¿ªÔÆÌåÓý

Date

Locked Re: Test Release 4.13.4 Timetable tool #timetable-tool

 

Paul,

Test release 4.13.6 (coming soon) has a new scale process. This will also apply to production 4.14 when it is released. N148 is being replaced by UK_N.

Before you upgrade to either of these releases, you will want to temporarily change the timetable scale to N. If you do nothing, the scale will automatically change to HO the first time you use the timetable tool.


Dave Sand

On Nov 23, 2018, at 9:22 AM, paul@... wrote:

Dave,

I've now downloaded 4.14.5. Thanks for the inclusion of N148 and the example script. I'm looking forward to setting up the timetable. I'd been wondering how I could easily set this sort of data up but this solves the problem brilliantly and I can now start experimenting with a script to do the integration with dispatcher.

A minor point - it appears that the 'Route Duration' field is not being set on the train display. This seems to be the case for new trains and trains migrated from the 4.14.4 version. The timings appear correct on the relevant segment's Train Graph. Also if you're considering any enhancements, a function to duplicate a train would be really useful, allowing for the situation where there a several trains with the same basic itinerary but at different times in the schedule.

Paul


Locked Re: Operational concept for a single operator #operationspro

 

Paul,

A car can only be assigned to one train.? When you move or terminate a train, you are telling the program the car dropped off is now available for the another train.? In your example, you have two trains moving in opposite directions.? If you build both trains, the cars assigned to each train are captured by that train. Since both trains are built, the cars can't be shared or swapped between them.

Now let's say you only build one train, that train captures a bunch of cars, and is scheduled to deliver those cars along the train's route.? If you move the train along the route, the cars that are spotted are now available for the train going in the opposite direction, or any other train.? So let's say the first train is half way through its route,? you move that train to the 5th location, 5 locations now have cars that were set out by the train.? If you now build the train going in the opposite direction, the cars that were set out are available to be pulled into that train.

So taking your example to an extreme, if you built 12 trains at once, not one of those trains would share a car.? But if you built and then terminated each train, the 12 trains could share many of the same cars.

You might want to try the demo files and play with them to see how cars move.

Dan


Locked Re: Raspberry Pi 3 Horse Power?

 

One thing I would say about Ras Pi boxes, they are very sensitive to the
speed of the SD card you plug into it. So the better cards are very much
worth the price.

I'd correct Jim's remark about Java, it is compiled, JavaScript is
interpreted. Ok, Java is compiled into P-Code, but I recall over 10 years
ago it was proved to be over 90% of the efficiency of native code. One thing
I fully agree with Jim is about automation, it requires a lot more blocks
and THEY MUST DETECT PERFECTLY.

On 'what's the slow point' I'd say that over 40 years of programming says it
is seldom where you think it is. Only when you can test and measure do you
find where the problem with speed lays. It usually turns out to be a big
surprise.

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


Locked How to deal with empty cars in OpsPro

 

I am not clear on how OpsPro handles empty cars. If I schedule an industry to receive an empty (E) boxcar, I would expect the program to find a E boxcar and route it to the industry. However, the documentation for OpsPro states "The program doesn't route cars with the default loads "E" or "L". So, if your schedule is requesting a car with the empty default "E", it can take awhile before that car shows up.". Why does OpsPro ignore (or defer?) the request for an empty (E) boxcar?
The documentation goes on to say that I can create a custom load called E(mpty) and implies that the industry would get a E(mpty) boxcar quickly just as if it was a load with implication that the boxcar would be considered loaded (thus routed) rather than empty (not routed or deferred).
If the answer is....the E boxcar will get to the industry eventually but not as fast as a loaded (aka...E(mpty)) boxcar, I can deal with that. However, if it means that OpsPro will never route an E and the industry will never get a E boxcar, then I am totally confused on why E (or L) even exists.
This brings up a related question....if a train stops at one location and picks up an E boxcar, will the train drop off the same E boxcar at the next location in its route that is requesting an E boxcar?
As a side note, I have been experimenting with this and I created some E(mpty) loads for boxcars but it is not clear how the program is treating these boxcars with a E(mpty) load if no industry is requesting a E(mpty) boxcar.
Obviously I am not clear on how to handle empties as I try to build schedules for my industries. Any help is appreciated.


Locked Re: Ops: "Percentage of custom loads generated by staging"

 

Hi Jon,

1) it applies only to the spur in question.
2) It can limit number of the cars with custom loads from staging.? 0% means no cars from staging to service the spur.
3) Yes, you can restrict which custom loads can ship from each staging track.

Dan


Locked Re: Packet Delays and Timeout - SPROG3 and JMRI 4.12

 

This is how SPROG in programming mode works. When you are programming (Read or Write) track power is OFF and you are using the programming track power levels. The whole layout/track you are linked to.?
When you open a throttle, the track power goes ON, and you can use the (single) throttle to test your settings.
SPROG in Operation mode has power ON, unless you switch to programming, when the power goes off, and you are in programming mode with lower track power etc.

