this looks like we are always producing a full dump, and we do it per task, but not tuned to the task?
seems like something we can avoid (if no cli --dump is requested)
apparently only main is actually storing the output of this call, accumulating further to letting us handle that thing strictly in the main
(and similarly with the initial load in fact -- as suggested in #32 )
https://github.com/vliz-be-opsci/py-trav-harv/blob/7088b65776926e73b42184e9b9e6af0e4ceed0c7/travharv/executor.py#L71
this looks like we are always producing a full dump, and we do it per task, but not tuned to the task? seems like something we can avoid (if no cli
--dump
is requested)apparently only main is actually storing the output of this call, accumulating further to letting us handle that thing strictly in the main
(and similarly with the initial load in fact -- as suggested in #32 )