
 special OLPC hardware modifications.

That would need to be accepted upstream by the alsa project, and then 
distros will automatically get the changes when they are part of a 
release.

Or (less effective but sometimes possible if in a hurry): Convince 
distributions to accept patches, even if upstream rejects it (or 
temporarily while upstream procedures might be slow).

For Debian, the package this would probably land in would be 
alsa-firmware-loaders (if I understand it correctly), and the people to 
contact about it are reachabel on this email address:

pkg-alsa-de...@lists.alioth.debian.org


Sometimes it might even ease upstream adoption if first working with 
distributions on shaping/polishing patches.


Kind regards,

  - Jonas

- -- 
* Jonas Smedegaard - idealist og Internet-arkitekt
* Tlf.: +45 40843136  Website: http://dr.jones.dk/

  [x] quote me freely  [ ] ask before reusing  [ ] keep private
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAknfJwUACgkQn7DbMsAkQLiJKACglphhhquwpqVytDJe1nPYRd5a
5OYAnRmVeYSjoluVH4OqbWv4cUxukrae
=Ve+F
-END PGP SIGNATURE-
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-11 Thread Peter Robinson
 There are two separate issues I encountered with TA (and Measure).
 One is the need for patches to the alsa audio to accommodate the
 special OLPC hardware modifications.

That would need to be accepted upstream by the alsa project, and then
distros will automatically get the changes when they are part of a
release.

 Or (less effective but sometimes possible if in a hurry): Convince
 distributions to accept patches, even if upstream rejects it (or
 temporarily while upstream procedures might be slow).

Fedora at least has a fairly strict policy of using upstream only and
will generally reject patches.

Peter
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread Sascha Silbe

On Wed, Apr 08, 2009 at 03:44:05AM -0400, Seth Woodworth wrote:


I'd like to encourage everyone to build on the great work of Arjun
Sarwal and help us extend the sensor interface with our peripheral --
the heart rate monitor.
Looks great! As I intended to do a similar device in the future (it's 
going to measure SpO2 (oxygen saturation) as well as that's what I'm 
interested in), I'd like to see the schematics. Are they available 
somewhere?
Do you just amplify the current through a photo transistor or is there 
something more fancy going on, e.g. AGC?


CU Sascha

--
http://sascha.silbe.org/
http://www.infra-silbe.de/


signature.asc
Description: Digital signature
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread Walter Bender
Alas, the blocker with Measure (and Turtle Art with Sensors) is that
we haven't come to consensus regarding a packaging strategy for the
binary drivers we include for support of the sensor input; these
differ by architecture, which wasn't a problem when we were building
for just one architecture. Perhaps the packaging gurus can chime in.
(See http://dev.sugarlabs.org/ticket/155).

-walter

On Wed, Apr 8, 2009 at 3:44 AM, Seth Woodworth s...@laptop.org wrote:
 -- Forwarded message --
 From: Tom Boonsiri tom.boons...@gmail.com
 Date: Wed, Apr 8, 2009 at 2:31 AM
 Subject: Heart Rate Monitor Peripheral
 To: seth.woodwo...@gmail.com


 Fellow developers,

 Is your Measure activity feeling neglected? If so, shame on you. =)
 I'd like to encourage everyone to build on the great work of Arjun
 Sarwal and help us extend the sensor interface with our peripheral --
 the heart rate monitor.

 Pictures of the device:
 http://farm4.static.flickr.com/3559/3419907541_f62b168dce_m.jpg
 http://farm4.static.flickr.com/3558/3420715230_00e29b7787_m.jpg

 It's a relatively simple device that measures the blood flow in your
 finger with an infrared sensor. Powered by USB, the device sends
 measurements to the XO via the AC/DC sensor interface (audio jack).
 Using a Measure variant with a heart beat detection method, we are
 able to display the heart rate (as shown in the pic).

