开云体育

Date

Locked Re: Genesis OEM TSU2

 

开云体育

Michael, I will try to get that value and check to see if there are bulbs or leds. Reads as a TSU2. But had problems reading and writing with the 12 volt power supply. Need to correct that first.

On the WOW diesel Quill. Only looking at V3 and V4 wow diesel the quill choice was not there until i updated the definition but then on reopening a saved engine in the roster it is not. I feel it is my set up. but not sure what to look for.

??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? Rick.


On 8/22/2018 8:26 AM, Michael Mosher wrote:

I've been corresponding with Nancy at Soundtraxx, it looks like the F45 does not have it's own set of default values, reused same decoder as the SDP45.? I did not have any info for the SDP45 either, she gave me one but it's for the older version with 1.5V bulbs, it's CV 256 is 52.? Athearn's announcement pdf states the recent run of F45 have LEDs.

All my Wow diesel V3 & V4 open from roster show the quill check boxes on the sound tab, even those created last year.? V1 & V2 do not since that CV is not supported in those versions.
Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/22/2018 12:57 AM, Richard Knight wrote:

Michael, I do not have the loco here. I forwarded your email to the owner. We will try to get you that value. Did you see that the quill choice appears (WOW V4) when reading a new locomotive but when opening from the roster it does not appear? Also he had a 12volt supply for the sprog and has ordered a more powerful P.S.? I talked to Josh at Soundtraxx about you not having the CV list for the F45.??? Rick.


On 8/21/2018 9:56 PM, Michael Mosher wrote:
Can you tell me the value of CV 256 for the F45?

Thanks,
Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/17/2018 12:53 AM, Richard Knight wrote:

Michael, Thanks. As i mentioned this was an over the phone "help" session that became complicated when we discovered the definition for the F45 was not listed. We did pick a loco with a similar prime mover but still ended up with an unresponsive decoder. After a reset and a retry today at my friends place I discovered the Sprog 2 he is using did not seem to have enough power to program and read correctly the TSU2. I use a Sprog 3 with upgraded power supply at home so my results have been better. Found a similar situation at the club with a 12 volt 125ma power supply? for the sprog not being able to read write a QSI V7 decoder. Rick.


On 8/16/2018 5:58 PM, Michael Mosher wrote:
The info has not been released by Soundtraxx yet for TSU2 equipped F45s.
They have info on Genesis locos here:

but the F45 not yet listed AND those files don't have everything needed to create a decoder definition.? For Tsunami2 and Econami decoders, JMRI needs the value of CV253 and 256 (read only values in the decoder); 253 is always 3 for Genesis & Athearn locos but 256 is unique for each model and the definition need that number for the identification routine to work.? I've been waiting for a few months now to get info from Soundtraxx but with their trade up program, new decoder models released and recent NMRA show they've been busy.

Now you can use other Genesis TSU2 definitions, the default values are not going to match what's in the loco so you'll have to read everything before making changes.? Another thing that might not match will be enumerated lists of sound choices, but so far the Genesis lists have been fairly consisted between models with same prime mover type, since F45 has 645, pick one that also has 645 (like a GP40-2) and probably the sound choices will be OK.

Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/15/2018 3:45 PM, Richard Knight wrote:
Hi, i am trying to help a friend, over the phone!, with his Athearn Genesis F45 with TSU2. The basic problem is this engine is not listed in the choices when reading the loc with decoder pro. The F45 and FP45 are listed in the TSU1 but some of the CV's have been changed with different options in the TSU2. ?I was using GP38-2 as a stand in but this became problematical also. Is there any chance that the F45 could be added to the list? ?Thanks to Michael Mosher I could get the new file on the WOWV3and 4 diesel, download and install into the current version of decoder pro. Thanks Again, Rick.






Locked Re: Dispatcher Automatic running not reversing at end of section

 

Hi Peter
While running have the log display open PanelPro->Debug->Display Message Log and
see if you get any messages. This problem is normally caused by block
connectivity issues, turnout connectivity issues or a missing signal controlling
the return exit from the last block.
Steve G.

