TODO (ComputeFStatistic-related)
-----------------------------------
* clean up exit codes (BOINC)

* test/compare various metrics
    + find/fix problem with Ephemeris-metric

* get rid of year-specification on command-line if possible (use ranges 00-04..)

* work on more general tilings

* speed up (interpolation vs oversampling?)

* modify ComputeFStat so that the search parameters can be referred to
  any arbitrary epoch and not just to the epoch of the first sample of the
  analyzed data set (as is now).

* determine detector from SFTs? (->yes, in SFT-v2)

* prepare 'unified' F-stat code for binary+isolated (->CFS_v2)

done
----
* use Coherent metric
* deallocate grid-memory
* print grid + metric infos for debugging
* find round-off errors? (-> gcc O2, 1e-5, seems known)
* use config-file
* activate cvs-log for lalapps/pulsar
* write metric wrapper for Ptole and Coherent
* adapt config-file more to python config-syntax
* specify sky-regions by polygon 
* implement new user-variable handling
* get rid of "coordinate-cut" limitations... (seems to work so far)
* clarify priorities between specifying Alpha+AlphaBand.. and a skyRegion!! (done: mutually exclusive)
* complete use of Ephemeris-metric
* commit stripped-down version of "test-SFTs" to allow everybody running testFstat-script
* quick-fix for mesh-covering problem at the equator?

* move to using SFT-IO lib (done in v2)