On Thu, Jun 12, 2025 at 08:13:08PM -0400, Andres Freund wrote: > FWIW, I find it utterly unsurpising that new users of %llu were introduced > after 15a79c73111f. For one, 15a79c73111f explicitly says "(minimal trial)" in > the subject line, it'd have hardly been sensible to introduce PRI* uses at > that point. Only a0ed19e0a9e, changed most of the uses (after e.g. the > io_uring uses were introduced). For another, how is everyone even supposed to > even have known about this new policy, it was just discussed in a long thread, > at a very busy time? Even now I suspect we'll grow more %ll[ud] users, it's > after all what we've been trained to do for quite a while now.
I don't think it is possible for everybody to be able to track that, so we can be really flex about that. So relying on individuals to notice it something when it is noticed is OK, and this was new code. In this case, the only reason why I bumped into this is pure luck: I was working on some backend area and wrote code that did the %llu with the cast dance, because I'm educated the same way as most people to make the code portable (use %llu and %lld, not the new thing). Then while grepping for similar code patterns and the git history, I've bumped into Peter's commit and just noticed the remaining %llu. And this was on new code, so fixing it seemed just the right thing to do at this stage. -- Michael
signature.asc
Description: PGP signature
