David Allsopp

> What you say makes sense -- and indeed matches my limited experience. That is, when I edited the compiler and then ran `make; make install`, running `./ocamlc` witnessed my...

@mshinwell remembers trying to tackle this with one of the debugger PRs. I think it's a good idea, but I expect we'll have to do this as we did with...

> Indeed that would be a good backward compatible way of fixing this, although keeping the `system` variable around carries a maintainance burden and we'll have to port each use...

Kind of a lot's happened in the last year 🙂 I don't think the lack of forward motion on this issue is a sign of a lack of importance, so...

I have been hacking (in fact, I am literally just now hacking) native cross-compilation into the compiler - the patch-set I'm working on at the moment does start to introduce...

I did some experiments today - it is reasonably non-invasive to add detection for this situation (using a very similar trick to what `runtime/Makefile` does with `primitives` and `primitives.new` and...

The bytecode compilation I think always proceeds with the boot stdlib - the only time `-I stdlib` is used in the toplevel is to do native compilation - which is...

In between things, I've been hacking a prototype, which suggests the diff is not as big as might be being feared - in fact, so far it's easier to debug...

I’ll push it and sync up with @shindere in an hour or so

> Suggestion: could someone who feels more comfortable than me with GitHub's interface perhaps assign this to me explicitly? It's not a commitment that I'll act on it in a...