Hi! Frankly, I do not want to heat up again the debate about the IETF OpenPGP WG and their behaviour. Some points anyway:
> Do you think Collin's worry could become real, that is, GnuPG keys could > not be uploaded to or retrieved from keys.openpgp.org at some point > in the future? 1. keys.openpgp.org is like the formerly PGP.com enterprise keyserver and contradics the PGP model of decentralization. That keyserver requires an email confirmation and thus can't synrconize with other keyservers. 2. The keyserver also does not send the user ID packet and requests clients to figure out which key signature belongs to which user id. The pleaded reason are the GDPR requirements. That is entirely wrong because with that reasoning they may also not distribute the keys at all or any key signatures. (A key or a signature is not different from a user ID, becuase it allows to identify a person). 3. keyservers are an ancient way of distibuting keys and actually dangerous because they wrongly make people believe that a key actually belongs to a certain person. Due to DoS problems we should not use keyservers anymore. The solution here is the Web Key Directory (which delegates the responsibility to the same entity which manages the mail address) and sending keys along with the mail. In fact, gpg has for long time a way to embed the key in a signature so that after sending a first signed mail, the peer has access to the key and can reply encrypted. This reflects real world communication models better than keyservers. > Or will OpenPGP and LibrePGP remain so close to each other that I can't say whether the IETF WG will revert their changes but I don't think so. 4. The new OpenPGP specification RFC-9580 is not the planned update to OpenPGP (RFC-4880) to adjust algorithm to modern requirements but a large rework without the consent of the major implementations (GnuPG and RNP). 5. RFC-9580 introduces extra complexity to allow the use of GCM as an optional new cipher mode. Despite that GCM is a fragile and easy to get wrong mode, it has no more use in todays zoo of algorithms except for backward compatibiliy. GCM was introcued to avoid patent problems with other modes but those patents have meanwhile expired. The claimed reason for GCM is that web broswers don't yet implement the OCB mode and thus Protonmail has minor performance issues for theyr browser based implementation. 6. The term LibrePGP was coined to make it easier to refer to a standard than to explain what rfc4800bis-2015 means. This is unfortunate but the IETF practically took over the well estabilished term OpenPGP which used to describe what the majority of PGP compatible implementations (GnuPG, RNP, BouncyCastle) actually implement. For details see our https://librepgp.org site. Shalom-Salam, Werner -- The pioneers of a warless world are the youth that refuse military service. - A. Einstein
openpgp-digital-signature.asc
Description: PGP signature
