Alec Theriault

While producing a test case, I realized __this has nothing to do with Happy 1.19.5 or 1.19.8 or partial type signatures - it is just a straight GHC compile time...

@wiz Do you have a commit of your project you could point me to where you are seeing this longer compilation times? I've tried compiling `master` (of Futhark) with Happy...

@nomeata Do you have an example of a slowdown caused by type-holes? If so, I would like to distill it to a test case for a GHC bug. I haven't...

Thanks @nomeata, I can reproduce now! I've opened https://ghc.haskell.org/trac/ghc/ticket/14766 for this. It looks partial type signatures are an issue. Copying over the table of compile-times from the ticket, I observed...

I want this a lot too for a project of mine whose grammar is quite large, and so I went ahead and took a stab at implementing it. [Here are...

Wait for feedback on design and merge-ability I think.

I think credit for the original `foggy` code and suggestion goes to @glguy. I don't have the original code anywhere, but I was able to recreate the following: ```haskell {-#...

I just ran into this exact bug (and the merged PR solves it for me). I see that the last release of Happy (1.19.5) on Hackage was over a year...

Fixed a while ago, no? https://hackage.haskell.org/package/base-4.10.0.0/docs/Prelude.html#t:Eq and later seem fine.

We're experiencing exactly the same issue (transactions enabled, fileDB, lots of concurrent writes with reads sprinkled in) on MapDB 3.0.8.