EMAN 2.1 beta-4 Release Date

2014-07-18

Summary

This the fourth beta release, a minor update to beta3. All of the major problems with the alpha releases have been fixed, and this release should be a fully usable replacement for EMAN2.0. We now strongly encourage ALL users to migrate. There are many many advantages to EMAN2.1 over EMAN2.0 (I was tempted to call it 3.0). Dropping BDB in favor of (Chimera compatible) HDF5 files and text files (JSON format) has made the whole package much easier and more stable.

Description

SPARX and EMAN2.1 are released together, as always. Both should be fully functional in source and binary releases.

If you experience problems with gold standard FSC curves that rise again at high resolution, please try the nightly snapshot, or use the --m3dold option in e2refine_easy.

Important performance note: The binaries are compiled for maximum compatibility, not performance. On a modern Intel processor, the 64 bit binary may run 30-40% slower than the exact same version compiled from source with ENABLE_OPTIMIZE_MACHINE. Using the 32 bit Windows binaries on the exact same computer is over 60% slower. That is, the same refinement may run up to 2.5x faster on linux than using the 32 bit windows binaries. The windows builds are provided mainly for interactive programs, not large scale refinement runs.

Mac binaries are also 10.9 compatible.

Version changes

While there are still some significant changes we plan for the near future, like a revamp of the particle picking interface, the system is now becoming quite stable again. There were relatively few changes in this release.

• A number of minor improvements in the single particle tomography programs
• Many small bug fixes

Please note that we are still in the middle of building a new interface for more convenient manipulation of MRCS files. Previously they had to be processed with e2proc2d.py and the --*threed* options. Now, files with a .mrcs extension will be directly treated as stacks of 2-D particles. Support is not yet complete, though, and may lead to some crashing. Please report any problems.

We believe we have fixed the bug with e2display on some non-US computers, but if it still fails, you may need to set these environment variables:

• LC_CTYPE=en_US.utf8
• LC_ALL=en_US.utf8