Hello Guido,

Thanks for taking time and enabling a discussion. First please allow me to urgently dispel the notion of "punishment", because it's actually all about positive reinforcement and earning, rather than taking away. In the demo system a parent can add any number of white-listed websites. You could also award 1000 hours of credits just for completing a 5 minute activity, for example.

The other part which I see needs better explanation is the connection to free education software. The goal is not to teach kids to embrace free software. That's not a bad goal or anything, but that's not what this platform can do. Rather, this platform (and proposed campaign) can result in the development of more free education software by giving people a reason to make it beyond mere altruism.

Basically, the above 2 paragraphs refer to 2 distinct communities: parents and developers.

For readers who have not seen the latest slides https://github.com/ccosse/NetDispenserEcoSystem/blob/master/Credit-Meter-Assisted-Learning.pdf, here is the idea: I began developing free education software 16 years ago, initially for my own 2 young children. I followed-through with some of the better ideas and got them onto Linux distros and into schools. My motivation was merely "faith" that something good would happen later, perhaps that a free education software revolution would take-off and there would be rewards down the line. Motivation aside, with the applications themselves there was always the problem of getting my kids to work carefully, to focus and to think.

Fast-forward to 2012 and I have been led, by "necessity", to develop the proposed system: a Raspberry-Pi WiFi HotSpot which firewalls them in unless they use credits which they've earned to open the firewall to their list of devices. They have free access to any number of white-listed websites, including the one(s) where they earn credits by completing activities. I call them the "credit-meter" and the "credit-feeder", respectively.

My results were very positive: the kids had a bird - bird-feeder relationship with the system, i.e. they served themselves when they needed more credits and it was essentially hands-off and drama-free for me. And they would have plenty to talk about at dinner, such as things they read, etc.

The big thing, however, was that any activity could be substituted and they would be motivated to make an effort just the same. This is the connection to free software development, i.e. the platform has the potential to accommodate anything you plug-in. It works as a single point of motivation for any content. I'm no biologist, but a biology activity could certainly work here if one existed. So, for the thing to reach its potential it should attract contributions from the largest number of developers possible. And this, in turn, would require complete openness and transparency so as to include FSF/E and anyone else.

Charging parents a subscription in order to compensate developers would result in value flowing in two directions: money from parent to developer, and applications from developer to parent. Let the money be handled by a trustworthy foundation, and let each parent distribute their own subscription fee among developers as they see fit. Require each activity to run independently of the platform and to be licensed as free software. The result is a transparent and healthy new market for free education software.

This plan offers an opportunity to begin a new chapter in free education software development by providing incentive, an organized community, a user base, a market and an increased likelihood of developer satisfaction, at the same time as focusing positive attention on the FSF/E. There is plenty of room for new innovations in education software, and there are plenty of creative people who might be interested if there was an active community and some incentive beyond altruism. The tools are there, the ideas are there, the developers are there, but without an organized community, without strength in numbers, those ideas will wither and die.

Best regards, Charles Cossé

On Sun, Jan 8, 2017 at 5:50 PM, Guido Arnold guido@fsfe.org wrote:

Hello Charles,

Sorry for the delay. I just found a few minutes to have a look through the slides of yours.

On Wed, Jan 04, 2017 at 02:28:59PM -0700, Charles Cossé wrote:

...

[2]Here are slides which describe the idea visually. The last slide explains the Edu-FLOSS connection.

I would like to re-develop the project featured in the slides as a community, inclusive of FSF/E.

Thanks for the offer! I have to say that I am sceptical about this project. My kids are not in an age where they would find themselves alone with internet access yet but I will soon get there. I am curious how happy you are so far with the outcome of the experiences you gained.

With the little experience I have, I'd rather send my daughter an email (if I can't speak to her in person) with URLs to read or tasks to do than forcing her to deal with a carrot-and-stick system.

If I come home and she hasn't done anything, I can still think of a way of punishment if I see the need for it. Cutting off access to information would not be my first choice. I mentioned my case earlier but reckon that this seems to be the only way that works with youths nowadays.

Back to your case: If the young person doesn't like the task at hand, the options are

• screw it and grab a book of her choice from the shelf (good case scenario)
• bypass the system (e.g. get access to the RsPi and replace the SDCard with a stock Raspian image (not really a bad case scenario ;)
• do it anyway to earn the frickin' credits and learn to hate the subject (bad case scenario)
• go outside to the next fast food joint with free Wi-Fi (and probably start to hate you for pushing her to spend the little money she's got for food neither you nor her want her to eat (worst case scenario)

There are certainly more options to list but I chose those to illustrate my scepticism in practice.

My main point thoug is that I also fail to see how this will teach the kid to embrace Free Software as an empowering, positive thing that is worth fighting for (or: endure minor inconveniences for now).

