Hi Quantopian team,
like pointed out in this thread, developing a complex algo is very frustrating due to the lack of stacktrace informations when an error occurs. I perfectly understand the security concerns behind this choice, but it fairly simple to implement a stacktrace rendering that trims off lines that don't concern the current file (in python it possibile to retrieve the raw stacktrace components, so no text parsing is needed).
An easy fix that will helps a lot in algo devel, in my opinion. What are your thoughts on the matter?
Thanks in advance.