Extractor tasks are no longer class attributes but rather object attributes. I have removed the extractor tasks list from most task protocol objects as the defaults are handled by ibllib. Note that the task in question is dynamic depending on the main sync anyway.
For more than 1 DAQ type we may need ibllib to infer the task type from the experiment description sync key, or by having make_pipeline use some task type map defined in the project_extraction repo.
Extractor tasks are no longer class attributes but rather object attributes. I have removed the extractor tasks list from most task protocol objects as the defaults are handled by ibllib. Note that the task in question is dynamic depending on the main sync anyway.
The following solution works for Bpod vs 1 other DAQ: https://github.com/int-brain-lab/project_extraction/blob/6d34f1769fed5c60a4dc4eb5bfd860d6d040e437/iblrig_custom_tasks/samuel_cuedBiasedChoiceWorld/task.py#L25-L27
For more than 1 DAQ type we may need ibllib to infer the task type from the experiment description sync key, or by having
make_pipelineuse some task type map defined in the project_extraction repo.