---------- Original Message ----------
From: "Peter Riggall via Groups.Io" <petermmrs@...>
Date: August 22, 2018 at 9:21 AM


Having resolved signalling problems, I have now set up Transits and tried to
operate a train automatically.

Both in simulation and real situations, when the train (just a loco at this
stage) arrives at the destination block it stops but fails to reverse
(although the dispatcher active train status still shows "running"), however,
if it is physically pushed back into the preceding Section (or the block
occupancy sensor of the last block in the previous Section) is set manually in
simulation mode), automatic running in reverse commences back to the starting
block.? The train then auto forwards and repeats the outward trip but again
stopping at the destination block until pushed to the preceding Section and so
on.? I have used the DCC monitor to observe the loco commands which confirm
this i.e. At the destination: a speed step 0 forward is issued but nothing
else whereas at the starting block a speed step 0 reverse followed by speed
step 0 forward followed by a speed step 7? forward is issued.

The Transit consists of four sections with direction sensors incorporated into
the SML.? The panel has the direction displayed and it shows the sections
being reversed.

I cannot work out why the train wont auto reverse at the ned of the outward
trip - any ideas anyone?

Thanks in anticipation.

Peter



Locked Re: Econami function pane

 

Thank you. That fixed it.? Steve?

On Wednesday, August 22, 2018, 9:26:30 AM CDT, Michael Mosher <mjmx1@...> wrote:


Its possible you have an old definition file.?
Pick Help - Locations...
Push the "Open user Files Location" button
Look in the decoders folder
Delete any old files for the ECO (and any other decoders in the normal installations)
Then back in DP
Pick Actions - Recreate Decoder Index
Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/21/2018 11:05 PM, Steve Teter via Groups.Io wrote:
Is it possible when I downloaded this version I didn't get all of the program?? Steve


Locked Re: Full timetable automation - AutoDispatcher 2 - abstraction - documentation - scripting

 

On Aug 22, 2018, at 3:55 AM, jamespetts via Groups.Io <jamespetts@...> wrote:

There is much to be said, however, for the view that, if one is proposing to spend a considerable effort coding a new feature, it is optimum to assess the requirements for that feature carefully before one starts and plan how those requirements can best be implemented, as doing so can make the overall amount of time spent fulfilling those requirements in the long term rather less than if one implements a system that fulfils some but not all requirements, but which cannot easily be expanded to fulfil other, closely related requirements.
I agree that you’re welcome to work that way. Many people do. How can I help you start?

But I _also_ want to encourage people (who want to) to get started with some code, as that gets something that’s working for at least some people much earlier. Code can be a big win, because JMRI users tend to get more joy out of using code on their railroad than they get of out reading requirements documents.

There are many ways to do this, and many of them work. Although I certainly have seen the NASA “make sure of accurate and complete requirements” process work[1], it can be slow and expensive, so even they are starting to move toward a more iterative development cycle[4]. In less high-stakes projects, like pretty much all hobby software, one has to balance the possibility of implementing into a corner, with the possibility that all the requirements documentation becomes just an esoteric form of science fiction.

In my experience, but perhaps not in other peoples, there’s a _much_ larger risk that nothing gets done than that almost-everything gets done but the last bit is harder than it could have been. So it’s good to start.

JMRI example: When LayoutEditor was implemented by David Duchamp, the implementations for turnouts didn’t cover the possibility of two turnouts on the layout that responded to the same address. This turned out to be hard to retrofit and as far as I know hasn’t been; people work around it with scripts and/or Logix. One might say “Proves the point! More early thought!”, but reasonable people can differ on that: Layout Editor was out and being used for almost two _years_ before the 1st mention of that on jmriusers. A quick scan of emails shows over 60 people using “Layout Editor” in emails before that, so there were a lot more people getting use out of it for quite a while than there were people stuck; and the stuck people didn’t lose _anything_ because without David being motivated enough to just write Layout Editor, they wouldn’t have had anything at all.

