Closed nickgsc closed 6 years ago
hey @nickgsc 🙌
that loooks great, however, is md5
not supported et-all anymore or is it just not the default anymore?
if it is safe to rely on its availble, i would prefer to use md5
as a default, to not break existing cryptex repo's, or do i get it wrong?
implemenation is great, just a matter of default sha256
(which would be better, i agree) vs md5
I had that thought too. I suppose defaulting it to md5 would be better for backwards compatibility of existing repos. It is definitely still a supported digest, it's just no longer the default on OpenSSL 1.1 (when it was on older versions of OpenSSL).
For reference, here are the available digests I was able to identify in the environments I have at my disposal right now:
OpenSSL 1.1.0f (on Debian Jessie):
Message Digest commands (see the `dgst' command for more details)
blake2b512 blake2s256 gost md4
md5 rmd160 sha1 sha224
sha256 sha384 sha512
LibreSSL 2.2.7 (MacOS High Sierra):
Message Digest commands (see the `dgst' command for more details)
gost-mac md4 md5 md_gost94
ripemd160 sha sha1 sha224
sha256 sha384 sha512 streebog256
streebog512 whirlpool
OpenSSL 1.0.2o (MacOS via Brew):
Message Digest commands (see the `dgst' command for more details)
md4 md5 mdc2 rmd160
sha sha1
awesome - many thx for your work ❤️ and your top-notch explanation 💯 going to merge it, and push a new release soon
hey @nickgsc i backported your changes to https://github.com/fastlane/fastlane/pull/12390 match
- where we forked cryptex initially
ohhh i see it already was there :( my fault.
I didn't realize that this had been forked from match, so I didn't even check that repo for a solution. I just rolled the fix directly here.
In any case, glad we have parity on that now between the two!
Fixes #9