Gabriel Scherer

I realize now that the tactic is only defined in the FunctionalExtensionality module, so maybe it is intentional that those kind of module-specific tactics are not found in the tactic...

It sounds like there is a mystery here (*somewhere* in the compiler things break if non-deterministic hashtables are used). I'm thinking of offering this mystery-hunt on Discuss as a good...

Nice find! Instead of `~random:false`, I wonder if we should consider more robust solutions: 1. We could use another datastructure than a hashtable, with a canonical iteration order. (For example...

Often we complain that our snapshot-testing strategy is too finicky, but I must say that it is quite nice that the compiler testsuite fails when we introduce build-reproducibility issues (a...

It's not clear from your report how to fix the output sanitizer, (putting myself in the shoes of someone willing to fix this, theoretically) could you include examples of the...

Before we embark on this kind of work -- not wrong, but it takes effort from somewhere else -- do we have concrete evidence that it is worth optimizing these...

__FUNCTION__ was implemented in #9554. Apparently it was already understood at the time that this is in fact more precise than "current function", but there were no clear guidelines on...

This looks reasonable from a distance, but I can't review it in-depth to check that it is correct. Can you explain (and then execute, hopefully) a testing strategy that would...

I am unconvinced by this proposal, but then maybe I don't understand the details. - I don't like the idea that release archives and opam-provided archives would have different content....

I don't want to give the idea that I disagree with trimming down the release tarballs -- I don't have a strong opinion. But my impression from a distance is...