EFOSC2 slit rotation problem

During EFOSC2 night 2012-10-21 the skylines were not aligned along the CCD pixel rows, at least for Gr#11. The image below shows EFOSC.2012-10-21T06:13:01.202.fits. 

PESSTO pipeline implications :this causes a problem for checking the wavelength calibration using the sky lines, as the extracted sky lines appear broader due to the rotation. As a fix, we are going to reduce these with PESSTOEFOSC2dSPEC, and reject the offset with the sky lines. Then, we will measure the shift manually, and apply it via specshift in iraf to the _t frames, before we run PESSTOEFOSC1dSPEC.