 If you're a developer interested in integrating biofeedback into your
 application, please email me for more details on how you can get your
 hands on a few of our peripherals. Otherwise, we are looking for
 developers who can help us evolve the Measure activity to better suit
 the lesson plan we have created for the device. The device is not of
 clinical quality and solely for educational purposes. We are
 definitely interested in feedback on our direction.

 In the short-term, it would be interesting to develop a gui where you
 could structure a family tree (even extending it to branches for
 relatives) and allow kids to record measurements for various family
 members. This could potentially evolve into other health education
 efforts with a wiki backbone to support health wellness.

 Please send me your feedback and your interest!

 Thanks,
 Tom Boonsiri -- OLPC GoldenState
 (tom.boons...@gmail.com)
 ___
 IAEP -- It's An Education Project (not a laptop project!)
 i...@lists.sugarlabs.org
 http://lists.sugarlabs.org/listinfo/iaep




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


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread Seth Woodworth
Right now this device works primarily with the XO-1, taking advantage
of the AC/DC conversion going on in the XO's audio-in port.  But if
the gain were tweaked a bit more it could work with other audio cards.

The fact that it is XO-1 Hardware specific means that packaging for
different architectures isn't a problem in the short run, correct?

On Wed, Apr 8, 2009 at 7:56 AM, Peter Robinson pbrobin...@gmail.com wrote:
 Alas, the blocker with Measure (and Turtle Art with Sensors) is that
 we haven't come to consensus regarding a packaging strategy for the
 binary drivers we include for support of the sensor input; these
 differ by architecture, which wasn't a problem when we were building
 for just one architecture. Perhaps the packaging gurus can chime in.
 (See http://dev.sugarlabs.org/ticket/155).

 So the sensors are the ones that can get plugged into the audio
 input/output on the XO? Is this custom to the XO or could be
 conceivably be implemented/used on other audio cards?

 Peter

___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread Peter Robinson
There's generally no issue with hardware specific packages in distros
(there certainly isn't an issue in Fedora), from the quick read I had
of the ticket it looks more like the turtle activity didn't work on
the non XO hardware. I presume that's because it couldn't load the
specific libraries. Generally you would put the hardware specific bits
in either a separate package from the app or a sub-package so that for
devices where the hardware isn't present it doesn't need to be
installed. This though would require the app to be able to detect the
whether drivers/plugins are present or not and deal with it. There
would be no issue installing them by default as long as it doesn't
disturb other sound cards that aren't supported.

Peter

BTW the idea of the hardware sensors is very cool. Is there somewhere
you can see what sensors are available for the XO and buy them?


 Right now this device works primarily with the XO-1, taking advantage
 of the AC/DC conversion going on in the XO's audio-in port.  But if
 the gain were tweaked a bit more it could work with other audio cards.

 The fact that it is XO-1 Hardware specific means that packaging for
 different architectures isn't a problem in the short run, correct?

 On Wed, Apr 8, 2009 at 7:56 AM, Peter Robinson pbrobin...@gmail.com wrote:
 Alas, the blocker with Measure (and Turtle Art with Sensors) is that
 we haven't come to consensus regarding a packaging strategy for the
 binary drivers we include for support of the sensor input; these
 differ by architecture, which wasn't a problem when we were building
 for just one architecture. Perhaps the packaging gurus can chime in.
 (See http://dev.sugarlabs.org/ticket/155).

 So the sensors are the ones that can get plugged into the audio
 input/output on the XO? Is this custom to the XO or could be
 conceivably be implemented/used on other audio cards?

 Peter


___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread Walter Bender
Peter,

In the ticket, I was trying to express my ignorance of how to proceed,
but not suggest that the distros haven't dealt with these sorts of
issues in the past.

There are two separate issues I encountered with TA (and Measure). One
is the need for patches to the alsa audio to accommodate the special
OLPC hardware modifications. The other is a means within the Sugar
bundling system to include binaries for multiple architectures. I
actually had written a work-around for the former, but I remain
ignorant about the details of how to best handle the latter. (At one
point, Sascha had helped me include a script that would build the
appropriate binaries at install time, but that solution depends upon a
build environment be installed, which is not the default for most OLPC
systems. Oops.)

Regarding available sensors, Arjun describes some on the Measure page
in the OLPC wiki

http://wiki.laptop.org/go/Measure

There are also many people working on USB sensor input devices that
would be great to support in general.

thanks.

-walter

On Wed, Apr 8, 2009 at 8:22 AM, Peter Robinson pbrobin...@gmail.com wrote:
 There's generally no issue with hardware specific packages in distros
 (there certainly isn't an issue in Fedora), from the quick read I had
 of the ticket it looks more like the turtle activity didn't work on
 the non XO hardware. I presume that's because it couldn't load the
 specific libraries. Generally you would put the hardware specific bits
 in either a separate package from the app or a sub-package so that for
 devices where the hardware isn't present it doesn't need to be
 installed. This though would require the app to be able to detect the
 whether drivers/plugins are present or not and deal with it. There
 would be no issue installing them by default as long as it doesn't
 disturb other sound cards that aren't supported.

 Peter

 BTW the idea of the hardware sensors is very cool. Is there somewhere
 you can see what sensors are available for the XO and buy them?


 Right now this device works primarily with the XO-1, taking advantage
 of the AC/DC conversion going on in the XO's audio-in port.  But if
 the gain were tweaked a bit more it could work with other audio cards.

 The fact that it is XO-1 Hardware specific means that packaging for
 different architectures isn't a problem in the short run, correct?

 On Wed, Apr 8, 2009 at 7:56 AM, Peter Robinson pbrobin...@gmail.com wrote:
 Alas, the blocker with Measure (and Turtle Art with Sensors) is that
 we haven't come to consensus regarding a packaging strategy for the
 binary drivers we include for support of the sensor input; these
 differ by architecture, which wasn't a problem when we were building
 for just one architecture. Perhaps the packaging gurus can chime in.
 (See http://dev.sugarlabs.org/ticket/155).

 So the sensors are the ones that can get plugged into the audio
 input/output on the XO? Is this custom to the XO or could be
 conceivably be implemented/used on other audio cards?

 Peter






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


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread Peter Robinson
 There are two separate issues I encountered with TA (and Measure). One
 is the need for patches to the alsa audio to accommodate the special
 OLPC hardware modifications.

That would need to be accepted upstream by the alsa project, and then
distros will automatically get the changes when they are part of a
release.

 The other is a means within the Sugar
 bundling system to include binaries for multiple architectures. I
 actually had written a work-around for the former, but I remain
 ignorant about the details of how to best handle the latter. (At one
 point, Sascha had helped me include a script that would build the
 appropriate binaries at install time, but that solution depends upon a
 build environment be installed, which is not the default for most OLPC
 systems. Oops.)

That would be just a packaging procedure. In the case of Fedora it
would be just the arch dependant rpms as opposed to using a activity
.xo file which I believe are arch independent. Unless I've missed
something here.

 Regarding available sensors, Arjun describes some on the Measure page
 in the OLPC wiki

Thanks!

 http://wiki.laptop.org/go/Measure

 There are also many people working on USB sensor input devices that
 would be great to support in general.

In fedora at least from a driver perspective it would land in Fedora
as soon as the upstream kernel has the drivers. I'm not sure whether
there is a defined api for accessing those sort sensors on linux. If
there was I would presume as the hardware is added that TA/measure
would automatically get support for the hardware as its added, if
there's not I presume there would need to be individual support added
for each device.

Peter
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread Peter Robinson
 Alas, the blocker with Measure (and Turtle Art with Sensors) is that
 we haven't come to consensus regarding a packaging strategy for the
 binary drivers we include for support of the sensor input; these
 differ by architecture, which wasn't a problem when we were building
 for just one architecture. Perhaps the packaging gurus can chime in.
 (See http://dev.sugarlabs.org/ticket/155).