As I said, I merely skimmed through the slides and may have missed the point, but didn't want to leave this initiative unanswered for too long.

Looking forward to your reply and a fruitful discussion!

Greetings,

Guido

-- Guido Arnold Free Software Foundation Europe https://blogs.fsfe.org/guido [] Edu team & German team OpenPGP Key-ID: 0x51628D75 [][][] Get active! XMPP: guido@jabber.fsfe.org || https://fsfe.org

---

Discussion mailing list Discussion@lists.fsfe.org https://lists.fsfe.org/mailman/listinfo/discussion

Hi Bastien,

Wow! What fantastic work at http://try.sugarizer.org! I just went through all of the activities and it was a great user experience, design and everything. I also tried kalulu-the-gathering but encountered an error when trying to run testAll ... but looking forward to exploring it, as well.

So about the try.sugarizer.org site, may I ask: how many developers, how much time to develop, how is it licensed, and of most interest ... what is the incentive and motivation of developers to participate?

I respond to more inline, below ...

On Mon, Jan 9, 2017 at 12:27 AM, Bastien Guerry bzg@gnu.org wrote:

Hi Charles,

I find the whole idea very interesting, thanks for sharing it.

At OLPC France (a grassroots organization, born in 2008 after I spent some time working for OLPC), we are developing http://sugarizer.org

It is a rewrite of the Sugar environment (https://sugarlabs.org) in HTML and Javascript. You can try it here: http://try.sugarizer.org

I've also dedicated some time last year to a free software to let children learn how to read, write and count. It is called "Kalulu", it is currently under heavy development, and will be available soon.

You can see the code here : https://dev.kalulu-education.org/cyprien/kalulu-the-gathering

but the developers still need to package and document it properly.

I'm mentioning those two projects because several of your concerns were part of the discussions we held for those two software.

For Sugarizer: children use it on their tablets, and they connect to a Raspberry Pi 3 located in the classroom. When children connect to the Sugarizer server on our RPi3, they can resume their sessions, store their activities, discuss together, etc. The RPi3 is also used to store videos and audio, because we found out accessing internet while in the classroom was neither easy nor very useful, attention-wise.

I see, and also in the framework which I wrote, the credit-earning website could run on the R-Pi as both R-Pi/credit-meter and online credit-feeder sites run on identical Gentoo Linux installations. However, in my scheme, running the activities on the R-Pi would defeat the purpose of the ecosystem by allowing the users to drift away and not continue engagement with developers, including the distribution of their proposed subscription fee among devs of their choice. In 2014 I did produce a 4G R-Pi SD image called "developer's sandbox" which ran both sites on the R-Pi.

As the scheme was difficult to explain, in part because it aimed to appeal in 2x different ways to 2x different communities, I put-together a demo site with both the credit-meter and credit-feeder *online* this time, and working with each other. Then also, to complete the picture, I added-in the interface where parents could add play-money to their accounts and give freely to (imaginary) developers. In other words, I assembled all the pieces and tried to make it like a "game" where people could try the various roles (parent, child, developer) without using real money. I resurrected that site a couple weeks ago at www.autoteach.net, and it is what it is ... but that's what I'd like to re-develop with help from interested people. Or perhaps as some collaboration with other efforts, such as sugarizer, that I was unaware of.

For now children are free to play with any activity, but having a framework where the teacher can say : "Children A has to learn X in order to access to Activity Z" would certainly be useful.

As for Kalulu, conditional learning paths is at the very heart of the conception of the software. Children are required to learn a lesson, then can reinforce what they learned by freely playing with treasures that unlock progressively. (The cognitive scientist Stanislas Dehaene has been involved in designing this software, and it implements well tested research about motivation and learning, at least for learning how to read, write and count.

So in general, I invite you to test sugarizer and I'll let you know when Kalulu is available for testing, as both may be of interest for your project -- either as softwares you can contribute to, or as platforms that would benefit from your ideas and code.

Both are HTML/JS and use an Apache 2.0 license.

All best,

-- Bastien

So I should explain that all I ever wanted to do was write more indie education software ... I have many ideas, but there's no incentive. My kids are grown but I've still got ideas and interest. So my goal is to be a member of a community which enables free, independent edu software developers to go about doing what they love. And the whole scheme I described is the only workable equation that I've been able to put together. But yeah, I want to be doing what you are doing, i.e. working on the software ... but I'm prepared to devote myself to community building and leading such a campaign if FSF/E can adopt / endorse ... I do truly believe that the whole ecosystem equation can not only work, but attract developers and produce a lot more free education software -- Javascript developed to run as plugin to credit-feeder website OR run statically and independently anywhere.

Best regards, -Charles

[[[ To any NSA and FBI agents reading my email: please consider ]]] [[[ whether defending the US Constitution against all enemies, ]]] [[[ foreign or domestic, requires you to follow Snowden's example. ]]]

The repository is here: https://github.com/llaske/sugarizer

As a general point, I'd like to urge people to avoid github.com. See https://gnu.org/software/repo-criteria.html for the reasons.