Don’t get me wrong. I agree there’s real value in somebody thinking through and _writing_ _down_ the tricky cases, things to look out for, etc. Careful & systematic thought is hard, and should be appreciated more. It’s a great contribution. But waiting for it is neither necessary nor sufficient to have something that runs and makes some people happy. The most important thing, by far, is for somebody to put their own time and effort into making _something_ that actually works for others.[5]

So, back to the question in the 1st paragraph: What can we (me or anybody else here) do to help some people get started on this effort? Maybe a couple people want to start on a Google Doc of important cases, and some others can generate some just-to-see-how-it goes scripts to try ideas out? Happy to help.

Bob




[1] I’ve been part of teams that wrote code to fly on the Space Shuttle; although not subject to the full rigorous flight-software sequence[2], there was certainly a very requirements driven process[3]
[2]
[3] (N.B. this is a couple years more recent than the versions we used)
[4[
[5] For examples of the full range, see and
--
Bob Jacobsen
rgj1927@...


Locked Re: Econami function pane

 

开云体育

Its possible you have an old definition file.?
Pick Help - Locations...
Push the "Open user Files Location" button
Look in the decoders folder
Delete any old files for the ECO (and any other decoders in the normal installations)
Then back in DP
Pick Actions - Recreate Decoder Index
Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/21/2018 11:05 PM, Steve Teter via Groups.Io wrote:

Is it possible when I downloaded this version I didn't get all of the program?? Steve


Locked Re: When reading sheets, incorrect setting values are returned

 

Robin,

That is a clear sign you need to go to Edit->Preferences->Defaults, and
change the selections to LocoNet from Internal. Then save and restart. That
should take care of it. Some versions tend to change the Defaults way too
easily by itself. The latest code avoids a few of the reasons. But starting
a profile without having the hardware connected is one of the simplest ways
to have the Defaults change. If you need to work without the layout
connected, make a separate profile that uses Simulator connection for your
LocoNet.

-Ken Cameron, Member JMRI Dev Team
www.jmri.org
www.fingerlakeslivesteamers.org
www.cnymod.com
www.syracusemodelrr.org


Locked Re: When reading sheets, incorrect setting values are returned

 

Robin,

2 things:

Make sure all of your preferences are NOT set to Internal.

What version of JMRI are you using. You should be at 4.12 at least

HTH

John Griffin?

On Wed, Aug 22, 2018 at 7:32 AM Robin Kerr <robkerr@...> wrote:
?
When I call up a known decoder and attempt to read the setting values on any sheet, the value 123 is returned in all cases, with the loco address as 15227!
?
If I create a new loco, and then read the individual sheets, they are read correctly, but as soon as I save the new loco, and then recall the decoder and attempt to read it, the above recurrs.
?
I have had no issues with my setup before, but today this is happening with any number of decoders.
?
My setup is a Digitrax DCS100, Locobuffer II, and a Powerpax.
?
Any help will be appreciated,
?
Robin Kerr



--
John Griffin
__________________________________
I've lost my dog and my remote...

Reward for remote.


Locked Re: Having a problem loading panel pro

 

That means that there’s a system name of “1” somewhere in your file. Perhaps it’s near line 20933, but the line counts are not always accurate.

System names should start with a prefix for the specific connection (a letter and optionally some numbers; i.e D or L2) and a type letter (T for Turnout, S for Sensor). So that 1 should be DT1 or L2S1 or something else. Finding it in the file and looking what’s around it will usually make it clear what it should be.

You _might_ be able to fix it by:

*) Make copies of everything
*) Run the jython/TurnOffXmlValidation.py script before loading your panel file (If you’re loading at startup, set the Preferences to run the script first; if you’re manually loading the file via the Panels menu, run the script from the Panels menu first)
*) Load your file
*) Store the contents out to a new file, with a different name
*) Quit, restart, load that new file and check if things are OK, particularly things with a “1” associated

The 2nd step will tell JMRI to proceed past the error, and _try_ to make it right. When you write out and test that file, you’ll see whether it was able to fix it or not.

If not, post the panel file and we’ll get it fixed.

Bob

On Aug 21, 2018, at 10:04 PM, amtkkydd@... wrote:

Hi Everyone I'm getting a parse error when I try loading panel pro here is the
exception:org.jdom2.input.JDOMParseException: error on line 20933: cvc-minlenght-valid: value '1' with lenght = '1' is not facet-valid with respect to minlenght '2' for type 'systemNameType'. I need some help to find this error Thanks Phil
--
Bob Jacobsen
rgj1927@...


