Roderick Chapman

Results 29 comments of Roderick Chapman

I also get this unimplemented feature reported on s2n-quic/quic/s2n-quic-core/src/counter.rs

+1 on this issue. I just read through this document and found the same issues with the safety macros. In addition: 1. The file structure has changed to add s2n_safety_macros.h...

I just did a build and test with GNAT 13.2.0 on x86_64 Linux, and all good.

Code which is common to _all_ hash algorithms (including RFSB) should go in the private part of SPARKNaCl.Hashing (like the TS64 function, for example.). Code that is common to all...

I agree - I think those short names came from TweetNaCl, so aren't really appropriate here, so please rename them as you see fit.

I think the default is "portable and should work on any 32 or 64 bit machine (but favour 32-bit if and where you can...)" Does that make sense?

I think I agree with Max - we already have more than one "family" of hash functions in the library - SHA2 and RFSB, so the package structure should reflect...

I'll take a look when I have a sec. I think it will be worth making the transition to the new syntax if and when Ada2022 becomes the default mode...

Where is that name-mangling documented? I couldn't find it in the user manual...