Closed giuliolunati closed 5 years ago
Rebased and merged... I really like this idea of us setting up some kind of mechanism to get error reports back! Here are some thoughts:
This will apply whether we use a whitelist ("you didn't whitelist something you should have") or a blacklist ("you blacklisted something you thought you could, but couldn't"). One thing that makes me suspect we might need a blacklist is because I'm not sure how this is going to work with tracing (for example). I don't completely understand what the stack will look like during a trace, where the print of what operation is executing is "running during any native or dispatcher". :-/ Will that mean we need all natives to be whitelisted? Or will the N_XXX not be on the stack? We will have to see when that gets further along...
But regardless: as I mentioned, when putting out the tutorial, people aren't necessarily going to be telling us when it didn't work or they got an error. They will just click to some other site! So it's important to have some feedback to know if there are client errors that are killing the interpreter...not just these emterpreter errors. Maybe if we get some experience with this error reporting method it could be used there.
Sorry I'm still slow...now in Philadelphia, just saw Chris last night at a Meetup!!
if error is caused by non-emterpreted functions, report them