Hello Guido,

On Tue, Jan 10, 2017 at 5:39 PM, Guido Arnold guido@fsfe.org wrote:

On Mon, Jan 09, 2017 at 11:54:13AM -0700, Charles Cossé wrote:

I'm puzzled with this too. What is meant by "engagement of the user with developers"?

I believe that the scheme addresses two things which have always been lacking in free software: incentive beyond mere ego and altruism, and follow-through in order that projects don't end-up "half-baked". So the "engagement" refers to both of these: users of the platform, i.e. parents, supporting activities with portions of their subscription fee, and the ongoing process of activity improvement which that would also tend to promote.

And, at the risk of redundancy: the subtle cornerstone of the whole scheme is that kids end-up learning very well, however that is _not_ their motivation. Their motivation is to _earn_ credits for internet access, but turns out that the results are the same, i.e. it is in their interest to focus and try hard. I'm not aware of any other scheme which results in kids asking for more math homework. The scheme is a way to harness that effectiveness.

Like Bastien, I think that endorsement of any organisation shouldn't

be necessary. To my knowledge, FSFE does not endorse any particular Free Software project for numerous reasons (e.g. where to draw the line what's worth endorsing?).

The "Current campaigns" page at: http://www.fsf.org/campaigns/ has a campaign for "High Priority Free Software Projects". Could this possibly fall under that header? Also, I just received FSF email attempting to raise $3,700,000 for the Talos Secure Workstation.

I would like to promote this scheme sometime soon by writing magazine articles, interviews, meetings with teachers, staff, parents, homeschool groups ... there are so many potential stakeholders who might be interested, all with the ultimate goal of producing a tidal wave of Edu-FLS development and focusing positive attention on the FSF/* which would potentially help other campaigns as well.

Best regards, -Charles

Hi Richard,

Richard Stallman rms@gnu.org writes:

I see two freedom concerns, in the project description.

Thanks for sharing those concerns.

1. The tablet.

I think all tablets are sold with a nonfree operating system, but maybe some tablet can run Replicant (the free version of Android).

I presume your software runs on Android. Is the entire app free software? Specifically, does it avoid nonfree libraries, including the Google Play Servics library?

The tablets we use in this school use Android, with a limited set of available services. We removed the Google Play services, but there are other parts of the tablets that are not free.

Our primary goal is to develop and test Sugarizer with children and teachers and we focus on that. We welcome help on using completely free systems on the tablets.

2. The Raspberry Pi.

The Raspberry Pi is an unfortunate choice because it can't run _at all_ without some nonfree software.

If what you need is a single board computer, how about recommending one of the models that can at least start up without nonfree software?

I'm all for it!

See fsf.org/resources/hw/single-board-computers for information.

Thanks for the information. As far as I understand, all single board computers have at least some flaws.

I just bought a BeagleBoard, which seems better than the RPi3.

I'll try it and we might consider replacing the RPi3.

Thanks,

Richard Stallman rms@gnu.org writes:

Can you help raise money to hire someone to work on reverse-engineering those interfaces?

Sadly no, we are a small association and we don't have time to raise money for this.

In the mean time, if I were setting up a school, as a matter of principle I wouldn't give the students tablets.

You mean: even if those tablets were running only free softwares?

If so, I'm curious to understand why.

After the six months I spent working for OLPC as a contractor (helping with the OLPC deployment in Haiti, which never happened btw), I'm also skeptical about using laptops and tablets in schools.

But our goal with Sugarizer is to convince the ministry of education it should encourages (and perhaps fund) free softwares in education, because the use and the design of these software is a collaborative work, bringing developers and teachers around the same table to find the relevant questions and answers about what students need.

...

Thanks for the information. As far as I understand, all single board computers have at least some flaws.

That is true. There are peripherals that need nonfree software. But your project does not use all the peripherals of that computer -- perhaps there is a model that can do _what your project needs_ without nonfree software.

This is one of our goal, but not our top priority for now.

Hi Dr. Stallman,

On Mon, Jan 9, 2017 at 12:39 PM, Richard Stallman rms@gnu.org wrote:

I see two freedom concerns, in the project description.

1. The tablet.

I believe that you are referring to Bastien's email here, which refers to his project and sugarizer. My understanding was that the tablets were only being used for their web browsers, and not installing software onto them.

2. The Raspberry Pi.

Okay, but this aspect is not directly related to the Edu-FLS aspect. It would be great if an FSF-endorsed campaign could not only endorse the Edu-FLS aspect, but also the device used at home (collectively referred to as the credit-meter), in which case we can find alternatives.

