Open eckter opened 1 month ago
Update: now that https://github.com/OpenRailAssociation/osrd/pull/9126 is merged, the error is still present but a lot less frequent.
It seems to come from this case that may cause discontinuities in rare cases. We still don't have a good solution for this problem. If the error is enough of an issue, we could look into manually editing envelopes to have a realistic result with an intersection. But I'd prefer if we could avoid it and just wait for the new simulation.
I'm moving this to minor as it's now very uncommon.
EDIT: actually not all that uncommon because of the number of allowances we run in stdcm post-processing
With https://github.com/OpenRailAssociation/osrd/pull/9137 we've mitigated the impact of this bug on stdcm
It's not quite closed as it can (very rarely) happen in standalone simulations. We've just dodged the problem for stdcm requests, where it has a much higher chance of happening.
What happened?
This error was thrown during the postprocessing.
What did you expect to happen?
No response
How can we reproduce it (as minimally and precisely as possible)?
Can be reproduced with the fuzzer, otherwise it's quite common when trying paths on a realistic timetable
On which environments the bug occurs?
No response
On which browser the bug occurs?
No response
OSRD version (top right corner
Account
button >Informations
)05351ec2518b7555501df832bae2696712164247