Locked When reading sheets, incorrect setting values are returned

Robin Kerr
 

开云体育

?
When I call up a known decoder and attempt to read the setting values on any sheet, the value 123 is returned in all cases, with the loco address as 15227!
?
If I create a new loco, and then read the individual sheets, they are read correctly, but as soon as I save the new loco, and then recall the decoder and attempt to read it, the above recurrs.
?
I have had no issues with my setup before, but today this is happening with any number of decoders.
?
My setup is a Digitrax DCS100, Locobuffer II, and a Powerpax.
?
Any help will be appreciated,
?
Robin Kerr


Locked Dispatcher Automatic running not reversing at end of section

 

Having resolved signalling problems, I have now set up Transits and tried to operate a train automatically.

Both in simulation and real situations, when the train (just a loco at this stage) arrives at the destination block it stops but fails to reverse (although the dispatcher active train status still shows "running"), however, if it is physically pushed back into the preceding Section (or the block occupancy sensor of the last block in the previous Section) is set manually in simulation mode), automatic running in reverse commences back to the starting block.? The train then auto forwards and repeats the outward trip but again stopping at the destination block until pushed to the preceding Section and so on.? I have used the DCC monitor to observe the loco commands which confirm this i.e. At the destination: a speed step 0 forward is issued but nothing else whereas at the starting block a speed step 0 reverse followed by speed step 0 forward followed by a speed step 7? forward is issued.

The Transit consists of four sections with direction sensors incorporated into the SML.? The panel has the direction displayed and it shows the sections being reversed.

I cannot work out why the train wont auto reverse at the ned of the outward trip - any ideas anyone?

Thanks in anticipation.

Peter


Locked Re: Having a problem loading panel pro

 

It would be useful to know which version of JMRI you are running now, I
suspect you just updated, so also the version you were running. The more
recent versions have tighter checking of the panel files. The results are
rejecting the load of a file to point to a problem instead of (many times)
silently not doing something during the load.

If you can post the file on the group site in the 'Problems being worked on'
we can see it and better suggest the right fixes.

-Ken Cameron, Member JMRI Dev Team
www.jmri.org
www.fingerlakeslivesteamers.org
www.cnymod.com
www.syracusemodelrr.org


Locked Having a problem loading panel pro

 

Hi Everyone I'm getting a parse error when I?try loading panel pro?here is the
exception:org.jdom2.input.JDOMParseException: error on line 20933: cvc-minlenght-valid: value '1' with lenght = '1' is not facet-valid with respect to minlenght '2' for type 'systemNameType'.? I need some help to find this error? Thanks Phil


Locked Re: Genesis OEM TSU2

 

开云体育

I've been corresponding with Nancy at Soundtraxx, it looks like the F45 does not have it's own set of default values, reused same decoder as the SDP45.? I did not have any info for the SDP45 either, she gave me one but it's for the older version with 1.5V bulbs, it's CV 256 is 52.? Athearn's announcement pdf states the recent run of F45 have LEDs.

All my Wow diesel V3 & V4 open from roster show the quill check boxes on the sound tab, even those created last year.? V1 & V2 do not since that CV is not supported in those versions.
Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/22/2018 12:57 AM, Richard Knight wrote:

Michael, I do not have the loco here. I forwarded your email to the owner. We will try to get you that value. Did you see that the quill choice appears (WOW V4) when reading a new locomotive but when opening from the roster it does not appear? Also he had a 12volt supply for the sprog and has ordered a more powerful P.S.? I talked to Josh at Soundtraxx about you not having the CV list for the F45.??? Rick.


On 8/21/2018 9:56 PM, Michael Mosher wrote:
Can you tell me the value of CV 256 for the F45?

Thanks,
Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/17/2018 12:53 AM, Richard Knight wrote:

Michael, Thanks. As i mentioned this was an over the phone "help" session that became complicated when we discovered the definition for the F45 was not listed. We did pick a loco with a similar prime mover but still ended up with an unresponsive decoder. After a reset and a retry today at my friends place I discovered the Sprog 2 he is using did not seem to have enough power to program and read correctly the TSU2. I use a Sprog 3 with upgraded power supply at home so my results have been better. Found a similar situation at the club with a 12 volt 125ma power supply? for the sprog not being able to read write a QSI V7 decoder. Rick.


