Long block signals behave differently in Experimental?

[scamps]

I tried the "put 3 stations between long signal loops" thing, which works in Standard. It did not work out in Experimental.

Attached two test saves - Standard (pak Britain) and Experimental (11.12).
https://www.dropbox.com/sh/7xsv2xemfdc6jwc/ANDUQOXc-t

If long block signals sould behave differently in Experimental, where can I find correct description?

Thank you.

[whoami]

Perhaps the same bug as discussed here: http://forum.simutrans.com/index.php?topic=12393.msg126489#msg126489

[scamps]

bump?

Edit: @whoami
OK, I followed the link you provided and thought that the mentioned bug was fixed prior to 11.12. Maybe I have misunderstood.

Editing post not to bump the thread again.

[whoami]

Hehe, I am going to bump you back. 
Uhm, I mean: We have to wait for the next ST-exp version.

[scamps]

New behavior in 11.13: now even one train can not travel this line, "wating for clearance"

[whoami]

For this, there might be a simple explanation: sometimes (also in ST-standard), especially on loading a savegame, an invalid reservation remains on track tiles. Use 'b' to view and fix the reservations.

[Carl]

Long blocks look to be working fine for me in 11.13.

[scamps]

Then it means I use signas in some wrong way. After upgrading to 11.13 I made a new line, but it made no difference. Train stands at long block signal, does not reserve anything and waits for clearance.

I would appreciate if someone checked out the save and explained what is wrong with this cofiguration.

[Carl]

I'm not able to look at the save right now, but this reminds me of a problem I had a few months ago. Does this help? http://forum.simutrans.com/index.php?topic=11749.msg115576#msg115576

[scamps]

Useful info, Carl, but does not seem to be the case.

[jamespetts]

Carl - thank you for testing.

Scamps - if you are still having problems, can you upload a saved game in which you can reproduce something that is clearly incorrect, and describe the incorrect behaviour precisely? That would be very helpful.

[scamps]

First post contains a link to two saves. One Standard game, one Experimental game. Same track confguration.  In Standard two trains can use the line, in Experimental 11.13 a single train can not use the line, "waiting for clearance" at long block signal.

And repeating the link   https://www.dropbox.com/sh/7xsv2xemfdc6jwc/ANDUQOXc-t

edit: failed to include screenshot directly in post, added under the same link.

[whoami]

I tested your savegames (I did not before). The funny thing is: the savegame from ST-std works fine in ST-Exp 11.13! signal-exp started to work on the left end (only) once I added signals on the bypass tracks and removed the "mirror" flag from the line schedule and the "reverse" flag from the trains. signal-std has a circular schedule, but changing it to a bi-directional schedule (still in ST-Exp) does not prevent it from working.
In signal-exp, I also extended the stations in a way that train length should not matter, but it did not help either. Forcing the trains to proceed or return to the previous stop does not prevent the problem from reappearing. No stale block reservations are visible.
The differences between the two savegames are schedules, train and station lengths, vehicle types, also slight differences in track layout. I still cannot exclude any of these as the reason for the misbehaviour.

[jamespetts]

Hmm - perhaps I have misunderstood the intention of long-block signals when applying changes to the routing and block reservation systems. Is the idea to allow one train to follow another into the same section, separated only by station stops rather than explicit signals?

[TurfIt]

No, rather the opposite. Long block should see through stations so reversing trains don't deadlock.  Let's you implement a 'One Train Working' system on a single track branch with multiple stations.

[jamespetts]

Hmm - so how should this example behave, in that case?

[scamps]

I used this as an example on long block signals
http://forum.simutrans.com/index.php?topic=10472.0

[jamespetts]

I have just fixed a bug relating to long block signals that would in some cases cause convoys to become stuck behind a long block signal unnecessarily, which I hope will fix this issue.