Check the preferences for the mode you are using.
I generally set Decoder Pro to SPROG (Programming) and Panel Pro to SPROG Command (Operating) to avoid confusion. Then open DP to program, or PP to run.


Locked Re: Raspberry Pi 3 Horse Power?

 

Dennis,
I agree with the maybe, depends a lot on HOW you're planning to automate. I use JMRI Dispatcher for my layout, and I've had 5 trains scheduled, with 3 active at once, with no issues running on my RPi3.?
--SteveT


Locked Re: Raspberry Pi 3 Horse Power?

 

Dennis:

I would say maybe... It all depends on just how many options the software has to deal with. Since you're posting here timing will be an issue, Java is interpreted which slows operation. The Pi is fast in 2000 terms but no longer.

Let's see who chimes in on this.

Also check prices for SSD a 240GB isn't that high so a PC and certain Mac's could be fitted for a more "solid stte" machine.

On another practical side... Automatically stopping trains requires a lot more detection to be sure trains are positioned correctly.

Our club's programmer built an automated system and it's been a bit of a larger project than first thought... C++ code not Java/JMRI...

Jim Albanowski


Locked Re: Panel Pro - looking for up to date Tutorial for CPE

 

John,

Using CPE and SML is fine. Doing so DOES change how the SML is created.

¡°Automatic" generation of SML is based on the block topology defined by the Layout Editor track plan.

"Manual" creation of the SML does not require any panel of any type. The signal mast pairs, turnout positions and block states are entered directly into the SML definition screen.

Some people create a LE panel to provide the block topology and then create a CPE (or PE) panel for running the layout. The LE panel is either hidden or sits behind the CPE panel.


Dave Sand

On Dec 4, 2018, at 2:41 AM, John Pearson <pearsonjohn99@...> wrote:

Dick
Thanks you for the EXCELLENT reply dated Dec 2nd it has opened my eyes to the differences in the different JMRI design systems. I agree with Alan B Pearce¡¯s comments ¡° it should be in FAQ¡± As a newby JMRI bless it is akin to drinking from a full bore fire hydrant !

To your note it is apparent that I should be using (CPE ). Control Panel Editor
My layout is quite small and caters around an interlocking controlled by a CTC machine. there are several back to back crossovers and a flat crossing with 4 90 degree diamonds, many of the signals are eye candy with of stage virtuals.
I have no room for ABS or long stretches of track.

Signal Masts are obviously the way to go ( thanks again for the explanation )

What I need now more than anything is a JMRI tutorial covering CPE & SML

BTW I have some of your SM LEDs ( RR-CirKits ) the Colors are excellent
Many thanks
John Pearson


On 3 Dec 2018, at 11:55, Alan B. Pearce <alan.b.pearce@...> wrote:

This needs to be in an FAQ somewhere. Thanks to Dick for a clear definition of the differences.

From: [email protected] <[email protected]> On Behalf Of dick bronson
Sent: 02 December 2018 23:47
To: [email protected]
Subject: Re: [jmriusers] Panel Pro - Layout Editor - looking for up to date Tutorial

John,

I'll make some (many) comments based on my personal opinions.

a) Layout editor (LE) is a vector view of your layout. I.e. the toe bone connects to the foot bone, the foot bone, connects to the ankle bone, .... It is designed to document and edit the connectedness of the layout. It only allow for one definition of each chunk of the layout. (vector) As soon as you have two turnouts that operate as one (e.g. crossover) or two detectors that operate as one you need to fight the system to figure out how to defeat its basic nature. (typically by making an internal shadow to define the duplicate item) It was never designed to make a CTC panel drawing in the first place. Doing so is sort of like trying to use a CAD program to paint a landscape.

b) The Control Panel Editor (CPE) is designed to create a pictorial CTC machine (Control Panel) view of your layout. It consists of various icons and images that allow you to recreate either a classic or modern CTC panel with surprising fidelity. It neither knows nor cares how one image or icon connects to another. The disadvantage of that of course is that you need to manual define the items and how they fit together or what they control. The advantage of that is that it will not try to force you into its own way of doing things.

c) The Panel Editor (PE) is the original JMRI panel editor which has been superseded by the Control Panel Editor which has many new features and ease of use advantages over the original. PE should have been depreciated years ago, but JMRI seldom does that for anything.

d) Signal heads refers to the class of hardware that can only control a signal one set of lights at a time. Typically it controls R, Y, and G lamps.

e) Signal Masts refers to an entire signal, be it one head or many heads, 3 lights per head or 7, etc. Typically it operates by sending aspect messages, not individual light commands. However if your hardware doesn't understand aspects, then JMRI can create signal aspects out of multiple heads.

