Ask questions to the candidates running for the GNOME Foundation elections

Ok. I was the original anonymous comment poster, but I got an account to clarify it. This was not at all meant as a troll but very serious.

As often, we tend to ask questions, we'd love to provide the answer ourself, so here is my point.

Here is the introduction to a research paper that is not yet written:

We sometime should remember that Computer Science is the science of information and thus everything should be done to ease the treatment of data.

In today's operating systems programs and data are stored within filesystems.

Among the programs available, database management systems were developped to empower data treatments. Today DBMS such as postgresql can handle complex queries, transaction support, etc. It is then quite surprising to see how underused DBMS are in our operating systems:

should the Free Software community wait until Microsoft does it ?

Today, both system and user data are stored as sparse files, with naturally every ``competing'' applications having it's own file for the same data (think for instance to bookmarks in two different browsersfootnote{I know that there are some tools to convert from one format to another, but it's neither easy to do on a regular basis in both ways (and what if three browsers...) nor ready to work with both browsers running at the same time, which is a legimate idea though}).

When comes the time of backup, the sparseness of data often forces to backup the whole disk.

We propose in this paper our vision which might provide a better integration of free software. Applying these concepts would require a little integration work from almost all participating projects, which would be needed to achieve the ``all data in databasis'' paradigm.

We will see that there exists two kind of data:

egin{itemize}

item a SYSTEM-related data: information about packages installed, and for each package system-wide configuration files;

item a USER-related data: all the data of a given-user, including some strongly-organized data such as mails, bookmarks, contact informations, xml documents (text processing, spreadsheet, etc. among them).

end{itemize}

Think for instance for the part of code in evolution which is redundant with (and less general that) postgresl...

I'll try to setup a website, since I got a working implementation (not specifical to gnome, though GUIs are done with gtk) as soon as got time.