Impressions from SugarCamp#3

Talking about data at SugarCamp#3

Talking about data at SugarCamp#3

The third edition of SugarCamp in Paris organised by OLPC France this last week-end, on the 12th and 13th, was a event rich of discussions (everyone is very passionate about what he does !), information sharing and ideas debated. Something not surprising considering the recent news related to OLPC is that most of the discussions were spent over the future of the couple OLPC/Sugar/XO and the importance, or need, of seeing these three names stick together. Most of what was discussed deserve deeper and longer thinking but there are some highlights that already stick to my mind:

  • Sugar has active users and already contributed to many success stories. As highlighted by Walter, 30% of the patches in the latest version of Sugar comes from end-users (kids!) and 10% of activities are also provided by them. It can be observed that they take appropriation of the tools given to them in order to develop the things they need. Rather than the traditional carrot and stick motivational model, it is here the combination of autonomy, mastery and sense of purpose that drives these kids to do something. They get satisfied of doing something that solves a problem they had and are very proud of showing it around. Furthermore does this not necessarily happen only in “big” deployments, many the “micro” deployments with only one or two classrooms also exhibit this kind of results. This is the kind of story that need to be shared widely to convey the interest of learning differently and show the meaningful role Sugar can play in it.
  • We need to construct evidence of the effectiveness of Sugar.  The hype associated to distributing funny looking computers to deprived children and see the magic happen is mostly gone. After almost ten years into the project and 2 millions laptop deployed, deciders expect to see concrete proofs of the learning progress of the young learners. Statistics become an hot topic, it is important to gather data and derive some meaningful information from it. This will also contribute to the self-assessment capabilities offered by Sugar which up to now only consist in looking at the content of the journal.  It won’t be easy though as the massive processing of the journal content raises some privacy issues. Gathering dedicated information will also bring its set of specific challenges (define target usages, get rights, define data archiving policies, …). Progress will have to be made on that front but in a very cautious way.
  • Sugar will no longer be the single Gtk+Python+Fedora couple it used to be. There was already ongoing work on using Web technologies to develop activities but the result still depended on the traditional Python-based Sugar. Sugarizer is bringing a new perspective to this acting as the “missing link” making it possible to run these Web activities on anything that can run a Web browser with decent HTML5 support.  Next to this, there is Bruno Coudoin who is happily experimenting with the couple QML+Javascript to re-think GCompris and from which inspiration could be drawn for a new Sugar. There are also the Web technologies centric operating systems WebOS and Boot2Gecko that are seducing as an alternative to the Fedora+X stack. They could even be the only way to keep the XO-1 (the majority of the machines deployed) alive now that Fedora is switching his focus onto the next generation of hardware. All in all it seems Web technologies is the future of Sugar, either natively or within some kind of containers (for instance to run on Android). This is a rather good news which nonetheless has several big issues: 1) What do we do for all the active community who learnt Python ? Will they have all to switch to Javascript/HTML/CSS ? 2) Will it be possible to develop and debug Web activities on the XO directly ? and 3) How will we deal with the fact that Javascript is very challenging to use without a good set of libraries ? All these questions will better come with good answers before Sugar goes to far down that line. It is also possible that several implementation of Sugar will exist, either as a temporary or stable situation. Say, one running on Android phones and another one for B2G devices. This will call for a canonical definition of what Sugar exactly means in term of learning features so that the different implementations of it can be compared.

It was also interesting to see more of non-XO/Sugar related education talks this year, this enlarge a bit more the scope of the community and is rather relevant because what we are concerned with is an education project after all 😉


I am a researcher mainly interested in : architectures for publishing, consuming and preserving Linked Open Data in low-resource contexts; complex systems; education; data visualisation; video games

Tagged with: , , , , , , , , , ,
Posted in Events
One comment on “Impressions from SugarCamp#3
  1. […] Per Child, de l’utopie à la réalité” echoes nicely with the comments I made about SugarCamp#3 on this blog. It recalls the history of OLPC and the link to its community. With a mitigated output so far and […]

Comments are closed.

%d bloggers like this: