[Stellenangebote] Doktorandenstipendium "Supporting Scientific Software Engineering", FU Berlin, AG Prechelt

[Lutz Prechelt]

Liebe Bald-MasterabsolventInnen,

bis zum 20. Oktober kann man sich noch bewerben beim 
Helmholtz-Kolleg GeoSim für ein dreijähriges Stipendium.

Die Prozedur ist beschrieben auf
http://www.geo-x.net/geosim/admission.html

Zu den Inhalten siehe die Beschreibung auf
http://www.drarbeit.de/ (suchen nach "Prechelt")
oder hier:


Dear would-be Ph.D. student in empirical software engineering,

the Helmholtz Kolleg GeoSim
performs research on explorative simulation in the geo sciences
(earth systems, water systems, atmospheric and climate systems)
and is offering 8 3-year positions for Ph.D. students.

http://www.geo-x.net/geosim

All of the slots in this Ph.D. program are for either
earth scientists or mathematicians, except one.
This one (hardly mentioned in the above materials at all)
would be in empirical software engineering (with me) and
aims at helping the scientific computing specialists to
identify and adopt suitable software engineering practices
and processes.

More concretely, the target people would primarily be
Ph.D. students in Mathematics (with Prof. Rupert Klein and
Prof. Ralf Kornhuber, both working next door) and the target
software processes could be any that they are not yet using but
that might help them and that they are able and willing to adopt.

Your research would be about this willingness and ability of
the mathematicians and helpfulness of the processes.

Candidate processes could be for instance
- from the realm of Agile Methods such as
  automated unit testing, test driven design, etc.;
- review and inspection techniques;
- suitable component structures and architectures;
- standards and practices around configuration management,
  build and release management, general project management,
  and organizational learning;
- suitable testing techniques
  (as these are simulations there is often no specification
   from which one could derive correct test results for any
   but the most trivial cases!).

During your research, you would need to
- find candidate techniques,
- evaluate their suitability,
- then their introducibility,
- then introduce them,
- then measure their effect
- all in the concrete practical setting of other people's
scientific work.

Difficult! 

It requires good power of judgment and good people skills.
A background knowledge of 
- differential equations and numerical computing
  (that is, the application domain) and
- many software engineering techniques
would be highly valuable, but neither is strictly required.

Do you think you would like to compete for this position?

If you want to pursue this, I suggest the following procedure:
0.	You check out the next application deadline on the GeoSim web site
(do not trust the information given in the present ad!)
1.	You read up some background
2.	You produce some preliminary ideas
3.	You talk to me about those
4.	You talk to the group of Rupert Klein
5.	You write your proposal (which needs to include specific 
    research directions in order to have a good chance 
    of acceptance).
6.	You submit your application materials along with your proposal.

Questions are welcome!

  Lutz Prechelt

Prof. Dr. Lutz Prechelt; prechelt at inf.fu-berlin.de
Institut f. Informatik; Freie Universitaet Berlin
Takustr. 9, R.014; 14195 Berlin; Germany
+49 30 838 75115; http://www.inf.fu-berlin.de/w/SE/