On 8/16/2018 5:58 PM, Michael Mosher wrote:
The info has not been released by Soundtraxx yet for TSU2 equipped F45s.
They have info on Genesis locos here:

but the F45 not yet listed AND those files don't have everything needed to create a decoder definition.? For Tsunami2 and Econami decoders, JMRI needs the value of CV253 and 256 (read only values in the decoder); 253 is always 3 for Genesis & Athearn locos but 256 is unique for each model and the definition need that number for the identification routine to work.? I've been waiting for a few months now to get info from Soundtraxx but with their trade up program, new decoder models released and recent NMRA show they've been busy.

Now you can use other Genesis TSU2 definitions, the default values are not going to match what's in the loco so you'll have to read everything before making changes.? Another thing that might not match will be enumerated lists of sound choices, but so far the Genesis lists have been fairly consisted between models with same prime mover type, since F45 has 645, pick one that also has 645 (like a GP40-2) and probably the sound choices will be OK.

Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/15/2018 3:45 PM, Richard Knight wrote:
Hi, i am trying to help a friend, over the phone!, with his Athearn Genesis F45 with TSU2. The basic problem is this engine is not listed in the choices when reading the loc with decoder pro. The F45 and FP45 are listed in the TSU1 but some of the CV's have been changed with different options in the TSU2. ?I was using GP38-2 as a stand in but this became problematical also. Is there any chance that the F45 could be added to the list? ?Thanks to Michael Mosher I could get the new file on the WOWV3and 4 diesel, download and install into the current version of decoder pro. Thanks Again, Rick.





Locked Re: Z21 and POM

 

开云体育

Oops yes, you were correct about the build.

In any case, package build 2659 and later have the required updates to make this work. ?There was a subtle bug that took me a while to find.

Paul


On Aug 21, 2018, at 3:03 PM, Moreno Mazzini <moreno.mazzini@...> wrote:

On Tue, Aug 21, 2018 at 06:31 PM, Paul Bender wrote:


?

On Aug 21, 2018, at 11:19 AM, Moreno Mazzini <moreno.mazzini@...> wrote:

On Tue, Aug 21, 2018 at 02:42 PM, Paul Bender wrote:
Moreno,Build 2696 or later should have all this fix. The problem was an incomplete interpretation of the LocoNet message received.

Paul
Sorry Paul, I tested last build/package available #2655, and now single CV programmer is ok for me, but I have error 306 again if I write a CV in a panel of a decoder definition.
?
Just looking at this again, 2696 is the latest build available at
?
You may need to refresh the page to see it. ?2655 definitely does not have the most recent fix.
?
Paul
Excuse me Paul, I have package 2655 (at I see the lastest build is 2696, but at I see the latest package is 2655).
Seems to me that are the same (build 2696 and package 2655). Both include your commit

Commit by
better interpretation of the loconet messages received.
Sorry, I use wrong word. I confirm I have 306 error when I write a CV in POM in a panel of a decoder definition.
Moreno


Locked Re: Full timetable automation - AutoDispatcher 2 - abstraction - documentation - scripting

 

There are, of course, lots of different possible approaches to achieving timetabled operation: some of them more flexible (and requiring more coding work to set them up) and some of them more limited (requiring less coding work to set them up). I believe that I have defined earlier in the thread what I consider to be an ideal system within the bounds of what can realistically be achieved if those who are inclined to do this sort of work are inclined to spend sufficient time doing so. Whether they are, of course, is another matter entirely, and no criticism is intended if they are not: this is, as Bob pointed out, a hobby for all concerned.

There is much to be said, however, for the view that, if one is proposing to spend a considerable effort coding a new feature, it is optimum to assess the requirements for that feature carefully before one starts and plan how those requirements can best be implemented, as doing so can make the overall amount of time spent fulfilling those requirements in the long term rather less than if one implements a system that fulfils some but not all requirements, but which cannot easily be expanded to fulfil other, closely related requirements.


