search

jpata / tthbb13

2 stars 5 forks source link

Dev tf #23

Closed tklijnsma closed 9 years ago

tklijnsma commented 9 years ago

This is the code used to generate the Transfer Function pickle file. It does not interact with the rest of the code, except for outputtree-strict.py, which needs the AccessHelpers from the TTHNtupleAnalyzer.

gkasieczka commented 9 years ago

Just saw: it also add the MEIntegratorStandalone to the .gitignore.

Probably ok for now, but shouldn't this actually be a submodule? If there are no objections I can promote it to one.

jpata commented 9 years ago

I think it looks good. I'm OK with the MEIntegrator being a submodule or in .gitignore. I used to use them in the past, but I found that they introduced somewhat more confusion in merges than they saved in storing the correct hash, but that confusion could probably be avoided with some effort.

See also: https://codingkilledthecat.wordpress.com/2012/04/28/why-your-company-shouldnt-use-git-submodules/

gkasieczka commented 9 years ago

Thanks - I'll merge this then.

Regarding submodules: as our current system is just to clone the most recent version from Lorenzo's repo we probably don't gain much either way.

Maybe, when things are a bit more stable, it will make sense to at least clone a specific tag from Lorenzo so we know which version the code should be on (but this can also be done in setup.sh)

jpata commented 9 years ago

Yep, thanks for merging. In general I like the idea of submodules and fully support using them in this case - just the last time I tried to introduce it to my physicists friends it ended in some very with some very unconvinced people needing to do git submodule update instead of the usual git pull due to my insistence on having new technology. Currently, our approach with the MEIntegrator code is that Lorenzo's master always points to the version that should be used with the master of tthbb13 and the concept of multiple branches does not exist.