3 projects I’d like to see happen

• Genealogy

An online collaborative genealogy tree that would connect all the persons who live or have lived. With photos, biography, maps…  The best trial is probably werelate.org.

• History

A Google Earth based collaborative historical atlas with links to sources, illustrations and videos. Should include various maps like frontiers, demography, battles…  Something like that :

• Science

A platform for online collaborative science that would allow everyone to propose a project or to join one. Should include wiki, Tex, peer-review system, bug report, photo gallery, etc. I already discussed this here. Colabscience.com could be a good starting point.

 

Open black block

I am currently studying a computational method proposed 3 years ago in an IEEE paper. The authors have evaluated the efficiency of this method by implementing it under Matlab/Simulink and by comparing the results with the Finite Element Method.

I now need to implement this method for my research. But although the mathematical description is complete, the Matlab/Simulink implementation description shows “black blocks”. Actually these “black blocks” are the key of the implementation. So I will probably need about one month to implement this method correctly if I have to figure out how to code these blocks.

I asked my Professor if it was possible to obtain the simulation file by contacting the authors. Here is his answer:

If the file is not available on Internet, it means that this is important for their research to keep it secret. If you want this file, we need to contact them and to sign a research agreement. It would be better if you can implement this method by yourself. 

At the time I just responded that this is an inefficient way to do research. I really don’t get why if someone publish a paper, the simulation file associated is not made publicly available too. When publishing, a scientist gets credit for all the work presented.  This is just not fair to keep something secret and to publish another paper using it, and to get credit another time for the same work.

If you want to keep something secret because it’s important for your research, don’t communicate about it ! Finish your research and then publish !

I should mention that the government is financing my PhD. It means that next month, taxpayers will pay for me to do something that as already been done. Besides the government will probably finance someone else in the future to do exactly the same work… Research could be more efficient by asking journals to force authors to make their data/simulation/program publicly available when publishing a paper. In the meantime, you can act: be an Open Scientist.



Bringing physics into massive colaborative science

Several months ago I tried to develop on this blog the idea of “Open Science”. In particular I underlined the fact that, to build an effective collaboration over the internet, we need new tools and guidelines [1]. The Open Science Summit 2010 seems to have speed up the whole process.

On the one hand, some days after the conference, a brainstorming about the definition of “Open Scientist” was lunched by Jessy Cowan-Sharp. The draft of the discussion is available for consultation and editable by everyone via EtherPad here [2].

On the other hand, Casey Stark @caseywstark and DJ Strouse @djstrouse lunched for the occasion, CoLab, a platform designed for open and massively collaborative science,

We want to make it stupid easy to center a discussion around protocols, data, plots, published papers, papers in progress, simulations, code, or any other component of scientific research, [...] to import a published paper and collaboratively highlight and annotate (it) in-line, [...] to host a working version of my paper online, collaboratively edit it [...]. In short, as a scientist, I should be able to easily and openly discuss any piece of my science with my entire scientific community.

The three first issues discussed on Colab deal with social science (local optimal scientific research environment, social network for science) and mathematics (P != NP). Now I had like to see physics, chemistry and open hardware projects enter the arena. And especially projects linked with global issues like drinking water, malnutrition, diseases, safe housing, rural electrification, reduction of CO2 emissions and clean energy.

But how should we tackle this kind of problems ?

Consider an open problem and various way to address it : technology A, technology B, law C, etc. Now, should we focuse on technology A as it seems the most promising or to the contrary consider at the same time all the propositions ? If the last approach is workable for “non-material” Polymath Projects for example, is it realistic to lunch at the same time 20 different experiments to develop 20 competing technologies ? Who is going to finance it ? Theses questions should be considered carefully before proposing an issue to avoid having the project at a dead end after some discussions.

In practical terms, I would recommend to propose (at least at the begining) issues whose solutions can be formulated in theoretical terms or require only small investment. Ideally everyone, not only laboratories, should be able to set up the required device/experiment and work on it. Any ideas ?

Notes

1. Open recherche (french)
2. Moi quand j’serai grand, j’serai open scientist ! (french)