Highlights of What’s New in Version 28.1.0, 23 July 2015

Here is a summary of what’s changed since release 28.0:

The ROTATE element was not affecting the floor coordinates. This was found and fixed by A. Xiao (APS).
The END_PASS parameter on SCATTER now works as expect, after removal of a one-pass offset.

A bug was fixed that caused a crash when a 1-line aperture search was performed. This was reported by Guohui Wei (JLab).

The TFBDRIVER element now has the ability to measure the beam phase for use in longitudinal feedback. Previously, only momentum-based input was available for longitudinal feedback.

The ramp_elements and modulate_elements commands now have the ability to write a record of their output values.
The run_setup command now has options, intended primarily for developers, to turn on memory usage and executing time monitoring during tracking.
The units given for loss rate the output files from touschek_scatter were incorrect and were fixed. Results were not affected. (A. Xiao, ANL)
The tune_footprint command was improved in several ways. It is now possible to ignore half-integer resonances. The upper and lower bounds of the chromatic tune footprints are now available for optimization. It’s now possible to turn off either chromatic or amplitude tune footprint deterimination.
The optimization_setup command allows suppressing particle tracking in order to improve performance in some unusual cases.
The correct_tunes command can now utilize any element that has the K1 parameter.
The chromaticity command can now utilize any element that has the K2 parameter.

Fixed a bug that affected tracking when orbit correction was used, start_from_centroid=1, and particle distribution was not random across processors.
Warnings about ρ > 10⁶ m are now issued by the parallel version, as for the serial version.
Memory usage logging to WATCH output files now sums the memory across all cores, rather than just the master core.
A memory leak was fixed in the ZTRANSVERSE element that sometimes caused the program to crash. This was reported by R. Lindberg (ANL).
The output of the beam charge in the ZLONGIT wake output file was corrected; previously, it only showed the charge on one core.
The frequency_map command now provides an estimate of the time needed to complete.