I am personally familiar with the DPM EBox-3350MX device, which is i486 architecture, and a diskless/fanless unit which also runs from an SD card. The first version of the project being proposed used this just fine. The only reason that I changed to the Raspberry-Pi is because the 3350MX costs $130.

-Charles

The raspberry pi has also sold 11 million units, granted they are not all being used as devices to teach kids coding as there are also units being used to control devices, cameras to capture wildlife photos, media (kodi) clients for example, but that is one big user base.

I think the fact that it has non free hardware blobs, which is more to do with Broadcom then the foundation. is a problem generally,

Perhaps we can add or create activities to teach about the 4 freedoms.

Paul

I am personally familiar with the DPM EBox-3350MX device, which is i486 architecture, and a diskless/fanless unit which also runs from an SD card. The first version of the project being proposed used this just fine. The only reason that I changed to the Raspberry-Pi is because the 3350MX costs $130.

-Charles

---

Discussion mailing list Discussion@lists.fsfe.org https://lists.fsfe.org/mailman/listinfo/discussion

On Wednesday 11. January 2017 10.19.35 Paul Sutton wrote:

Hi Dr Stallman

Just had a look at this list, so basically there does not seem to be a full solution out there that does not require some sort of non free software to get it working.

What about mentioning Arduino in that list, Granted it is not a single board computer, but if people want solutions to a hardware project and want it to be as free as possible this could be an option, of course if it fits in to being truly free, then it can be recognized as such.

Paul

...

See fsf.org/resources/hw/single-board-computers for information.

On this topic, the FSF is considering the EOMA68-A20 card in its "Libre Tea" form for RYF certification:

https://www.fsf.org/blogs/licensing/support-the-libre-tea-computer-card-a- candidate-for-respects-your-freedom-certification

More information here:

https://www.crowdsupply.com/eoma68/micro-desktop/updates/fsf-ryf-background

As this latter article notes, the plan had been to produce the EOMA68-jz4775 card instead because the SoC involved (the Ingenic jz4775) doesn't need any non-free software to function completely. However, it was determined that if the Allwinner A20 used in the EOMA68-A20 doesn't expose its GPU, doesn't require GPU support for productive use, and thus doesn't need non-free software to function, it could potentially be RYF-certified instead.

Some of this deliberation occurred in the context of the availability of FSF- endorsed operating system distributions. Various "libre" GNU/Linux distributions actively support ARM but not MIPS in recent versions, with the latter having been a focus in the past because the Lemote Yeeloong got a certain amount of endorsement:

https://blogs.gentoo.org/blueness/2014/07/02/continued-support-for-the-lemot... yeeloong-gentoo-mips-is-alive-and-well/

But there's something of a chicken and egg problem with MIPS systems and libre distributions, currently, in that the distributions won't support an architecture without any currently-produced (and easily-purchased) devices that can be used without non-free software. This is also true of emerging architectures such as RISC-V, but those architectures need to work towards a basic level of adoption first, anyway.

Meanwhile, there's no particular incentive in producing devices for such distributions without any particular interest from those distributions themselves, particularly when those distributions can find other architectures and systems to support, even when the systems being supported stand on somewhat uncertain foundations, as shown by that FSF summary. However, the situation around SoC vendors, documentation and suitability for Free Software is improving.

So, I think that the situation is getting better, but it would be nice if more cooperation between different Free Software interests occurred. Libre distributions should be able to support more than two architectures, particularly when their upstreams (for example, Debian) are providing support for far more than that. And in doing so, they would at least encourage architectural diversity in the hardware realm.

Paul

P.S. I do wonder whether the reasoning around RYF endorsement for the A20 could be used with the MIPS Creator CI20, which uses the jz4780 that is unfortunately encumbered with the proprietary and generally hostile PowerVR technologies, simply by disabling the GPU. However, I don't know enough about the jz4780 to be certain that doing this wouldn't disable all display support and severely limit the product's functionality.

Hi Charles and all,

the Kalulu contribution for the XPRIZE Global Learning challenge has been published, you can clone it from here:

https://dev.kalulu-education.org/cyprien/kalulu-the-gathering

or

https://github.com/Manzalab/kalulu (a mirror repository.)

Then test and/or fork it at will.

Have a good week!

Hi Hilaire,

Fernandes hilaire@drgeo.eu writes:

What is Kalulu?

Kalulu is the name of the free software that has been developed by a company, Manzalab, for the Global Learning XPRIZE challenge.

The main website for Kalulu is this one:

https://www.kalulu-education.org

The software is based on cognitive research and currently available both in english and swahili. It provides activities to let children learn the basics on how to read, write and count.

You can download/install/test Kalulu from these repositories:

https://dev.kalulu-education.org/cyprien/kalulu-the-gathering https://github.com/Manzalab/kalulu

You can see a quick screencast on how it looks like so far:

https://vimeo.com/199333164/

Best,