Improve throughput

[thumphries]

Here's a bit of low-hanging fruit:

• Do some time profiling in a large codebase like amazonka
• Measure the impact of -O2
• Parse and search batches in parallel, measure impact

[justanotherdot]

What's the best way to test this on a large base with hgrep? afaik, I always have to specify a Haskell file to hgrep (can't just glob easily, especially if there are non-Haskell source in the path).

[thumphries]

I'd recommend Gabriel Gonzales' bench tool (hackage name is just bench) and something like

cd amazonka

hgrep Instance /**/.hs

(Sorry for formatting, iOS keyboard seems to have no backtick)

On 6 Oct. 2017, at 2:36 pm, Ryan James Spencer [email protected] wrote:

What's the best way to test this on a large base with hgrep? afaik, I always have to specify a Haskell file to hgrep (can't just glob easily, especially if there are non-Haskell source in the path).

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

[mitchellwrosen]

I did a simple profile run and 99% of the time in spent parsing modues. Concurrently parsing would therefore help a lot.