On Wed, 8 Jul 1998, Adam Back wrote:
This is necessary because for new algorithms, or new features etc.,
there may well be multiple proposals for the same algorithm. So
allocating #12 to the succesful AES candidate is in itself useless,
because there may be 2 or 3 incompatible implementations.
AES will require a single specified base mode for any two AES
implementations to interoperate - much like DES usually refers to the 56
bit ECB mode. It is implicit that it run using PGP-CFB for OpenPGP.
There should also be a single key setup specified for the same reason.
Things like HAVAL needed further specification since there were several
flavors of "HAVAL" (same for BlowFish). The ones being reserved are (or
should be in the case of AES) unique, or specify a specific flavor (e.g.
160 bit 5 round HAVAL).
Better would be to allocate #12 to Tom Zerucha, #13 to cryptix, and so
on on demand, and then the variant which is chosen as the standard for
inclusion in OpenPGP 2.0 can use that ID.
What if I am working on multiple algorithms - I would need a range.
Otherwise these things should be implemented in the experimental range
(100-110) and implmentors can discuss between themselves which one should
be codified into a small number.