John Ingve Schjølberg

Looking at the `SubConfigurations` function, the `mgr` parameter could easily be replaced by an interface returning the raw manager instead of passing the raw manager directly: ```go // AsRawManager provides...

Of course, one solution is to move this function into the generated code, but that would necessitate changes to the `SubConfigOption` interface and further increase the number of APIs exposed...

Yes, that sounds like a good idea. It would also help with making the API clearer since the call types are meant to be used by generated code.

> I'm curious whether you are using a tablet, or whether Chromium thinks you are using a tablet (or in tablet mode) when you're not actually using one? That appears...

I think you might be able to implement this by overriding the acceptor and command queue modules. The command queue sends a "checkpoint" command every k views or if the...

Looking at some profiles, it doesn't seem like the conversions are slowing things down much. Here's a memory profile: ```text File: hotstuff Type: alloc_space Time: Apr 16, 2022 at 12:03pm...

Hi! > Namely, the methods CreateThresholdSignature() and VerifyThresholdSignature() actually create (and verify) an aggregate, not threshold, signature, which is insecure when partial signatures sign the same message. Indeed, the `bls12`...

> I wonder if we should at least rename `CreateThresholdSignature` and `VerifyThresholdSignature` to `CreateMultiSignature`and`VerifyMultiSignature`. I discussed this issue with @meling yesterday, and he suggested refactoring the `CryptoImpl` interface such that...

Also, we should have a new name for the interface that abstracts the multi/threshold signatures. Does `QuorumSignature` make sense? Edit: Or maybe we can just call it `Signature`

Running into this same issue with [PoCL](https://portablecl.org/) also.