Jacob Finkelman

Thank you for the prompt fixes. I only had a brief chance to look at this today, and it requires more thought than I had available. I will try and...

Thank you for all the work you put into this! The details here are tricky and overwhelming and I really appreciate you sticking with it. Compatibility with npm is nice...

I doubt there is in the short term. The Cargo Team is ~still~ intermittently trying to figure out what semantics we want. I doubt this will get merged in this...

Without diminishing the pain this is causing. Why has this been such a large issue in our community when many other ecosystems do not provide an opt out? pypi, npn,...

I wonder if the issues with the excessive deltas/commits could also be reduced by starting the process shortly after having done a index squash, and by doing another index squash...

That talk was fascinating. Thank you for sharing and presenting respectively. Here's some of what I got out of it. The Huffman/Arithmetic encoding discussed above reduces the arbitrary case down...

That is a very cool algorithm for reusing the accidental values. How do you construct the encoded keys such that you have a high probability of these happening? > It...

Thinking about this overnight. The math I did yesterday isn't quite helpful. In the regimes where I thought a `ApproxSet` with `epsilon` >= 2 was the smallest option, two or...

You are absolutely correct! Both about my mistake and its implications.

@konstin and @zanieb you have extensive testing of error messages, does this change make things easier or harder to understand? Similarly for your use cases / benchmarks how does it...