19913d85-f662-4518-a9a1-7adc1793dfd4.png){kind=small}
Channels
#announcements
- s
schrockn
09/10/2018, 12:26 PMya - s
schrockn
09/10/2018, 12:26 PMoh totally - s
schrockn
09/10/2018, 12:26 PMdon’t do that - s
schrockn
09/10/2018, 12:26 PMi think a clear error message is sufficient - s
schrockn
09/10/2018, 12:27 PMwhat do you think of @lambda_solid - s
schrockn
09/10/2018, 12:27 PMwhere that all ended up - f
freiksenet
09/10/2018, 12:27 PMI don't like the name, but that's just a knee jerk reaction 😄 - f
freiksenet
09/10/2018, 12:27 PMit's fine - f
freiksenet
09/10/2018, 12:27 PMI think - s
schrockn
09/10/2018, 12:27 PMlol - f
freiksenet
09/10/2018, 12:27 PMI'd call it transfrom - f
freiksenet
09/10/2018, 12:28 PMbut it's not ideal too - f
freiksenet
09/10/2018, 12:28 PMregardless, it's fine - s
schrockn
09/10/2018, 12:28 PMbetter than flags i think - f
freiksenet
09/10/2018, 12:28 PMdef better than flags - s
schrockn
09/10/2018, 12:29 PM - s
schrockn
09/10/2018, 12:30 PMonly question is using DagsterTypeError versus a new exception type - f
freiksenet
09/10/2018, 12:30 PMI also don't like re-raising - f
freiksenet
09/10/2018, 12:30 PMthat happens everywhere - f
freiksenet
09/10/2018, 12:30 PMbut it makes sense to add context wher ethe error is happening - s
schrockn
09/10/2018, 12:30 PMya we have to do that - s
schrockn
09/10/2018, 12:30 PMi just think having a different exception type - s
schrockn
09/10/2018, 12:30 PMmakes it more clear and less error-prone - s
schrockn
09/10/2018, 12:31 PMso the re-raise is for that case and only that case - f
freiksenet
09/10/2018, 12:31 PMwhat would it be called? - s
schrockn
09/10/2018, 12:32 PMso i think evaluate_value is not the best name either - s
schrockn
09/10/2018, 12:32 PMbut if we stick with that - s
schrockn
09/10/2018, 12:32 PMEvaluateValueError or something - s
schrockn
09/10/2018, 12:32 PMjust make it very clear that it is part of the contract of that function - f
freiksenet
09/10/2018, 12:33 PMokay, fair enough