So the sensors are the ones that can get plugged into the audio
input/output on the XO? Is this custom to the XO or could be
conceivably be implemented/used on other audio cards?

Peter
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread Rafael Enrique Ortiz Guerrero
Hi


In a related note
Kristianpaulhttp://co.sugarlabs.org/wiki/index.php?title=Usuario:Kristianpaulaction=editredlink=1of
Sugarlabs .co is testing measure in Soas (with arjun's help) in an
effort
to  adapt it to all kinds of hardware.

http://co.sugarlabs.org/go/Sobre_Measure

Rafael Ortiz


On Wed, Apr 8, 2009 at 7:40 AM, Peter Robinson pbrobin...@gmail.com wrote:

  There are two separate issues I encountered with TA (and Measure). One
  is the need for patches to the alsa audio to accommodate the special
  OLPC hardware modifications.

 That would need to be accepted upstream by the alsa project, and then
 distros will automatically get the changes when they are part of a
 release.

  The other is a means within the Sugar
  bundling system to include binaries for multiple architectures. I
  actually had written a work-around for the former, but I remain
  ignorant about the details of how to best handle the latter. (At one
  point, Sascha had helped me include a script that would build the
  appropriate binaries at install time, but that solution depends upon a
  build environment be installed, which is not the default for most OLPC
  systems. Oops.)

 That would be just a packaging procedure. In the case of Fedora it
 would be just the arch dependant rpms as opposed to using a activity
 .xo file which I believe are arch independent. Unless I've missed
 something here.

  Regarding available sensors, Arjun describes some on the Measure page
  in the OLPC wiki

 Thanks!

  http://wiki.laptop.org/go/Measure
 
  There are also many people working on USB sensor input devices that
  would be great to support in general.

 In fedora at least from a driver perspective it would land in Fedora
 as soon as the upstream kernel has the drivers. I'm not sure whether
 there is a defined api for accessing those sort sensors on linux. If
 there was I would presume as the hardware is added that TA/measure
 would automatically get support for the hardware as its added, if
 there's not I presume there would need to be individual support added
 for each device.

 Peter
 ___
 Devel mailing list
 Devel@lists.laptop.org
 http://lists.laptop.org/listinfo/devel

___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread Richard A. Smith
Walter Bender wrote:

 Peter,
 
 In the ticket, I was trying to express my ignorance of how to proceed,
 but not suggest that the distros haven't dealt with these sorts of
 issues in the past.
 
 There are two separate issues I encountered with TA (and Measure). One
 is the need for patches to the alsa audio to accommodate the special
 OLPC hardware modifications. The other is a means within the Sugar
 bundling system to include binaries for multiple architectures. I

What binary drivers are you talking about?


-- 
Richard Smith  rich...@laptop.org
One Laptop Per Child
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: [IAEP] Fwd: Heart Rate Monitor Peripheral

2009-04-08 Thread John Gilmore
 Alas, the blocker with Measure (and Turtle Art with Sensors) is that
 we haven't come to consensus regarding a packaging strategy for the
 binary drivers we include for support of the sensor input; these
 differ by architecture, which wasn't a problem when we were building
 for just one architecture. Perhaps the packaging gurus can chime in.
 (See http://dev.sugarlabs.org/ticket/155).

I'm guessing below, because I can't find the source of the binary
driver you are discussing.  The bug report doesn't point to it, and
it isn't in an obvious place in the OLPC git or the Sugar gitorious.

It's not a binary driver, it's just a library implemented in C or C++
rather than Python.  This is only novel to the Sugar team -- the vast
majority of packages in every Linux distro are already this way.

There's already a python-alsaaudio package in Ubuntu:

  http://packages.ubuntu.com/intrepid/sound/python-alsaaudio

If this is the module you're calling, your sugar-measure-activity
package just has to mention that as a dependency, and all should be
well.

If you have your own unique python-alsaaudio module, you'll probably
want to merge it with the already-existing one, and fix up Measure
so it can call the existing public one.

John
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel