The main new features in this release are random delays at locations, random points and signal failures, and temporary speed restrictions. These are described in section 6.13 of the user manual and on-screen help.
The Actions Due and Performance logs look the same as before but now they are Windows forms rather than panels, and as such can be moved outside the main display area. This will be useful for those with high resolution monitors where ROS doesn’t fill the entire display area, and for those with two or more monitors.
More detailed preferred direction checks have been incorporated to help identify discontinuities that are hard to spot by visual inspection.
A ‘flush’ command now follows every write to the performance log so that external inspection tools have access to up to date information.
Several Beta versions available on Discord preceded this release to allow wider testing in the community and I am grateful for the work done by members and for their suggestions and error reports. The result is better for their input.
Please don’t try to load any session files saved by Beta versions prior to Beta4 because they are likely to fail. Code for forward and backward compatibility had not been completed in those versions.
Bugs
There have been quite a number of bugs reported since the last release as follows:-
An error in the timetable validator was found to cause a crash when track element identifiers contained unexpected characters. The situation should merely have reported the syntax error which it now does, rather than crashing the program. Thanks to Amon Sadler for reporting this error.
An error in the timetable ‘New Entry’ function was found to occur with an empty timetable when the keyboard shortcut was used rather than the button. The program failed to make the necessary check which it now does. Thanks to mathstrains19 for reporting this error.
An error was found when joining two trains under signaller control when they should have joined under timetable control. This situation is now detected and a message given rather than the program crashing. Thanks to Micke(Commuterpop) for reporting this error.
It was found that a train at a station that had suffered a random delay could have the delay removed by taking signaller control. This happened on restoration of timetable control and was caused by that function forcing a recalculation of the train’s departure time which bypassed any existing random delay. It was a relatively easy fix to avoid the recalculation and carry out the necessary actions within the timetable restoration function itself. Thanks to FinsburyPark for reporting this error.
Another error was found by FinsburyPark when skipping timetabled events, where one more event was skipped than should have been. This was found to be due to my having added the word ‘Timetable’ in the floating information window to make it clear that the times listed didn’t take account of random delays. The program however interpreted it as a timetabled event that needed to be skipped, with the result that one too many events were skipped. Once identified it was a simple matter to make the program ignore this line. Thanks again to FinsburyPark for reporting this error.
This wasn’t really a program bug as such but a further anomaly caused by Windows’ management of high resolution monitors. The ‘Actions Due’ panel displayed larger than it should have done, resulting in the wrong train being identified on clicking a headcode in the list. The solution is to override high display settings in program compatibility mode. In response to this I strengthened the advice in the ‘High Resolution Monitor Scaling Issues’ page of the website. Thanks to josco for reporting this issue.
Download it here