Release Notes for JMRI 3.11.4 release
Date: March 20, 2015
From: Bob Jacobsen
Subject: Test version 3.11.4 of JMRI/DecoderPro is available for download.
This is a test release. Please post a note if you encounter any new or old bugs! And please back up your JMRI files before installing this, in case you want to go back to an earlier version.
This is the next in a series of test versions. We expect this series to end in the next JMRI production release around the middle of 2015.
Some of the changes involved are quite extensive. They may require a certain amount of experience before they are working well. Therefore, this test version should be considered experimental.
Known problems with this releaseNone at release time.
New warnings for this version:
This release, and all later releases in this cycle, require Java 1.8 or higher. To use this or any later JMRI releases, you'll have to update the Java on your computer to Java 1.8 or higher.
JMRI 3.11.3, and therefore this version, removes some files that were present in earlier versions. Failure to remove these can result in some portions of JMRI (notably the web server) failing to function correctly. If you are updating an existing JMRI installation to this version on Linux, you must remove any older version of JMRI or unpack this version of JMRI in an alternate location. The recommended update procedures for JMRI on Microsoft Windows (run the installer) and Mac OS X (open disk image and drag to final location, e.g. Applications folder) remove/replace the older version of JMRI, ensuring that these obsolete files are not present.
This release uses JDOM 2.0.5, which requires some code changes if you are directly calling the parts of JMRI that interchange XML objects.
There are now separate settings for the default Service Mode Programmer and for the default Ops Mode Programmer. If you're using two or more system connections, this lets you use one connection for ops mode (on track) programming and another connection for your programming track. If you're using two separate connections, you should go to the Defaults pane on the Preferences window and check that those defaults were set properly.
If you are making a direct connection to a Intellibox unit via serial or USB, you will have to reselect your preferences. See the Uhlenbrock support page for instructions.
If you are using an NCE USB adapter, you will have to reselect your preferences the first time you use a JMRI 3.11 version (if you set them using JMRI 3.10 or earlier). See the page on configuring JMRI for NCE for instructions.
The Save button in the Preferences window now only prompts for restarts if required. There may be instances when pressing save incorrectly fails to prompt to restart JMRI or prompts to restart JMRI when it should not.
Because of internal code changes, this version of JMRI may not be compatible with CATS 2.34 or earlier. Keep using JMRI 3.10 with CATS until there's a CATS update that says it will work with JMRI 3.11.
The XmlIO web service has been removed. All users who were relying on the XmlIO API are recommended to begin using the JSON API. inControl web throttle users are recommended to begin using webThrottle instead, or invest in devices that can support webThrottle, the iOS WiThrottle, or the Android Engine Driver.
Please use these links for downloading:
Changes since test release 3.11.3:
ESU - ECOS
- Fixed an issue where the Loco Database from the Ecos wasn't always loaded correctly
Egbert Broerse improved the signaling system documentation
Balazs Racz: For single output signal head, whenever the underlying turnout changes (due to change in the turnout table, turnout feedback, or JMRI listening on the command bus and picking up a turnout change event), the signal head will adjust the displayed output. Also fixed a bug at JMRI startup: The single output signal head implementation was generating a turnout change command every time JMRI started up (at panel.xml load time).
Oscar Moutinho updated the DB-HV-1969 signals definition to improve speed control and icons
The signal speed definition has been updated for consistent operation. Also, a release test has been added to keep it that way.
New / Updated decoder definitions
Economy Serie 6: version (CV7) starts with 51 instead of 52 (Alain Le Marchand)
Doehler & Haass
Global update (Pierre Billon)
- Added: 1 new Decoder / 4 Fw versions (DH22A)
- Added: 3 new Fw version(s) (DH05C, DH06A, DH10C, DH12A, DH16A, DH18A, DH21A)
- Added: 1 new Fw version(s) (FH05A, SD18A, SD21A, SH10A, generic profile)
- Added: SUSI pane, moved relevant options there (All decoders)
- Added: Start/Brake/Shunt pane, moved relevant options there (All decoders)
- Change: Adv. mapping now in standard mapping table (All Decoders)
- Change: Railcom pane (grid + qualifier) (All Decoders)
- Change: Adapted Trix 66840 def to match above changes to the core D&H XMLs.
- Fixed: Consistent Reset function / translation (All Decoders)
- Fixed: Removed non applicable Motorola options (FH05A, DH05A, DH05B, DH10A, DH10B, DHP160, DHP250, DHP260)
- Info: New ProductID format for future versions (e.g.: DH05A_2014.10)
- Info: No new models created for new Fw on long deprecated models (DH05A/B, DH10A/B, DHP160/250/260). Use generic profile if needed.
Add DCCconcepts Zen series of decoders (Alain Le Marchand)
Peter MacKenzie added definitions for the Hattons MD4 and Gaugemaster Opti DCC25 decoders.
Dan Boudreau made the following changes to operations:
- New feature, you can now have the program assign a locomotive to a train based on the train's tonnage, route grades, and engine horsepower. See the operation's help "Optional Train Requirement" for more details.
- Changed the date and time format when logging rolling stock and trains to be more MS Excel compatible. Makes it easier to combine several Excel sheets and then sort by date and time.
- Export of cars and locomotives now includes the "Moves" count associated with them.
- Improved how the program deals with cars on a FIFO or LIFO track. Previously, cars on a FIFO or LIFO track were stuck behind cars that couldn't move. Program will now bypass a stuck car and process the next cars in the order defined by FIFO or LIFO. A car could be "temporary" stuck if the car's destination was full and couldn't accept the car.
- Improved how the "Edit Car" window blocking feature deals with cars that are defined as "Passenger" and also in a kernel. Now the program won't ask if all passenger cars should have the same blocking number if the car is also in a kernel
- Made improvements to the car router to allow routing through staging. This allows users to model industries that aren't on the layout. Feature is still disabled by default.
- Fixed a problem where a caboose or car with FRED would lose its departure track in in staging after reseting a train that also departed and returned to the same staging track.
- Fixed a problem where a car's final destination and custom load didn't reset properly when a car was departing staging and was assigned to staging that was full. Thanks to Dave Waraxa for finding and reporting the two last problems.
A Warrants panel is added to Preferences. Warrants can now interpret the Aspect Speed Map four ways: % of Normal, % of Full Throttle, Miles per Hour or Kilometers per hour. A global Throttle Factor to aid trains in running at scale track speed is added. Also added is a combo box to set the layout scale. See the warrant's help documentation for details.
- The algorithm to change speeds due to track conditions is improved to more accurately predict speed change start and end points.
- Warrants now respond to block speed changes.
- Aspect speed changes for warrants may be indicated by actual prototype speeds - e.g. mph or kmph.
- Individual engine addresses may be calibrated with a "Throttle Factor" so they can run at scale track speeds. The NXWarrants have a new feature to calibrate an address.
- The "Offset(sec)" column in the Signal Table at Add Items->Occupancy Blocks now can adjust the the end point of a speed change (+/-).
- Warrant help documentation is updated.
The name of the active profile is displayed on the PanelPro main window and in the status bar on all Roster windows (including the DecoderPro 3 main window). (Randall Wood)
Fixed an issue where editing the DecoderPro 3 Preferences could cause the configuration to be corrupted or destroyed. (Randall Wood)
Lisby made an enhancement to the train tracking as implemented in Block.java: When the existing code has not been able to track the train, i.e. the train ID in terms of block value will be lost. The enhancement stores the candidate Blocks that a train may have arrived from when a block becomes occupied and then when a block becomes unoccupied, it informs its neighbors who can then choose to use their candidate list to determine where their train came from.