On 2020/01/23 20:49:01, hanwenn wrote: > I mean, if we were writing a new SSL library, I surely would want scrutiny on > every cast, but we aren't.
The question is, do you want scrutiny of any cast? If you do, then keeping the warnings enabled and keeping the signal-to-noise ratio high by casting only where it is well justified is the right approach. If you don't care, disable the warnings. > Can you point me to a conversion warning in LilyPond which could cause a > genuine, user-visible problem? I'm not going to spend the time to sift through past commits, current warnings, and existing casts to answer this properly. One involving a shorter integer that sticks in my mind (because I was recently working on resolving it) involves MIDI track number, a 16-bit number. It would not surprise me to find places in the MIDI code where conversions to 8-bit numbers occur; perhaps there's a questionable one in there somewhere. Now, I would prefer to leave off defending my stance on this, and let the senior contributors decide what is going to happen in this project. https://codereview.appspot.com/557190043/
