On Tue, May 7, 2013 at 12:01 AM, James Tranovich <ja...@openhorizonlabs.com> wrote: > Hello all, > > What can I do to help make rusti "halfway tolerable"? > > Not to belabor the point, but rusti definitely needs some love -- my git > version of rusti as of one or two days ago always crashes with > "memory_region::~memory_region(): Assertion `false' failed." It will return > the evaluated expression correctly before that, though, most of the time.
That's one I haven't seen yet. Can you file a bug tagged with A-rusti? > I stopped before diving into that particular rabbit hole, because it seems > that rusti might need a more invasive and "proper" fix. I think you're right that we may need to rethink the whole approach to rusti. Right now, it literally just takes the user's input, wraps it in a main function that it gins up, and JITs that. For example, "let x = 3;" will produce a "warning: unused variable: `x`" error, which would be great if you were compiling `fn main() { let x = 3; }` but hardly the right behavior for a REPL. > Just wondering what the general plan of attack was and if there was anything > I could do to help. We have a plan to make a plan, but no plan yet. But an obvious problem that we can do something about at the moment is that there are no tests for rusti (https://github.com/mozilla/rust/issues/5469), which astrieanna and I are (slowly) working on fixing. Lindsey _______________________________________________ Rust-dev mailing list Rust-dev@mozilla.org https://mail.mozilla.org/listinfo/rust-dev