Threshold Call: upcoming second public draft (2pd)

[Peralta, Rene C. (Fed)]

Dear MPTC-forum,

We have been refining some elements of the submission process described in NIST IR 8214C ipd (initial public draft of the Threshold Call).

A second public draft will be published to allow for additional public feedback.

Before that, we would like to ask for your opinions on various topics, including:

- Implementations with external dependencies

- Multi-part specifications

- KAT values

- Phases and deadlines

- ...

In the next days and week, we'll initiate a few separate email threads for these topics, motivating concrete comments about considered options. We hope to hear back from people interested in the process. Please reply in the corresponding email threads.

Thank you for your attention

René Peralta

(on behalf of the MPTC Project).

[Nigel Smart]

Hi All

In the initial public draft for document M4 it talks about measuring...

Perf2. Processing time (in seconds) and/or processing (e.g., number of
processing cycles).
Perf4. Networking time (in seconds).

We are not sure if this makes sense. For example we are unclear how
to do this, and in addition it is common for a party to have the following
execution...

Send stuff
Do stuff
Receive stuff

Thus a party may not be slowed down by the need to send data, as
it can use the "wait" time in a round to do useful computation. Hence
total time will not be
           Processing Time+Networking Time

Thus we think "total execution time" per procedure might make make
sense, but measured in two settings LAN and WAN [e.g. on a single
AWS data centre, and on two widely separated AWS sites].


Memory complexity is also a little undefined, as it depends on what one
means. Do we mean RAM or Disk Space? After all we could trade one
for the other in a given implementation.

Yours

Nigel