Student Application Template

• Name: José Luis Sanroma Tato

• Contact/Email: email: josel.sanromatato@gmail.com

• Background: I am MSc candidate in computer Science (I expect to present my Maste thesis in June) at the University of Castilla-La Mancha, Libre Software enthusiastic and FSF member 12590. Since 2012 I am research assistant in the ARCO Research group (https://arco.esi.uclm.es/en). My work in the research group was focused on building a highly scalable and opportunistic architecture to build Debian packages for different desktop and embedded systems, automatically taking care of the dependencies. I built some parts of my project as isolated environments using virtual machines to allow more flexible resource and task management.

  • This work is part of the VIII Libre Software University competition in Spain and is licensed under GPLv3. Project website

    The repository of my thesis is public here. I don't have a debian package of this work yet. I am fluent in C++, C, bourne shell, Python, Java, Latex, Make and also networking and distributed systems. In my relationship with Libre Software, I have been usin Debian for more than 6 years and I have organized two install parties to help the new students to install Debian at the university where more than 40 people attended. I have also reported some bugs to Debian. Besides, my Master Thesis

• Project title: Opportunistic and distributed infrastructure to build Debian packages into a web of trust.

• Project details: Building software on shared machines increase the risk.

  • Debian itself has a web of trust in which all the Debian Developers are part of a trusted network. A Debian Developer is already trusted to build software because they upload the packages to Debian and their GPG key in in the Debian keyring. Building an opportunistic architecture in which we know what trusted computers are on-line, Debian developers may donate CPU cycles in order to build debian packages in their machines.

• Synopsis: Building packages on shared machines increases the risk. This project tries to solve the problem using opportunistic computation to build and infrastructure to build Debian packages into a web of trust.

• Benefits to Debian: Distributed and opportunistic infrastructure to build Debian package.

• Project schedule: This work will be directed by three iterations divided in the three moths of the coding phase

  • Iteration 0:
    • Study the problem and the possible solutions that could be applied to finally choose the one thats suits better the requirements.
    • Understand everything with GPG keys.
    • See what parts of my Master Thesis are useful here and use it.
    • First increment of the project with something running
  • Iteration 1:

    • Build the opportunistic system, things such as Ice middleware could be useful here and more concretely IceGrid

    • Work with GPG key to provide a web of trust. Simulating an example organization.
  • Iteration 2:
    • Last changes to finish que work
    • Bug fixing
    • Write Documentation

• Exams and other commitments: Master thesis presentation on June the 20th.

• Other summer plans: Some trips in sporadic weekends.

• Why Debian?: Since 2007 I am using Debian as a main operating system. Debian is the GNU distribution I agree the most. Most of the packages we use nowadays are made for Debian. I always wanted to do something to collaborate with the Libre Software and Debian in some return way in general.

• Are you applying for other projects in SoC? No.