Jeffrey Walton

@chittti, We recently made the necessary changes to support the newer KDFs, like Scrypt and Argon2. Also see [PR 611, Add KeyDerivationFunction interface](https://github.com/weidai11/cryptopp/pull/611). We cut-in Scrypt and Argon2 is next....

@Bweyer6, I looked into Argon2 a while ago. The problem was/is, the code is very complex and I have not mustered the energy to untangle it.

Thanks @yahyazadeh, I'll have to do some digging into this issue. Sometimes sharp edges surface because of the standard used in the library, like P1363 or ANSI X.9 versus the...

@exception0x876, Please forgive my ignorance... Is that the BSD's device? Or does the Armada 375 have special circuits providing AES? (Or maybe a combination?) The reason I ask is, I...

> The device has so called "Security Acceleration Engines" Ack, thanks. I'd be interested in it, too. We started on some of the specialized ARM implementations. The holdup is the...

> The errors above happen when including the includes to crypto++ into an project with the latest c++ preview version (aka c++ 20 on MSVC at the moment). I don't...

@AraHaan , Are you experiencing the issue with the latest version of Visual Studio? According to [Release Notes Preview](https://docs.microsoft.com/en-us/visualstudio/releases/2019/release-notes-preview), that is Visual Studio 2019 version 16.8 Preview 3.2 (September 29,...

> Would it not be much better to clean this up? I believe this was (is?) a design decision. The flat directory structure means GDB and some other braindead tools...

> VectorSource and/or VectorSink cannot compile when passed a vector that takes a custom allocator. > In my case I have a codebase that uses a secure allocator for all...

@ehildenb, @dwightguth and @gtrepta, Yeah, we should be able to supply the function stand-alone with adjustable rounds. Give me a few days to look into the changes. Many of the...