Fixed the logic for reading in database files. The previous logic used the default database directory name which may cause a runtime error if a user defined database name was provided.

• Summary

Chainbreak Scoring: Enabled chainbreak scoring for carbohydrates NEWS: * Added cutpoint variants for sugars * Refactored ChainbreakEnergy - Made work with cutpoint variants that do not require 3 overlapping points - Removed duplicated code - Added lots of explannatory comments * Refactored add_cutpoint_variants() and add_single_cutpoint_variants() - Removed duplicated code - Enabled to work with sugars ---------------------------------------- Unit test status: Pass Integration test status: 25 rings: change in number of ResidueTypes reported inverse_rotamer_remodel, loop_creation, antibody_graft, SnugDock, hybridization, mr_protocols, antibody_legacy, antibody_H3_legacy, remodel_disulfides, remodel: cosmetic output changes swa_rna_gaga_23_append_and_kic_close, swa_rna_gaga_22_prepend_and_kic_close, swa_rna_gaga_10_append_and_ccd_close, swa_rna_gaga_18_rebuild_bulge: strange output of "10 BulgeApplier" in new (huh?) 10 other loop-related tests: trajectory changes (I don't THINK I broke anything here, as the unit tests for ChainbreakEnergy and all loop and closure unit tests passed, and I had another set of eyes double check my work; however, if you are concerned, please e-mail me. I did slightly change the logic of ChainbreakEnergy, but only in a way that should make a difference in results in a specific case I do not believe can ever happen with any ResidueTypes currently in Rosetta -- or ever likely in Rosetta. I am of course willing to revert if needed.)

• Summary

Compile fix for gcc 4.6.3 with warnings-as-errors; there is a pseudo-uninitizalized variable in a pilot app that makes that compiler grumpy, and was holding up the release script. Also added copyright header and vi-emacs settings header to that file. No test changes expected, as pilot app has no integration test.

• Summary

Minor bug fix. Compiled on windows but not linux.

• Summary

Minor modification to previous bug fix suggested by Seth Cooper.

• Summary

I accidentally removed the boost_1_55_0 directory when pushing fixes for the windows build. So sorry for that. I'm adding it back as a quick fix since I don't know how to properly revert a commit on the master and I want to make sure things are back to normal as soon as possible for everyone.

• Summary

Updates for windows port.

• Summary

Fix for access violations in the WIN32 BOINC build. No tests run; changes only affect WIN32 platform.

• Summary

add options to have PyMolObserver only update on energy on conformation changes. Options: -update_pymol_on_energy_changes_only -update_pymol_on_conformation_changes_only These should be passed along with usual show_simulation_in_pymol option. Note that by updating on energy changes only, pose length changes will not occasionally cause a segfault which sometimes happens when the observer kicks in and shouldn

• Summary

Fix a warning I was getting in antibody/metrics.hh due to an uninizialized templated struct. Fixed by making a default constructor that initializes templated members to 0, this should be fine as it seems usage of this struct is templating to primitives (eg. core::Real)

• Summary

Changing build settings for phenix builds. Updating option.py which fails on some python versions.

• Summary

Fixing gcc build which gives an error from the double-->float conversion in the center of an array initializer list in Frank's cryst_design.cc application.

• Summary

[minor update of SandwichFeatures] - typo corrected (@overviewprotein_tools/scripts/clean_pdb_new.py) [integration tests comment] - should not change

• Summary

[minor update of SandwichFeatures] - added a new column 'average distance between (two) sheets via central residues in edge strands' (prints until 1 floating point like 6.4) - added a new column 'minimum distance between (two) sheets via central residues in edge strands' (prints until 1 floating point like 6.4) - debugged for a case when strands are < 3 for topology identification like in '1A21' - default of exclude_sandwich_that_has_near_backbone_atoms_between_sheets_ = 'false' - default of exclude_sandwich_that_is_linked_w_same_direction_strand_ = 'false' [integration tests comment] - should not change other than features and features_database_schema (if any)

• Summary

pull from remote

• Summary

Bugfix for -missing_dens_to_jump

• Summary

Legend: