search

star-bnl / star-sw

Core software for STAR experiment
26 stars 63 forks source link

StTriggerData2022 version number wasn't incremented #609

Closed genevb closed 6 months ago

genevb commented 7 months ago

A user sent me a log file of processing a MuDst from FastOffline of Run 22 data for which running his macro produced the following error in SL23d:

The StreamerInfo of class StTriggerData2022 read from file /star/data100/reco/production_pp500_2022/ReversedFullField/dev/2022/078/23078044/st_physics_23078044_raw_3500050.MuDst.root
 has the same version (=1) as the active class but a different checksum.
 You should update the version to ClassDef(StTriggerData2022,2).
 Do not try to write objects with the current class definition,
 the files will not be readable.

I spotted this update, which didn't include a ClassDef incrementation and was made after Run 22 finished: #450

I asked the user to run their macro in SL22c, which was from before that update to StTriggerData2022.h, and the error went away. So it probably should have been incremented.

-Gene

akioogawa commented 6 months ago

I'm puzzled. #450 did add non-IO variables with //!, thus won't need version change?

genevb commented 6 months ago

I see your point, @akioogawa, and I don't have an explanation. I believe you are correct that changes to non-streamed (or "transient") members don't require a streamer version change, so in principle this shouldn't be needed. I did not provide a "how-to-reproduce" for this as the same user is still having other problems with their macro and opened Issue #611. I don't know if anyone else is reading Run 22 MuDsts and seeing the same messages or not, but in principle it should be straightforward to test whether incrementing the version number truly resolves it.

plexoos commented 6 months ago

Fixed in #612