开云体育

I can't see the LNWI being any real use here.

However, what could work, at broadly similar amounts of money would be:

Layout-LocoBuffer-RaspberryPI(running JMRI, connected to WiFi system)

Laptop running JMRI, talking via WiFi to RaspberryPI using either screen sharing (VNC and similar), or via the settings for a networked connection between two instances of JMRI.


RaspberryPI v3 would be powerful enough, the new v4 has more power. Once configured to run stuff at startup, and with the addition of a shutdown button (numerous cheap solutions on internet), the PI would be an automatic "black box".


- Nigel

James,

Try Options >> Add >> Add Background Image…

Dave Sand

I have a version of JMRI 4.14 and it is working very well.

I have attempted to place a logo jpg image without success on the layout panel using layout editor.

I need a step by step procedure so I can place the logo of the layout on an area of the layout panel.

Thanks

Hi

I was viewing a layout editor panel via the webserver and noticed that the crossover was showing the incorrect occupation status. Only a change to Block 1 of the crossover is recognised, and this shows all 4 areas of the crossover as occupied. Changes to Blocks 2 - 4 make no difference on the web panel. This is the case whether it is closed or thrown.

I was initially using v4.16 on a raspberry Pi. However, I've since recreated the problem on 4.16 and 4.17.3 running on Ubuntu 19.04, java 11.0.4.

Showing the problem is very simple:

• Create a layout editor panel
• Put a crossover on the panel (using created turnout)
• With the Tables
  • Create 4 Blocks
  • Create 4 Sensors
  • Link sensors 1-4 to blocks 1-4
  • Create 1 turnout
• Edit the crossover and assign the blocks and turnout to it
• Start webserver and view panel
• Alter state of sensor 1 using Sensor Table
• Notice the native panel shows only top left quadrant as occupied
• Notice that the web panel shows all 4 quadrants as occupied
• Changing sensors 2-3 gives correct occupation on native panel but does not change the web panel.
• Notice that web 'preview' panel (on the web home page) always shows the correct occupation (page refresh needed between each change)

I've created a simple panel using the above steps and uploaded to /g/jmriusers/files/ProblemsBeingWorkedOn/Andy%20Brown%20-%20161045/CrossoverBlockOccupancyIssue.xml

In my current application I have blocks 1 & 2 with one sensor and 3 & 4 with another. Hence, when in the crossed state the problem is not so obvious, but when closed it's really obvious.

The same issue is visible when the blocks are set to reserved (white). It is also the same with a double crossover. Slips only have one block so the problem does not occur.

Regards
Andy

Dave,
The feature is working as intended which is good. Sometimes people come up with something I haven't tried which is fantastic. I think (??????) that what I am looking for is a global feature. Some kind of an internal clock that will take the oldest car on a track and move it first and not start over at A1 every time. This feature would take quite a process to integrate into the program and would have to be able to turn off if someone doesn't want it. I see a lot of my cars not move because the search never gets to them. Most of the time I have figured work around to get what I want because why change for one person and I get that.
Good program
Thanks
Matt G?

Gary
I have two computers linked together. Bought them on sale and the cost was not much. I have one for CATS and the other for operations. I do not have my CATS panel on this computer but there is something in the CATS settings that listens and communicates with operations. The automations are in the tools area of the trains page. You create an automation that has something for a trigger. Say build train 2131 when train 1234 arrives at location ABC and the list goes on and on. My CTAS computer tells operations train?1234 occupied "station" ABC. The JMRI trains automation sees this and kicks in the automation for 2131. You can have operations and CATS both running on the same computer but there was a problem with something, (I cannot remember what that was now) where I couldn't perform something with both on the same computer. I'll remember right after I send this. But, two cheap computers running together works just fine. The automations also make it hard to build a train out of order if you are running trains in real time and train 1234 needs to be at location ABC before train 2131 should be built.

