[Uuidrev] Fwd: Concerns regarding backwards compatibility in draft-davis-uuidrev-uuid-long-00

[Brian Mottershead <[e-mail]>]

Thanks, I do think that identifiers longer than 128 bits are needed,
especially for many cryptography-related and hash-related identifiers.

As for the 3968-bit max length, I am not sure where that came from. Such a
specific number seems odd, considering that (a) it seems very long; and (b)
it isn't actually long enough if the motivation is to be able to store very
long cryptography keys.

The latest RSA public keys, for example, which one could imagine would be
useful as identifiers, are 4096 bits.   The other thing that seems odd is
that in the Long UUID prefix, the length field is 16-bits, and specifies
the length in octets.  This allows the length of the Long UUID to be 65536
octets.  Why accommodate such an enormous length and then impose a limit
that only uses a fraction of it?

But, obviously, such long Long UUIDs are never going to be translatable to
UUIDs in the manner I suggested, because they can't satisfy the requirement
that the 128-bit "short UUID" part suffice as an identifier.  The trailing
bits in that case are not just nice-to-have, but critical to the identifier
role.   Those wouldn't be in the sv0 UUID-compatible space that I suggested.

Brian


On Fri, May 15, 2026 at 3:29 PM Michael Richardson <[e-mail]>
wrote:

>
> Brian Mottershead <[e-mail]> wrote:
>     > I would like to express a concern regarding
>     > [draft-davis-uuidrev-uuid-long-00] the recent Internet-Draft
> specifying
>     > "Long UUIDs". In essence, this draft proposes a new universally
> unique
>     > identifier format to be standardized by the IETF. "Long UUIDs" are:
>
> It's not chartered, it's an idea.... so this is exactly the time to talk
> about this.  Thank you for the email.
>
>     > Although the identifier is described as a "Long UUID", it is actually
>     > an entirely new identifier format that is neither backwards- nor
>     > forwards-compatible with standard UUIDs:
>
> That's a fair comment.  I'm not sure what motivates 3968 bit sizes.
> Do you feel that you have a need for things longer than 128, and if so,
> why?
>
> I don't think that anyone expects Long UUIDs to be a drop-in.
> I would suggest a name not including "UUID" in the name would help.
> LUID, ULID, ???
>
>
> --
> Michael Richardson <[e-mail]>   . o O ( IPv6 IøT consulting )
>            Sandelman Software Works Inc, Ottawa and Worldwide
>
> **       My working hours and your working hours may be different.
>  **
> ** Please do not feel obligated to reply outside your normal working hours
> **
>
>
>
>
>