Mike J Innes
Mike J Innes
Note that it's possible for someone to grab the commits here and carry on work in a new PR, which might be a good option if people need the patch...
It would be really great to see Ink and Python working together. If you want to get going with this, I think your best bet is probably to not worry...
Yeah, it'll definitely need some updates.
FWIW, I don't think this is an issue addressed by the FAQ; there's no option to rebuild here so it's not obvious that the package needs to be re-installed.
Doing this in a Cassette-like way (generated function + reflection + return CodeInfo) is the right way to go. Cassette itself doesn't let you work on typed IR, but it's...
FWIW, the swap issue here comes from our inlining of SSA values. This can happen across slot reads/writes, which is invalid since it can change side effects. In a case...
Yeah, this is a useful reference, but I don't think it's nearly as bad as it looks. Most of it is runtime support (e.g. GC and IO) that we don't...
> swap them out for our array type Lucky we have a compiler that will make this very easy :)
Since we're immediately throwing out the SSA values themselves anyway, I expect this will be easy to update if it does change. The main change on 0.7 is linear IR,...
Even with phi nodes it should be straightforward to turn them into locals. We'll probably want some way to coalesce and/or remove locals entirely to clean up the code, which...