later

Dave
Trempealeau WI

Matt and company
I've used the interchange designation for a length of track here for quite some time. It works flawlessly. When a spur is filled, JMRI will place a car on the alternate track with a destination of the filled spur. The car moves at the next session to its destination. So I'm wondering what is it that you are not seeing with this feature? Gary, the same question. What does your build report tell you about the car and why it couldn't move? Get into the habit of looking for the car number from the bottom up. The last paragraph will show and in that paragraph it will tell you why the car didn't move. This has to be a missed setting.
later
Dave
Trempealeau WI

I have an MGP LocoNet Bridge.? While it's fantastic for setting values and fine tuning MGP components located in a LocoNet;? It's not designed as a layout or locomotive running device.? (In the terminal mode I can throw and close any LocoNet aware turnout, so the possibility is there, but not the implementation.)

Ken Moordigian
Jackson Livery
818-522-4292

On Sun, Aug 18, 2019 at 12:33 PM, dick bronson wrote:

Wayne,

Conventional English understanding would make you think that a 'Shorted' LocoNet would be a bad thing. However it just means that pins 3 & 4 are both connected to the LocoNet. In fact this is a good thing, not a bad thing, because it improves the redundancy of the connections. The LocoNet is a duplex connection only at the computer end. On the network itself there is just a single electrical signal? (LocoNet) and a ground. Both the LocoNet and ground are normally two wires with the same information for redundancy. The Railsync lines (1 & 6) are differential signals that drive the boosters. (Rail A and Rail B) If one of those goes bad, then there are issues.

Dick :)

Dick,?

I get what your saying and that does make sense. However, I still have the initial problem I started with and only a couple cables had four lights. As you mentioned LocoNet being duplex at the computer end.. Maybe one of the cables that is "shorted" is messing that up??


Wayne

Hi Dave,
What version (exactly) of JMRI are you running? Turnout position on web panel should reflect the same state as on the JMRI panel.
Also, what browser are you using? If it does not support WebSocket, you won't get updates. Test this at:?

I've not heard of those popups. It's very likely they are not coded into the client-side web panel. If you'll post your panel and advise how to duplicate the popups on the JMRI panel, I'll take a look.

--SteveT

I'm using layout editor and two-sensor feedback. When I throw a switch using the web server panel, it never changes position when the feedback event comes in. It will change if I refresh the browser. Is there a setting to allow this to happen? The Browser (client) would have to request the update from the WebServer?

Also using Entry Exit routing, the "cancel" and "stack" pop-ups don't appear on the web server

Thanks,
Dave

Have a few routers and other stuff laying about so might give that some consideration.

Thanks for replies.

Regards,

CFJ

One third-party wireless LocoNet connection device is a LocoNet-Bluetooth "smartphone interface" device from Mollehem. An english web page for this device is at .

JMRI supports this device via the "BT-LocoBridge" type of LocoNet interface type.

I do not have any practial experience with this product, so cannot advice of its suitability for any particular application. I believe that someone from Mollehem by the name of "Anders" pays attention to this list, and can provide further details.

Regards,
Billybob

CF:

There's a lot of other information going across the LocoNet interface using the LocoBuffer than a WiFi throttle uses. The LNWI is designed solely for WiFi throttles, though very flexible working with iOS/Android devices quite nicely.

While I think it would be possible to write a JMRI throttle connection or the PC or Mac equivalent of WiThrottle or EngineDriver I'm not sure i would be desirable enough to do that...

If you want to cut the cord I would look at WiFi (or other radio) USB bridges... PC to widget to radio to widget plugged into the LocoBuffer...

Jim Albanowski

Yes.

CFJ,

You want to substitute a WiFi connection between your computer and LocoNet instead of a LocoBuffer connection?

Remember, the LNWI is primarily intended to provide a mechanism for phones to control locomotives. So the communications protocol is focused on the messaging required to control locomotives.

