soulomoon
They are pretty dependent on the hieAst. Since we are hover only on the `print`. The `print @Int` as a whole has `types: [5]`. Actually, the type follows the evidence...
You can setup ghc session and build the HieAst for a simple file @jhrcek , without the HLS. It should speed up investigation on hieAst. I hope it helps. ```...
> Awesome, thanks! ~I'll play around with this snippet.~ I turned your snipped into runnable script for local poking around. It required me to declare Outputable instance for HieType to...
> I figured out a nicer way to show hie file (with full types reconstructed instead of with `TypeIndex`es) and it seems that in both cases the type info is...
WIth ``` - name: cabal package: Cabal version: 3.6.3.0 modules: - src/Distribution/Simple/Configure.hs ``` `cabal bench -j --benchmark-options="profiled-cabal"` | version | configuration | name | success | samples | startup |...
> the `ghcRebuilds` number seems to indicate that we aren't editing the file in the benchmark passes. Could you ensure that each benchmark iteration also edits the file? Sure, I...
I need to take some more time to come up a more decent benchmark on this. Convert it to draft now.
The overal improvement is not really obivious. Slightly better mem usage and speed. | version | configuration | name | success | samples | startup | setup | userT |...
After update to master, seems it does show decent improvement. | version | configuration | name | success | samples | startup | setup | userT | delayedT | 1stBuildT...
Another run, the overall result is similar, but the detailed traces comparison indeed show some decent improvement. | version | configuration | name | success | samples | startup |...