For debugging it might be useful to know when output was produced. ATM I could pipe output to ts to get stdout annotated. But that would not anyhow help with stderr. Hence I thought it might have been nice to be able to store outputs with
It is useful e.g. to see when command actually output something after "thinking hard". We could then also improve plotting graph with annotation on when tool was producing any stdout and/or stderr output on the timeline of resources consumption.
Sure thing, with buffering all those timings might be "approximate".
For debugging it might be useful to know when output was produced. ATM I could pipe output to
tsto get stdout annotated. But that would not anyhow help with stderr. Hence I thought it might have been nice to be able to store outputs withIt is useful e.g. to see when command actually output something after "thinking hard". We could then also improve plotting graph with annotation on when tool was producing any stdout and/or stderr output on the timeline of resources consumption.
Sure thing, with buffering all those timings might be "approximate".