result in links being added, and running the same process twice gives different IDs to different links. The link IDs are currently the smallest integers that are not currently being used in the network - so the method generating IDs should already be giving the same set of IDs (given the exact same set-up and inputs).
The problems I see:
when these IDs are distributed to the different links that are being added there is no order to who gets which ID - hence different answer each time
we sometimes make small changes, it would be good to have more meaningful indexing that would persist even when input/setup are not entirely the same
Right now processes in GeNet such as:
result in links being added, and running the same process twice gives different IDs to different links. The link IDs are currently the smallest integers that are not currently being used in the network - so the method generating IDs should already be giving the same set of IDs (given the exact same set-up and inputs).
The problems I see: