8 then 11 now 17 ??

Phil G

(Further reply while my previous is being moderated)

Phil G, thanks for mentioning Azuls Java 17. I'll investigate that.

To Bob: I'm running Windows 10 on an Intel i5-8500T @ 2.11 GHz, supported by 16 GB RAM; 64-bit OS, which doesn't seem too shabby to me.

If Java is available from other vendors (I had no idea), and if Phil G is able to run Java 17 on Windows 8 (!), then there's something here to be investigated.

Phil G: exciting! Can you tell us whose Java 17 you're using?

Thanks, guys.

Hi,

Thanks for the responses. I wanted to import my own decoder profile not one of the standard ones already supplied with the software and yes I followed the import procedures.

I did not update version 4.26 to 5.10, but did a fresh install in both cases. My previous version (I think) was 4.16 and I uninstalled this first, then installed 5.10. When I couldn't import my decoder profile, I uninstalled it, and then installed 4.26 and everything worked fine. So still puzzled why 5.10 would not work properly.

Regards

Roger

What is the URL of the page that isn't working? That looks like it is still going to a /panel page, instead of a /frame page, but I can't see the URL in the screenshot.?

I'd also like to see the websocket.org page, as this symptom looks 99.9% like a WebSocket issue.

Hello SteveT -

The websocket protocol works on the older iPads (I checked on echo.websocket.org). I also enabled "frames" and this didn't solve the problem. See the attached pictures for a panel that worked and one that doesn't work.

I appreciate any other suggestions you or others have.

Thanks!
Michael

--

Michael
Modeling the CNW low line

Dear Ken et all

?

success! Linking it as a turn out control works. Including the blocks.?
there is a minor bug?

if the turn out is set to go and the signal is go then the block is occupied, the signal returns to stop. All good. But if the turn out is then switched to closed the signal then moves to go. So if seems to be responding to alternative commands not to the logic of the layout. As writing this had a thought if in the logic edit. What happens if you select lock to turn out??

Michael,

The most likely cause is that your pad's browser does not support WebSocket protocol.

There is a test for this on the current JMRI web "About" page.

?

If that's the issue, the workaround is to use the /frame server instead of the /panel server. You need to enable "frames" in JMRI, Edit, Preferences, Web Server.

--SteveT?

Eric,

You did not answer my question about the initial turnout states after starting PanelPro and loading the xml file.

There is another step. ?Before starting PanelPro, start the NCE command station from a power off. ?After the command station is running, turn on track power. ?Now start PanelPro and load the xml file. ?Before doing anything else, what are the turnout table states?

Dave Sand

I should add that the newer iPads work just fine.

--

Michael
Modeling the CNW low line

Eric,

How do you know that it queries the physical turnout. ?Do you see a turnout message in the NCE Monitor?

Dave Sand

Hello -

I recently upgraded JMRI to 5.10 running Java 11.0.11. I've been using the webserver to run panels on multiple iPADs using Safari:

The two older iPads do not render the panel graphics correctly: Part of the panel is showing part is not. I've done the usual fixes (restarts, refresh). The operating system and browser cannot be further updated on the older iPads.

Any help you can provide would be greatly appreciated.

?

Thanks!,

Michael

?

--

Michael
Modeling the CNW low line

Dave,

That is exactly what I was talking about. When you do that click on the turnout, either on the Main Panel graphic or the Turnout Table, JMRI first queries the physical turnout. If the physical turnout state is different than the digital twin in JMRI, JMRI updates the digital twin's state in Panel Pro to reflect reality.

If you click a second time, a command is issued to change the state and then Panel Pro updates to reflect the new state.

?

I wanted to know if that functionality is, or could be, exposed somewhere as a function that could be called by a program.? If so, one could periodically run a polling program and update the panel for anything that got changes locally. Since most turnouts are not being actually changed, the power / bandwidth used should be minimal.

Hi Peter

?

Just saw this now. I will connect via discord.

- John

?

On Thu, Jan 9, 2025 at 11:04 AM, jngeddes wrote:

1. Re load

- Does the indicator of 1.00 == no extra load?

no.? ?0 = no extra load.? Remember it is exponential, so on a five point scale, 1 is only a bit more 2 as a fair bit more ... 5 is a hell of a lot.

- Load indicator seems to positive and negative but I can't determine why or whether this matters

- I can't figure out the load algorithm (3 load steps, load multiplier 100) load goes from 1.00 to 1.30 to 2.20 to 3.70

That is not a load indicator.? That is, behind-the-scenes, a momentum calculation.? It is combination of the load and brake. I intend to hide it in the final versions.

- When I set the load multiplier to 20 or 200, I get the same progression. A bug? Or am I doing something wrong?

umm... I see a huge difference in the momentum calculation.

Setting up these parameters for reasonable loco performance is requires a fair bit of "try it and see" given that I am also working with decoder momentum, so understanding the relationship of load step to throttle impact would be really helpful. I am assuming that that is what the load indicator is intended to provide.

Ok, I'll document it.

?

2. Re brakes

- I have been using the ESU three brake option - functions 4, 5, and 6.

- To keep it simple I have 3 brake steps

- Brake steps 1, 2 and 3 activate F4, F5 and F6 as expected. I like how it highlights the function in the normal function button section of the throttle.

- What is the slider bar to the immediate left of the brake panel showing?

That is an attempt to represent the air reservoir. If the reservoir is empty You can't apply brake, or full brake.

i.e. if you dump the brakes, the reservoir empties.? You can't fully reapply till it fills.

?

3. Re Accel/Decel settings

- The Repeat Delay setting implies it goes up from 1 msec. But it seems like there is some slower minimum speed step repeat delay. about 1 sec? True?

I don't see that. For me it steps at about the rate set. It won't be exactly that as it sends then receives a response, so there can be a little bit of delay.

?

The combination of the repeat delay(s) and the step effectively set the base momentum (no load) on the loco.

?

If you want to work with me on this I am happy for you to email me or contact me on the DCC-EX Discord server. ?? ?'flash62au (Peter)'

?

--

Peter Akers?
[Brisbane - Queensland - Australia]?

On Wed, Jan 8, 2025 at 08:05 PM, Jon Schmidt wrote:

Are you saying that the same schedule could be assigned to Spur A in match mode, and Spur B in sequential mode?? How?

For any given spur, select the schedule you want, and then select which mode to run it.

?

Dan

Hi George,

?

re Onboard Loco Simulator

?

Quite the project. Interesting. I agree -- wouldn't it be nice if such functionality could be built in to all decoders (assuming it didn't add a lot more cost). Perhaps it will come someday. The functionality of the ESP32-based technology is very impressive.

?

In terms of your idea for ED buttons, in my limited experimentation, I came to the conclusion that your idea of using the DOWN button to invoke a given function (brake) would be handy. I might still lean to a horn-like momentary action rather than a latching action but that is hard to confirm without actually being able to try it.

?

?

- John

I did see that the schedule and mode are in the location export.? I'm confused, tho.? Are you saying that the same schedule could be assigned to Spur A in match mode, and Spur B in sequential mode?? How?

?

Thanks for enhancing the schedule export.

?

Jon