4 ms·
You write "intended for people with little background on crypto to be able to use it without being concerned about long configurations" but the documentation sa
by eesmith 8y ago
You write "intended for people with little background on crypto to be able to use it without being concerned about long configurations" but the documentation says "intended for educational purposes only".
These are not compatible.
You write "insecure alternative to OpenSSL". Long experience says that it's hard to turn an insecure system to a secure one. Why would people with little background on crypto want to use an insecure system?
Why is this not built on top of one of the existing cryptography packages for Python? Quoting comments from https://www.reddit.com/r/Python/comments/5gn3q8/what_crypto_library_should_one_use_in_python/ https://www.reddit.com/r/Python/comments/5gn3q8/what_crypto_... :
> Cryptography seems to have a good understanding of "you shouldn't roll your own crypto", and the main devs appear to be pretty well known in the crypto community. ... It is also feature-complete if you need to use any of the hazmat low-level stuff ... I was really satisfied with it. Straightforward API, good documentation, easy to find examples online, no problem with it so far.
> If you want to use a modern, safe library where it's improbably that you will break your encryption by making a stupid mistake, go for libsodium (PyNaCl or libnacl, the latter has a worse interface but exposes more low level functions) ... NaCl was written (by Daniel J. Bernstein, a renowed cryptographer) specificially to make it easy for non crypto people to get safe encryption.
Don't get me wrong - building your own system is a great way to understand the problems involved. But your statements make it seem as if you are trying to solve a real problem, and I don't think that problem exists.
- orancanoren 8y agoThis project started only for educational purposes but lately I have been thinking whether it could become more than that or not. I do not have what it takes to build a secure library myself, I don't have the necessary depth of background and seeing all the other libraries where it is a hassle to integrate into your system, I thought maybe the community may want to use something simpler. Other than that I completely agree with you that building your own crypto is `the root of evil`. Either way, I do appreciate your input on this.
- eesmith 8y ago"all the other libraries where it is a hassle to integrate into your system" Is the 'cryptography' package a hassle to integrate? The quotes I gave say otherwise. From what I hear it is also simple to use. If that is your goal, then you need to make a persuasive argument that there are usability problems with the existing libraries (even those libraries which are designed to be easy to use), and that your package will be easier. Is there any reason to not build an easy-to-use package on top of, say, the 'cryptography' package, perhaps by using the 'hazmat' subpackage? I do not think that "building your own crypto is `the root of evil`". I don't know who you are quoting. The original Biblical reference assigns that to the love of money.