Over the last year I’ve been trying to understand why GPG isn’t popular. Based on the features I think it’s a pretty valid thing. This article changed my mind.

Turns out GPG is too old ¯_(ツ)_/¯

I like signing my commits, it feels good to know that my identity is actually attached to my code. So I put in some work to reconfigure git to use a different signing tool, I didn’t think it would be such a big deal, turns out git fully intergrates GPG. I’m confused. Why does git need to be hardcoded to use GPG specifically?

What rule says we can’t have git configs like:

[sigining]
  defaultMethod=minisign

[signing.minisign]
  always=true
  signCommand=minisign -S -s {secret-key-file} -x {sig-file-name} -m {target-file}
  verifyCommand=minisign -V -P {public-key-file} -m {target-file}

Where the verifyCommand exits 0 if the signature is good and 1 if not.

I’m open to hearing cons. These are some I can think of:

  • User’s have to configure git with each signing and verifying program
  • Upstream security conserns from signing programs
  • Signing programs changing their interfaces
  • SteleTrovilo@beehaw.org
    14·
    1 year ago

    The article that changed your mind really shouldn’t have. It’s mostly full of hyperbole. Like this:

    “PGP does a mediocre job of signing things, a relatively poor job of encrypting them with passwords, and a pretty bad job of encrypting them with public keys. PGP is not an especially good way to securely transfer a file. It’s a clunky way to sign packages. It’s not great at protecting backups. It’s a downright dangerous way to converse in secure messages.”

    Literally none of this is true - the author is presenting their particular opinions as general fact. I use AES through PGP, knowing that even future quantum computers can’t break it.

    I wish they’d cut out all the 90’s references and pointless exaggerations, and stuck to facts. Then again, the facts-only version of this article probably wouldn’t make a strong case against PGP.

    (Also, one of the links in the article, with the dodgy-and-harmful link text “Full disk encryption isn’t great”, includes advice to use PGP in it. Maybe the author should have read the references they were citing.)

  • metiulekm@sh.itjust.worksEnglish
    3·
    1 year ago

    Just wanted to say that you actually can sign git commits using SSH keys! :) It’s nowhere close to being as flexible as what you suggest, but it’s also not GPG, so there’s that.

    • danhab99@programming.devOP
      2·
      1 year ago

      This is litterally just my feelings and not based of fact: but that feel icky to me to sign with ssh. IDK why but I’m just like that

      • metiulekm@sh.itjust.worksEnglish
        1·
        1 year ago

        I am no cryptographer, but I understand that in the SSH protocol, the keys are only used for signing anyway: that is, the user is authenticated by saying “I want to authenticate with some key, and here’s some data signed by this key”, and this is completely separate to encryption. It also seems that GitHub encourages using separate keys for commit signing and general SSH access, which might alleviate some of the ickyness.

        You are really not wrong though, I feel like people only started using SSH for this because it kinda worked and they already have been familiar with it.