New version from Andrea, with some bug fixes.
Same code, but rebuilt, linking against version v4r52 of the Fr package. This fixes the problem encountered by a few people, that ascii2frame produced bad frames (indicating the wrong byte ordering) when run on little-endian (Linux) machines.
Also fixed a few mistakes in the usage information for ascii2frame. In particular, there is actually no "-o" option, despite the fact that such an option was listed when you typed "ascii2frame -h".
Changed code to write out the azimuth values in radians rather than degrees, as specified in version 4 of the frame specification.
Fixed degreeToDMS() to calculate degrees/minutes/seconds correctly (thanks to John Whelan for finding the bug).
Added more digits of precision to many detector geometry values, using the values listed in lsd.pdf from LAL 0.10 .
Fixed azimuth values for bar detectors. (Had been expressed in degrees east of north, but the frame spec needs the angle north of east.)