Morten Dahl
Morten Dahl
I would say "fix later", wdyt @jvmncs ?
Some references: - https://endler.dev/2020/rust-compile-times/ - https://matklad.github.io/2021/09/04/fast-rust-builds.html - https://github.com/Swatinem/rust-cache - https://pingcap.com/blog/rust-compilation-model-calamity - https://github.com/RazrFalcon/cargo-bloat
cc @voronaam just to make you aware about this one; some of the ops in https://github.com/tf-encrypted/runtime/pull/508 would likely benefit non-trivially from this since a lot of projection operations are being...
This was mostly done for `AsyncSession` and `SyncSession` by switching to `ArcArray`, but something similar could be done for `SymbolicSession` to potentially reduce Moose compilation time
Quick idea for how to think about dialects/levels (some inaccuracies, bare with me) 
Below is my proposed design for sub-computations. It is based on generalizing a computation to not be a single graph, but rather a labelled set of graphs. The graph with...
Jamboard: https://jamboard.google.com/d/1Dr294lS34XxOwrxDImaRFwNl0w280ba-mjUn5ss1LCc/viewer?f=0
@rdragos @voronaam @jvmncs maybe something to discuss next week
@rdragos this might be something for you to spending a bit of time thinking about next week
cc @voronaam