f) SSL stands for Simple Signal Logic. Think 'Simple' as in 'Simple minded', not simple as in 'Easy'. It does basic ABS on single track fairly well. However as soon as you add a crossover or anything more complex you need to cross your eyes and stand on one leg to figure out how to make it work correctly. (mostly done by adding invisible signals and calling them 'distant')

g) SML Signal Mast Logic. IMHO This is the way signal logic should always be done. It treats signaling as aspects (per the prototype). It calculates aspects as they relate to pairs of signals, i.e. the signal you are facing paired with the next signal along the path you will be taking. It has built in rules for a growing number of prototype railroads. This saves you a lot of research into how signaling should be done on your railroad. It takes the most complex trackage in its stride. For each signal pair you simply list the correct turnout positions along that route and list all the blocks between this signal and the next.

h) NEVER mix SSL and SML unless you enjoy cat fights and detective mysteries. They do not play at all nicely with one another and you will be left wondering just why in the world your signals behave the way they do.

i) Crossovers in CTC territory are two blocks. When calculating any block requirements always think about how many different trains can occupy any given area at the same time. That is the number of block detectors required. For single track it is one, unless the track is long, then it may be several so one train can follow another. For a single turnout it is one. For a crossover it is two, one on each side if they are passing. For a yard ladder it is ........ one. (unless you can switch two parts of the ladder at the same time somehow) Of course a yard ladder is not usually signaled, but it makes a nice example. <G>

Dick :)

._,_._,_


Locked Ops: "Percentage of custom loads generated by staging"

 

I am a little confused by this option on the Schedules page.
1) Does it only apply to the current spur, current schedule, or universally?
2) If it applies to the current, does it apply to all of the custom loads listed in the schedule?
3) Do I assume that the 'generate load' options for the staging track control which custom loads are actually built?

Thanks in advance,
Jon


Locked Re: Initial setting of sensors in Layout Editor

 

Thanks Steve & Ken.
Thats my first visit to Routes.
Works a treat? !


Locked Re: Operational concept for a single operator #operationspro

 

Scott....all that makes sense and helps me a lot. I understand how the second train works as long as it is moving in the same direction following the first train. How about if the two trains are moving in opposite directions at the same time. Example....you have 10 locations with one train starting at one end and the other starting at the other end at the same time. Both will visit each location but in opposite order. Does that make a mess of the picks and drops? How does the train build handle that since the cars at a location toward the tail end of a route may change vs what the build was expecting.

Even more things to consider...but I am learning.
Paul D


Locked Re: Initial setting of sensors in Layout Editor

 

There is a script called SetAllUnknownSensorsToInactive that is part of JMRI.? I use it for my testing.? You can have it run at start up by defining it in the Start Up section of the Preferences.

Dale Gloer


Locked Raspberry Pi 3 Horse Power?

 

Does the RPi 3 or 3+ have enough horse power to run a large HO layout with automated trains?

Say there are two large independent loops with staging yards, one with 2-4 trains and the other with 4-6 trains. No Bidirectional trains at this time.

Also looking at LCC.

Would like to run headless with the RPi and not worry about mechanical Failures with hard drives, etc.


Locked Re: Initial setting of sensors in Layout Editor

 

You have a few options. For Internal sensors, there is an option to set them
to initialize to INACTIVE. But for 'real' sensors, the issue is getting the
status from the layout. Here it depends on the type of hardware you have.
Many systems have a way to be 'polled' which forces the hardware to tell you
the current status. Some hardware unfortunately does not have that option,
some only report when they change state. Here you could fake the initial
state, like using a route to set the status, but that is only a guess in
that case.

Now when I'm working on a simulated layout, I use scripts to set the various
systems of sensors and turnouts to INACTIVE or CLOSED as part of my startup.
Here since there is no layout, I'm doing this to make testing simpler.

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


Locked Re: Packet Delays and Timeout - SPROG3 and JMRI 4.12

 

Hi All,

No Sprog hardware but in an earlier message Rockpro says "Thinking all this might be endemic to some of the other warnings and things".

What warnings are these?

Is there any weird activity in the main JMRI console when this behaviour occurs?
If so, please post here, would make it much easier for people to track down issues,

Steve.


Locked Re: Initial setting of sensors in Layout Editor

 

Hi Paul, create a new route,

PanelPro > Tools > Tables > Routes > Add

Select any turnouts / sensors you'd like setting and their states and save the route.

then go to?

PanelPro > Edit > Preferences > Startup > Add > Route > Select Route to activate on startup,


Steve.


Locked Re: Packet Delays and Timeout - SPROG3 and JMRI 4.12

 

Same here.?? Pi-Sprog (just got it the other day).??? Using Steve Todd's image on a Raspberry 3 B+.?? Everything works great except track power goes off automatically goes off randomly for no apparent reason, especially when programming locos.?? Have to check the Power Control applet all the time to ensure track power is really on or not.

More of an annoyance than anything else.?? Happy with the purchase and glad that I don't need to mess with service mode programming using the Digitrax DCC system ayout any longer.??


Locked Initial setting of sensors in Layout Editor

 

At start up sensors are set to condition "Unknown".
Is it possible to set them to "Inactive" at start up?


Locked Re: Defining a Departure Track

 

Hi Jon,?

Thanks very much.? I've no idea how I missed that.??

Chris