OK, I don't know how to produce a live video of my screen in W

10, or how to produce a dump/stack trace. If someone can walk me

thru either or both (and I can remember exactly what I did just

before the POOF...) I'd be glad to provide that information. Most

likely to happen when modifying the path of line 2 created from

paralleling line 1, since line 2 is an offset copy of line one

(and attributes of the pins in line 1 are not copied)  it usually

requires modifying when creating roads.

The one feature I'd really like to see is the ability to split

lines like you can tracks to create road intersections and while

at it, an easier way to add pins to a track (An "add pin"

selection on the right click menu in modifying lines/tracks?).

Thanks again for all your hard work!

Could someone please walk me thru downloading and installing

Xtracks (with latest Beta) on my linux mint v.19.1 desktop

computer as I would really like to work on my desktop and large

monitor than on my small laptop running win 10. Plus I could get

the laptop and it's second mouse off my desk.

On 8/27/20 10:58 PM, Adam Richards

wrote:

Ian,

None of those terminations should happen - likely

the issue is not a user error but a system error. 

To diagnose we either need a methodological

summary (first this, then this) or we need a dump/stack trace to

see the offending code.

For the first type, a video would do. 

To get better restart - set the checkpoint

frequency to 1 - and the auto save to 1 or 2. 

The labels should be saved when the tracks are

saved - so either auto save or explicitly save should make sure

that are saved correctly. We might insert an extra save after

leaving manage layers, I guess.

Adam

On Thu, Aug 27, 2020 at 6:15

PM, scenicsRme via groups.io <playswithtrains2=aol.com@groups.io>

wrote:

I'm still getting crashes whe working with parallel

lines, both





creating and modifying after creating. Yes, I wish it

gave a





warning that it did something illegal and gave an

opportunity to





save, but as said, you click on something(???) and POOF!

you're





staring at your desktop. When you restart the program it

asks if





you'd like to continue, but what comes up is usually a

number of





operations in the past, which is aggravating in itself

when you've





been working to perfect a line path, say, but it also

drops the





layers labels (keeps numbers and colors), you have to go

back into





manage layers to restore them and the most irritating

for me is it





restarts at less than full size so the status bar

doesn't show.





You have to minimize the screen, then maximize it to

restore it to





full size to make the status bar visible again. Note,

these are





observations not complaints since they have work

arounds.

On 8/27/20 10:35 AM, Adam Richards





wrote:

Ian,








You said the magic word - "Unload".   

















Found a problem with param files that include Adjustable

Track -





wrong assumption made - trying to free storage starting

at zero -





system abend due to addressing negative memory... 

















Fix pushed. 

















Adam

--

Adam Richards

Gene,
I think significant barriers are -

• (AFAIK) no one has documented the .RRC format,
• There is no open-sourced code to read,
• I think it is about 12 years since there was a code update which was when it was sold (so the developers are probably not available). 

This would reduce one to trying to reverse engineer via experimentation - which is a LOT of work.

One of the things you get with your "free purchase" of XTrackCAD is open extensibility and the prospect that, even if it ceased to be maintained at some point, someone would be able to put together a format converter with a sensible amount of effort based on a documented format (and maybe even some code reading). That's how we have a converter to JMRI Panels, for example, someone independently took up the cudgels. 

I know this is likely disappointing news. The best I can offer, I'm afraid, is that if you can print a plan view in 3D RR, in V5.2 you can import it as the background layer image in XTrackCAD and "trace" over it with tracks. 

Adam

Ian,

sorry but I can't reproduce that problem with 5.2 on Windows 10 either.

Regards,
Martin

Am 26.08.2020 um 16:36 schrieb Adam Richards:
> 1. kato-n.xtp has some definitions where the path statement is
> incorrect. It would not work well in Train mode. There are track
> segments not in path statements.  These have been corrected in N-Kato
> Track & Structures.xtp
>
> 2. The reason the auto-upgrade did not work is twofold. The name is
> changed (which should have resulted in a "Not Found" with an "OK"
> button) and crucially that I forgot that Windows uses "\" instead of "/"
> as a path delimiter....  I have a fix for that.
>
> 3. The abend is different. I can't get that to happen with V5.2 forced
> to load the V5.1.2a file - I get the usual CheckPaths() "Path error"
> messages and life goes on - at least on Mac.  So that remains a mystery
> - perhaps Martin can try to reproduce?

Hi Adam, I have tried to copy and rename the files back to the originals but that doesn't seem to work. I have also tried to set the system so that it starts with a new clean trackplan. Unfortunately it sticks at the Tomix N parameter. Sigh.

No, there was no warning, just no program. Gone in a Flash!

On Wed, 26 Aug 2020 at 11:54, Adam Richards <adamjmrichards@...> wrote:

Actually, the way this is supposed to work (now Beta 2) is that if the program sees a param file that was open in the product params library, at startup it will try to substitute the current level params library seamlessly.  This was to cope with the very type of error here where an error would be fixed with the latest version. The problem here, though, is that the kato-N.xtp file at v5.1.2a has been renamed to N-Kato.xtp at some point for v5.2.0 Beta2. So that logic won't work to get past this as intended. 

I'll let the team decide if we need a conversion table or logic of some sort to find the right file, or if that rename is causing more harm than good - especially now we have Martin's lookup code to find param files by their Content field. 

The CheckPaths() code was added and is performing more checks than it used to - to try ensure that all track segments are in at least one path.  That is the warning error here.  I'd have to guess that problem was fixed for N-Kato... 

The terminating error is most likely something else in that old file that is not in accordance with the latest level.  Did you get an actual error message before shutdown? 

Adam

--

-- Ian

Actually, the way this is supposed to work (now Beta 2) is that if the program sees a param file that was open in the product params library, at startup it will try to substitute the current level params library seamlessly.  This was to cope with the very type of error here where an error would be fixed with the latest version. The problem here, though, is that the kato-N.xtp file at v5.1.2a has been renamed to N-Kato.xtp at some point for v5.2.0 Beta2. So that logic won't work to get past this as intended. 

I'll let the team decide if we need a conversion table or logic of some sort to find the right file, or if that rename is causing more harm than good - especially now we have Martin's lookup code to find param files by their Content field. 

The CheckPaths() code was added and is performing more checks than it used to - to try ensure that all track segments are in at least one path.  That is the warning error here.  I'd have to guess that problem was fixed for N-Kato... 

The terminating error is most likely something else in that old file that is not in accordance with the latest level.  Did you get an actual error message before shutdown? 

Adam