On Fri, Dec 15, 2017 at 04:40:42PM +0100, Emilio Cobos Álvarez wrote: > Hi, > > Tigercosmos sent a patch today adding #![deny(warnings)] to a bunch of > crates[1]. > > Just wanted to give a heads-up / ask whether there's any objection to > the change before going ahead and r+ it. Does it sound reasonable to > everyone? It may make the servo-with-rust-nightly build fail a bit > earlier, but probably that's good, actually.
>From a Firefox point of view, #![deny(warnings)] is already causing problems, and I don't think it's a good idea to have that in source, essentially for the same reason it's not a good idea to have -Werror set by default on C/C++ projects. Now, to be more specific about what problems we're already seeing with #![deny(warnings)], they are around the fact that when you bisect, the build fails on older code that triggers warnings that didn't exist in the stable rust compiler back when the code was written, but that do in the newer compiler, which you're using while bisecting. I'd actually go as far as saying that #![deny(warnings)] shouldn't even exist in rust at all for this reason, and CIs should be setting the equivalent to -Werror (which I think -D warnings is). Mike _______________________________________________ dev-servo mailing list dev-servo@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-servo
