Adjust forget commands to respect the direction of locos

[nic547]

For context: I'm trying to write my own software that interfaces with DCC++ EX. There I need to deal with the fact that only a limited number of locos can be "active" at a time. I would like to "forget" locos that aren't actually in use to keep the number of active locos down. I don't think there's a good way to figure out which locos are parked on the layout and which ones have been removed from the layout, so my plan was to forget both, as a loco being "unknown" and not getting speed reminders is kinda the default anyway.

Currently the "forget locos" commands send e-stops without the direction bit set, e.g. they're backwards e-stops. When a loco is still on the layout, this can lead to the headlights facing the wrong way, which is a bit unfortunate IMHO.

With this PR, instead of sending the "backwards e-stop" packet to locos the CS will send the e-stop packet with the direction the locos should already have. While I personally only require the change to the "forget this loco" command, I have also adjusted the "forget all locos" command to be consistent. Instead of sending a broadcast e-stop, it now only sends the e-stop to locos in the speed table.

• DCC::forgetLoco calls the setThrottle function twice - I'm not sure why, but assumed it's to ensure that the loco actually gets the packet, since it's not gonna get any reminders. DCC::forgetAllLocos doesn't send it twice. Should I adjust one of them?

• Using DCC::setThrottle instead of DCC::setThrottle2 results in a call to DCC::updateLocoReminder. If I understand the code correctly, calling the updateLocoReminder with for a loco that isn't in the speedTable should do nothing. Is it ok to rely on that behaviour?

This PR is probably a breaking chance, while not explicitly documented I would assume that other software might rely on the current behaviour (at least mine does). So it's not quite obvious if this change is worth it or if there are better options.

Also thanks to haba on discord for helping me a bit with this.

[Asbelos]

I think calling updateLocoReminder for a loco that isn't in the list will add it to the list.. On 4 May 2022 13:02, nic547 @.***> wrote: For context: I'm trying to write my own software that interfaces with DCC++ EX. There I need to deal with the fact that only a limited number of locos can be "active" at a time. I would like to "forget" locos that aren't actually in use to keep the number of active locos down. I don't think there's a good way to figure out which locos are parked on the layout and which ones have been removed from the layout, so my plan was to forget both, as a loco being "unknown" and not getting speed reminders is kinda the default anyway.

Currently the "forget locos" commands send e-stops without the direction bit set, e.g. they're backwards e-stops. When a loco is still on the layout, this can lead to the headlights facing the wrong way, which is a bit unfortunate IMHO. With this PR, instead of sending the "backwards e-stop" packet to locos the CS will send the e-stop packet with the direction the locos should already have. While I personally only require the change to the "forget this loco" command, I have also adjusted the "forget all locos" command to be consistent. Instead of sending a broadcast e-stop, it now only sends the e-stop to locos in the speed table.

DCC::forgetLoco calls the setThrottle function twice - I'm not sure why, but assumed it's to ensure that the loco actually gets the packet, since it's not gonna get any reminders. DCC::forgetAllLocos doesn't send it twice. Should I adjust one of them?

Using DCC::setThrottle instead of DCC::setThrottle2 results in a call to DCC::updateLocoReminder. If I understand the code correctly, calling the updateLocoReminder with for a loco that isn't in the speedTable should do nothing. Is it ok to rely on that behaviour?

This PR is probably a breaking chance, while not explicitly documented I would assume that other software might rely on the current behaviour (at least mine does). So it's not quite obvious if this change is worth it or if there are better options. Also thanks to haba on discord for helping me a bit with this.

You can view, comment on, or merge this pull request online at:   https://github.com/DCC-EX/CommandStation-EX/pull/233

Commit Summary ef009a0 Adjust forget commands to respect the direction of engines

File Changes (1 file)

M
DCC.cpp
(13)

Patch Links: https://github.com/DCC-EX/CommandStation-EX/pull/233.patchhttps://github.com/DCC-EX/CommandStation-EX/pull/233.diff

—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you are subscribed to this thread.Message ID: @.***>

[nic547]

Ok, I have adjusted that part. I extracted the calculation of the speed byte into it's own function, that I then use in the forget commands, so that I can call setThrottle2 and avoid the call to updateLocoReminder.

Is there an easy way to validate that the speed reminders aren't sent anymore?

[nic547]

Superseded by #242