Locked Re: Genesis OEM TSU2

 

开云体育

Michael, I do not have the loco here. I forwarded your email to the owner. We will try to get you that value. Did you see that the quill choice appears (WOW V4) when reading a new locomotive but when opening from the roster it does not appear? Also he had a 12volt supply for the sprog and has ordered a more powerful P.S.? I talked to Josh at Soundtraxx about you not having the CV list for the F45.??? Rick.


On 8/21/2018 9:56 PM, Michael Mosher wrote:

Can you tell me the value of CV 256 for the F45?

Thanks,
Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/17/2018 12:53 AM, Richard Knight wrote:

Michael, Thanks. As i mentioned this was an over the phone "help" session that became complicated when we discovered the definition for the F45 was not listed. We did pick a loco with a similar prime mover but still ended up with an unresponsive decoder. After a reset and a retry today at my friends place I discovered the Sprog 2 he is using did not seem to have enough power to program and read correctly the TSU2. I use a Sprog 3 with upgraded power supply at home so my results have been better. Found a similar situation at the club with a 12 volt 125ma power supply? for the sprog not being able to read write a QSI V7 decoder. Rick.


On 8/16/2018 5:58 PM, Michael Mosher wrote:
The info has not been released by Soundtraxx yet for TSU2 equipped F45s.
They have info on Genesis locos here:

but the F45 not yet listed AND those files don't have everything needed to create a decoder definition.? For Tsunami2 and Econami decoders, JMRI needs the value of CV253 and 256 (read only values in the decoder); 253 is always 3 for Genesis & Athearn locos but 256 is unique for each model and the definition need that number for the identification routine to work.? I've been waiting for a few months now to get info from Soundtraxx but with their trade up program, new decoder models released and recent NMRA show they've been busy.

Now you can use other Genesis TSU2 definitions, the default values are not going to match what's in the loco so you'll have to read everything before making changes.? Another thing that might not match will be enumerated lists of sound choices, but so far the Genesis lists have been fairly consisted between models with same prime mover type, since F45 has 645, pick one that also has 645 (like a GP40-2) and probably the sound choices will be OK.

Michael Mosher
Webmaster, ECSFM            
Member SFRH&MS              
DCC Master PVSMR            
On 8/15/2018 3:45 PM, Richard Knight wrote:
Hi, i am trying to help a friend, over the phone!, with his Athearn Genesis F45 with TSU2. The basic problem is this engine is not listed in the choices when reading the loc with decoder pro. The F45 and FP45 are listed in the TSU1 but some of the CV's have been changed with different options in the TSU2. ?I was using GP38-2 as a stand in but this became problematical also. Is there any chance that the F45 could be added to the list? ?Thanks to Michael Mosher I could get the new file on the WOWV3and 4 diesel, download and install into the current version of decoder pro. Thanks Again, Rick.




Locked Re: Econami function pane

 

Is it possible when I downloaded this version I didn't get all of the program?? Steve


Locked Re: Econami function pane

 

I'm using the first one on the list econami diesel ECO-21 pin.? This was selected by JMRI when it read the decoder.? I noticed the UK version looks fine.? ?Steve


Locked Re: Econami function pane

 

开云体育

Which Econami definition are you using. I took a look at "ECO-21P Diesel" and it looked fine.?

--?
Dave in Australia

The New England Convention 2018

On 22 Aug 2018, at 12:48 PM, Steve Teter via Groups.Io <s_teter@...> wrote:

I was programing an early econami 21 pin decoder today and noticed there was no legacy function pane and the regular function pane seemed to be incomplete.? Am I missing something?? I'm using 4.13.2 on a windows 7 machine.? ? Steve Teter??


Locked Re: Saved roster

 

开云体育

There's only one Roster File for each decoder and programming changes should be saved to that.

--?
Dave in Australia

The New England Convention 2018

On 22 Aug 2018, at 8:07 AM, paxton58@... wrote:

When I update CVs with the programming file and save the updates to the roster, they are not saved to the roster file in the JMRI file.?? Are there two different roster files and how do I update the correct roster?