If it helps at all, one of my postdocs is running the 64-bit Mac version and he says there is no problem there... /Derek
On Mar 23, 2010, week13, at 1:42, Paul Emsley wrote: > Alice Dawson wrote: >> I'm running Coot 0.6.1 on a Mac (running 10.5.8), installed via Fink. >> >> I'm running some Shelx refinements. When I try to save the coordinates >> (as a .ins, from a .res that I read in) I get this error >> >> (first-coords-imol) >> (first-unsaved-coords-imol) >> INFO:: Setting fileselection with file: benz32_bind-coot-0.ins >> save coordinates for molecule 0 to file /Users/axdawson/work/ptr1/ >> Coord_submission/pemet/benz32_bind-coot-0.ins >> /sw/bin/coot: line 5: 26481 Bus error /sw/bin/coot-real >> "$@" >> >> This happens even when I make no changes to the coordinates before >> trying to save. It doesn't happen with pdb format coordinates. Any >> ideas? >> > > Dear Alice, > > I think that this is the same problem that Derek Logan complained about > recently. I did investigate the problem and it seemed to result from > setting the precision or other output format specifiers for the floating > point numbers in the .ins file. I don't know why Coot on a Mac should > do this - I suspect a Mac problem rather than a Coot problem. > > I have patched things up in the pre-release to not crash on writing the > .ins file - but that doesn't help a great deal because the resulting > .ins file is empty. > > Real solution still pending. I'm hoping the Bill and I can find some > time to bash on it soon. > > Paul.