lhames added a comment. I missed this thread earlier -- thanks to Dave for pointing me to it.
@kuhnel -- Thanks very much for working on this. Out of interest, did you see https://llvm.org/docs/ProgrammersManual.html#error-handling ? If not (and if you find it helpful) then maybe we need to make that document more discoverable. If that document is not helpful then we should improve it. > I guess the major difference in perspective may be due to different valuation > of the cost/"disease" it's intended to prevent - for myself I think the cost > is worthwhile but I understand that's not the case for you. I've certainly found Error / Expected to be valuable for my work, but I designed them with my use-cases in mind so take that with a grain of salt. ;) As a general observation: critiques of Error/Expected include the avoidable ergonomic issues (potentially addressable by API improvements), unavoidable ergonomic issues (we really want language support for this, but can't have it), and questions around correct application of Error/Expected -- sometimes error handling is just hard and Error/Expected feel awkward because they force you to confront it, and other times they're just not a good fit for whatever you're trying to do and you have to recognize when they should be dropped in favor of a more appropriate tool. Especially on that first point (the avoidable ergonomic issues) I think that Error/Expected could benefit from review now that we've gained more practical experience with them. I'm not in a position to drive any such effort, but would be happy to help out. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D105014/new/ https://reviews.llvm.org/D105014 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
