Joe Eli McIlvain

Is there anything the community can do to help with this? It seems we're waiting on the source code (#9) to proceed if the maintainer doesn't have time to do...

Note that I've also opened an identical PR in the upstream repo: https://github.com/AlloyTools/org.alloytools.alloy/pull/182

Note that I have tested this locally (by rebuilding the `.jar` on my fork, and copying it into `~/.vscode/extensions/arashsahebolamri.alloy-0.7.1/org.alloytools.alloy.dist.jar`) and it works correctly: 

This was discussed in the August 2 sync call, and I had the following notes in the Zulip thread: > I haven't proven it, but I think what's happening is...

Looks like this has been a bug since pretty much the very beginning (8 years ago). I would take the position that silently truncating the range is the most consistent...

Yeah, we definitely don't want to leave gap in between in whatever the desired behavior for this method is. We discussed this a lot on the sync call and my...

After discussing this topic with Sean today, here's a path forward that at least he and I are in agreement on (others here can chime in as well). This mostly...

Sounds great, if you're willing to continue on that path. Thanks!

That's an interesting point, and I definitely like the reduced conceptual complexity of that approach (especially that it avoids the mental overhead of remembering which cases truncate and which cases...

@SeanTAllen - I don't understand - maybe I'm missing something. To elucidate, are you able to produce a code sample that reads from uninitialized memory without using FFI?