Annual Report 2011 Max Planck Institute for Astronomy

components. The owner of the ticket is in charge of
the underlying task. Additional information, such as
due dates or dependencies on other tickets can be reflected.
All information is kept in a database. A simple
interface allows to query and filter the tickets and,
by that, to monitor the progress of the project or to
identify open issues for upcoming milestones.
• The version control system. Each trac project is associated
with a subversion repository. For software
development projects it is the place to store the code,
but it can also be used to archive and control versions
of documents.
• Seamless interaction. Within trac it is very simple
to introduce references to tickets, wiki pages, files in
the repository, blog entries etc. All content within
trac is also searchable.
• Expandability. trac capabilities can easily be expanded
by adding plugins. There is an active community
developing a wide variety of plugins, such as the
blogging system, user account management or Latex
support in wiki pages. trac and its plugins are developed
in the Python programming language; with
reasonable effort it is possible to develop custom solutions
needed for a project.
trac and Linc-nirvana
Linc-nirvana introduced trac in 2010 and uses it since
with great success. More than 70 team members and reviewers
from inside and outside of MPIA have access to
Linc-nirvana’s trac.
The project uses tickets to organize and follow up
most development activities. Each team member involved
in an activity gets informed by email about updates.
The tickets increase the efficiency of status meetings
– they help to stay focused and provide all important
information on its subject in a commonly accessible
1 Push-Pull
2 Push-Pulls
4 Push-Pulls
8 Push-Pulls
IV.4 Special Developments in the Technical Departments 81
place. The tickets have demonstrated to be very helpful
when collaborating with partners in different institutes.
All project related documentation is archived in trac
and is easily accessible in various places on the wiki,
depending on their context. So are meeting minutes and
notes, which are stored as blog posts.
Thomas Bertram, Florian Briegel
Interaction Matrix Calibration for Adaptive Optics:
What is the best method?
For an adaptive optics (AO) system it is necessary to
characterize how the deformable mirror (DM) interacts
with the complete optical system up to the wavefront
sensor. This procedure, commonly referred to as “calibrating
the interaction matrix” (a matrix which is then inverted
to produce the matrix for reconstructing the wavefront)
is a process which is a key to achieve an optimal
control of the DM. Members of the technical department
at MPIA, working together with graduate student Xianyu
Zhang and colleagues from INAF, have determined an
optimal method for performing this calibration. These
results appear in the article “Calibrating the interaction
matrix for the Linc-nirvana high layer wavefont sensor”
which was recently published in the journal Optics
Express.
Traditionally, interaction matrices have been calibrated
by commanding the DM to a sequence of well-defined
shapes, and, at each step in the sequence, taking
a measurement of the light pattern seen by the sensor.
But this process is affected by various sources of noise.
Fig. IV.5.3: See text for details
4 Frames
3 Frames
2 Frames
200
150
100
50
0
1 Frame
Credit: MPIA