Version 1.9

Version 1.9.2, 02 Nov 2022

    • Modified to recognize new data file header flag CXHF_ST_2GOAL (Maestro v4.2.0 added support for two alternate goal targets in a trial's searchTask operation). Data file version incremented to 24, but no actual changes in data file format.

Version 1.9.1, 07 Jun 2021

    • Modified to support 200 (instead of 50) "sorted spike train" channels in the Maestro data file. Nowadays, researchers generally record multiple neural units using the Plexon or Omniplex system external to Maestro and perform "spike sorting" analysis offline to isolate the spike train of each unit. Some use editcxdata() to store the neural responses with the original Maestro data file containing the behavioral responses, trial/target definitions, and other information. The spike trains are stored by appending records to the data file. Originally, there was support for 13 distinct spike sort channels, identified by ID 8..20 in the first byte of the record (to distinguish them from other record types). Circa 2013 (v1.6), this was increased to 50, with spike sort channel IDs 8..57. With this change, JMWork can read/write data file records with sorted spike train times on any of 200 channels. To do so, we keep the same channel IDs 8..57 in the first byte of the record, but we also specify a bank number 0..3 in the second byte (prior to this change, that byte was always 0).

Version 1.9.0, 05 Mar 2020 (rev 19 Mar)

    • No functional changes or bug fixes in this release.

    • Java 8 runtime environment now required. This release was compiled against a Java 8 SDK. As such, it requires a Java 8 or better runtime environment. Java 7 and earlier VMs are very out-of-date, are no longer supported, and are considered insecure. Users are strongly encouraged to upgrade their machine to the latest update of the Java 8 VM. We have verified that JMWork runs on both a Java 8 and a Java 11 VM under Windows and macOS. Note that Java 8 and 11 are both long-term support releases, with planned support through at least 2023 and 2024, respectively.

    • Better desktop integration for Windows. JMWork is now distributed via a self-extracting installer. The application itself is packaged in a WinRun4J-based EXE wrapper (JMWork.exe) that locates a suitable Java VM on the host machine and launches the application with that VM.

    • (Revised 19 Mar 2020) Minor bug fixes: (1) Changed the keyboard accelerators for the check box menu items in the View menu. The function-key accelerators are reserved in macOS, and using the Command key as a modifier is problematic with a check box menu item due to an internal JDK bug). Now the View menu items use the Shift key as the modifier. (2) Fixed bug in computation of trial target trajectories when velocity stabilization is enabled with a sliding-window average.