While I have not specifically tested the LNWI and its communication protocol, I _suspect_ that the protocol used by the LNWI will not provide all of the capabilities of a LocoBuffer. For example, I would guess that you could _not_ apply firmware updates to your DCS210/DCS240/DT402/DT500/etc. via the LNWI communication protocol. This is just one of perhaps a hundred different LocoNet messages which I suspect cannot be communicated via the LNWI protocol.

Regards,
Billybob

Currently I have JMRI (on laptop with a loconet cable) working with loconet via a locobuffer.

Is it possible to have JMRI (on laptop with a wifi connection) working with loconet by substituting a LNWI to laptop wifi for the wired connection?

CFJ

Thanks to everybody for all the help on this.

Lot’s of interesting tests. It’s clear that programming isn’t always reliable, but there didn’t seem to be a case where Paged mode worked and Direct Byte did not work. So that makes the case easy.

Starting with the next test releases, Digitrax programming will first try Direct Byte mode.

Thanks again.

Bob

On Aug 12, 2019, at 4:44 PM, davethomas1359 via Groups.Io <davethomas1359@...> wrote:

I have a DCS51 but had a DCS50 previously. I could only get it to read an older QSI decoder in Direct Mode with a PR3.

On Monday, August 12, 2019, 04:31:48 PM GMT-4, Breezlys via Groups.Io <livinthelife@...> wrote:


On Sun, Aug 11, 2019 at 02:34 PM, Bob Jacobsen wrote:
But there’s still the open question as to whether (some) DCS100, DCS200 and DCS50 command stations have trouble reading and/or writing in Direct Byte mode some or all decoders. QSI decoders were mentioned several times in that context.

Is there anybody out there with one of those command stations and QSI decoders who could test reading and writing in Paged Mode and Direct Byte Mode?

Bob


--
Bob Jacobsen
rgj1927@...

Bob,

I tried QSI V7 and V9 decoders with both a PR4 and with a DCS100's programming outputs.

-The PR4 using Paged mode correctly read CV's 1 and 66 on both the V7 and V9 decoders, but corrupted CV29 (wrote 255 to it) on both of them. Writing 34 back to CV29 worked okay (verified by reading in Direct).
-The PR4 using Direct mode correctly read CV's 1, 34, and 66 on both the V7 and V9 decoders. Writing to these CV's worked okay.

-The DCS100 using Paged mode correctly read CV's 1 and 34 on the V7 decoder, but repeated read attempts all returned the same incorrect value for CV66. However, it did NOT corrupt this CV as a subsequent read with Direct returned the correct value. I did not attempt any writes with this combination.
-The DCS100 using Direct mode correctly read CV's 1, 34, and 66 on the V7 decoder. I did not attempt any writes with this combination.

-The DCS100 would not read the V9 decoder in either Paged or Direct mode. I did not attempt any writes with this combination.

HTH,
Steve
"Breezlys"

--
Bob Jacobsen
rgj1927@...

Ken,
No, I have not tried 4.12 as the documentation recommends.? Should have read that instruction more closely.
I will set a restore point on my PC and then give 4.12 a shot tomorrow.
There was some logic I'd included for a grade crossing signal that must have been overwhelming the CTI network on start-up.? This came to light during the upgrade attempt - actually when I tried to go back to 4.11.6. For some reason, the network tolerated this logic okay before the upgrade attempt, but was adversely affected after the upgrade and continued not to like it after reverting back to 4.11.6, even though the panel xaml file I used was written by 4.11.6.

I have that issue cleared up and the panel is loading once again without hiccups in 4.11.6.
So, fingers crossed I will try 4.12 tomorrow, and if the panel loads without issues, I will move on to 4.17.3 with great optimism and report back on this thread.

Mike

I've not used Layout Editor for a while. Just before I stopped working with it, I had an odd problem. The scroll bars disappeared. I set the option to have both vertical and horizontal scroll bars, but they didn't appear. I still have the problem. There must be some setting I accidentally changed. Can anyone tell me what it would be?

I could post the panel for someone to check if I can learn how to do it.

Don Weigt
Connecticut, USA