Should we use license for data or databases generated in the openMicroanalysis project?

[Hendrix Demers]

After thinking about which license the project should use, I ask myself that question: Should we use license for dataset or databases generated in the openMicroanalysis project?

Data I had in mind was maybe a database of k-ratio from various publications or phirhoz distribution obtained by Monte Carlo simulation. Building the database can be a long process and maybe the author when to specified how the data can be used.

I read Robert L. Price Editorial in Microanalysis & Microanalysis (Microsc. Microanal. 20, 996–998, 2014) touching on data repositories and check the different projects mentioned. My finding is

• Dryad use Creative Commons Zero (CC0) which is define as OpenData
  
• Material Project seem to use no license, but ask to cite the project.
• figshare use different license (Creative Commons, MIT, GPL, Apache License, etc) depending of the content (figures, datasets, code) see their Copyright and Licensing page.

Creative Commons licenses seem to cover all possibility for data publishing/sharing. I think we should have a license associate with each data or database generated by this project so the user will clearly know what he can do with the data. The main author of the dataset or database should free to choose the license he prefer.

What do you think?

Hendrix

[Eric Jensen]

Since the project aims to be as open as possible the CC0 license or no license with citation sounds simplest and simplicity will attract more users.

[donovan]

Since my Matrix.mdb Penfluor/Fanal database of Monte-Carlo k-ratios is already distributed with my free CalcZAF/Standard distribution, the "cat's out of the bag", at least for me. But the more open the better I guess, though anyone with any integrity should at least cite their database sources I would think.

As for Matrix.mdb, we have over 200K k-ratios calculated as of a few months ago. And we are getting ready for a new update to the database with even greater coverage of the periodic table.  By the way, if any one is interested, the latest matrix.mdb can be downloaded from the latest CalcZAF distribution:

http://probesoftware.com/smf/index.php?topic=81.0

It is in Microsoft Access database format (for fast SQL queries).
john

On Thursday, June 4, 2015 at 1:37:49 PM UTC-7, Hendrix Demers wrote:

After thinking about which license the project should use, I ask myself that question: Should we use license for dataset or databases generated in the openMicroanalysis project?

Data I had in mind was maybe a database of k-ratio from various publications or phirhoz distribution obtained by Monte Carlo simulation. Building the database can be a long process and maybe the author when to specified how the data can be used.

[Hendrix Demers]

I don't think it is too late. The way I see it, it is to clarify what the user can do with the data. Also I was thinking more about stand alone data than data distributed with code, but I think we should apply a licence in both cases. You can describe it with the data or on the website, but also add a licence which make it clear what what the user can do. For example, I have use other people data from their program for my one code, but I need to ask them permission to distribute the data. If they have a licence, it would be more easier for both of us. 

I discover today to use all Creative Common licences, you have to cite the source. From the FAQ: "All CC licenses require users to attribute the creator of licensed material".

https://wiki.creativecommons.org/wiki/Frequently_Asked_Questions#How_do_I_properly_attribute_a_Creative_Commons_licensed_work.3F

On a side note, which version of Microsoft Access database format you are using? I have problem reading the old 97 version on windows 7 computer.

Hendrix

[donovan]

Hi Hendrix,
The matrix.mdb database is written in Access 2000 format so should still be readable by most apps.
john