suspicous NEMA phase names

[namdre]

In the complex test results (i.e. complex/traci/trafficlight/NEMA/offset), the states change but the phase names do not:

state="grrrrrrGGrrr" name="4+4"/>
state="grrrrrrGGrrr" name="4+4"/>
state="grrrrrrGGrrr" name="4+4"/>
state="grrrrrrGGrrr" name="4+4"/>
state="grrrrrrGGrrr" name="4+4"/>
state="grrrrrrGGrrr" name="4+4"/>
state="GGGGGrrrrrrr" name="4+4"/>
state="GGGGGrrrrrrr" name="4+4"/>
state="GGGGGrrrrrrr" name="4+4"/>
state="GGGGGrrrrrrr" name="4+4"/>

Since the first traci action comes at step 30, this seems to be an issue with the basic operation (or maybe a faulty configuration?).

@qichaow @mschrader15 can you take a look?

[mschrader15]

I will look into this

[mschrader15]

@namdre It looks like that has been an issue since commit 77cdc5e24bd6bc15cee3db98f9e9fa134d7580f1 tests/complex/traci/trafficlight/NEMA/offset/tls_state.complex (here for reference)

I think it was an error in initialization that I thought I solved with Lines 607-609 in 1cc8ec9. I will do some more testing to confirm that I fixed this.

[mschrader15]

The same error was actually present in all tests https://github.com/eclipse/sumo/blob/01964e4731d18e4697c8476edb05686a2295c2c2/tests/sumo/basic/tls/NEMA/basic/tls_state.sumo

[mschrader15]

@namdre I am not seeing the error in tests run on HEAD. I believe that 1cc8ec9 resolved this.

[namdre]

You are right. Sorry for the bother )-:
I was looking at those tests to see why they produce different results after your commit but failed to see that it solved this very problem.

[mschrader15]

No problem!

For future reference, do you prefer PRs that address multiple issues or should I try and keep one PR per Issue? My mess of PRs yesterday was pretty confusing lol

[namdre]

One PR per issue would be preferable but it's not an iron rule. Sometimes one simple change fixes multiple bugs and it would be silly to try and separate it.