Hi Markus - I have integrated the GCR and G64 patches. I left the GEOS patching D64 to G64 since that really should be a special forked tool you manage if that is OK.
That's ok.
The 2nd patch leaves one bug (no new bug): If the following conditions all are met, then parsing that sector fails:
- Sync length is shorter than usual, just 10 - 16 bits.
- Header sync is starting at end of track in image with 8 bits or less.
- Header sync is therefore continued at start of track, again with 8 bits or less.
Therefore the sync is not recognised neither at the end of the track nor at its start.
Well, these conditions are very unlikely to occur, but there is a possible problem. That's the "FIXME"-Comment in the source code (with a typo in the german text after it).