Recent searches
Search options
My coworkers ask for help with a #Rust compiler error... *once*...
...and then I run and try to mechanize my explanation and put it back in the compiler.
Our dev tools need to talk to humans in the way humans talk. Nobody is an expert on everything, helping newcomers (with better tools, better docs, better errors) helps *everyone*.
That's really the trick behind the rustc diagnostic output: it's not about a technology, its an attitude.
rustc doesn't have good diagnostic messages because of some technological breakthrough. It has good diagnostic because we look at how people use our tools, noticing when things confuse people, and changing the output to match people's mental model.
@ekuber I love this attitude and I aim to emulate it.
I do technical standards work and one of the ways I’ve adapted that is that I’ve stopped accepting the preface “this might be a stupid question…”
I interject that you have the luxury of asking me (or the group) about ___. The next person reading the spec might have the same question and not have that luxury.
Please ask your question for them and let’s make it more obvious.
@ekuber@hachyderm.io@scouten YES! For every one "this might be a silly question" there are another ten out there not asking the question but just as confused. There's no harm on iterating to make things clearer.