OQS presence at relevant conferences - Message

[RodriM11]

As detailed in the last TSC meeting, we have decided to re-establish our presence at conferences, to (a) communicate that OQS is active and that its core mission — research, new algorithms, new primitives, hybrid constructions — remains very much relevant despite the standardization of the first PQC algorithms, (b) gather feedback from users about what they want from the project, and (c) attract new contributors.

This issue is for members to propose / discuss certain messages that we would like to convey at these conferences. Since deadline for certain conferences is fast approaching, I will provide a deadline of two weeks (until May 14th) for this discussion. After it, and based upon the feedback gathered, I will create proposals for conference talks and submit them to open CfPs, from the list in #276.

[RodriM11]

As I stated on the TSC meeting, I believe renewing our presence at conferences is important to convey the message that the mission of OQS remains the same, despite the first PQ standards already present and available. That is, to continue to be the forefront project in terms of PQ research, via continue to provide new PQ algorithms, features, hybrids and so on.

It would also be interesting IMO to discuss the "historic" phases of the projects, in a similar manner as what the HR of March 26 identified, to state that the project is currently under a clear post-standards phase.

Furthermore, I think we could also talk about new "additions" that either have landed, are working on, or are planning to address: 1) inclusion of relevant schemes to be standardized (i.e. HQC, eFrodoKEM), round 2 signatures for NIST on-ramp, standards from other competitions (e.g. NTRU), 2) New features for schemes (e.g. external mu, increased platform support, sign/verify message API, OpenSSL 4.0 support), 3) Features of the various wish lists of the different subprojects.

I believe these talks would help attract potential new collaborators, and also help us understand what features the community is enjoying from OQS, and which ones they would like to were them present.

[baentsch]

Agreed wrt to conference presence. What's imo required for that is a deliberate decision as to which conferences would reach the contributors we'd primarily like to attract -- and the message(s) that would be enticing them to the project.

In line with thinking that the OQS goal is to help new PQ algs prove their value, one suggestion could be to consider events at which authors of new PQ algs would be to entice them to help add their algs to OQS. Arguments to interest them could be simplified integration and testing of their algs in larger software stacks as well as community effort to further strengthen their code, e.g., by a good CT testing harness. In that realm, may it be an idea to document (as paper or otherwise) what OQS does in the space and how it compares to the state of the art (@bhess)?

In turn, I'd personally give reduced attention to conferences where already standardized algs are likely topics: Contributors to such algorithms and code bases probably are more motivated to help with more production-oriented FOSS libs like OpenSSL or BouncyCastle.

I guess what I'm trying to say is: Can we agree on what are "relevant" conferences in the sense of this issue?

[ashman-p]

It is good to have a set of target conferences where the OQS project is known to attend and participate (make major announcements, etc.). Secondarily, we may want to have some stated guiding principles that help determine others to participate in should the opportunity arise. This gives maximum flexibility.

I also think that OQS should continue to cultivate and leverage synergies with other open-source projects. There is not a competition with these other projects; in fact, it’s more collaborative. The community is more encouraged seeing that. It is also helpful when there is a smooth transition between code bases from OQS to, say, OpenSSL or Wolf.

[RodriM11]

What's imo required for that is a deliberate decision as to which conferences would reach the contributors we'd primarily like to attract

Yes, I agree. I created #276 as a separate issue to try and "divide" the debate between the message(s) to convey, and the venues in which to do so. But I am aware that the two are very related.

In line with thinking that the OQS goal is to help new PQ algs prove their value, one suggestion could be to consider events at which authors of new PQ algs would be to entice them to help add their algs to OQS.

I think that would be one target to follow. But, on one hand, it may not always be easy to know which conferences those may be. On another hand, and with the specific topic of enticing the contribution of new PQ schemes, I personally think that we should not leave it to just meeting at conferences, and should take a more active approach. I was actually planning on raising this topic at the next meeting and will undergo it myself if nobody else wants to (i.e. carry out liboqs#2098).

Arguments to interest them could be simplified integration and testing of their algs in larger software stacks as well as community effort to further strengthen their code, e.g., by a good CT testing harness.

I agree, and furthermore will add that the presence of such algorithms under a common, easy to use platform will certainly increase experimental use and study, which can help with selection / round advancement opportunities.

In turn, I'd personally give reduced attention to conferences where already standardized algs are likely topics: Contributors to such algorithms and code bases probably are more motivated to help with more production-oriented FOSS libs like OpenSSL or BouncyCastle.

IMO conferences in which people "know the drill" of OSS collaboration are more likely to attract collaborators / contributors (when interest align), so I think we should be present at those too (at least for now, obviously with time we may revisit certain presences depending upon updated / new interests, etc...).

[RodriM11]

It is good to have a set of target conferences where the OQS project is known to attend and participate (make major announcements, etc.). Secondarily, we may want to have some stated guiding principles that help determine others to participate in should the opportunity arise. This gives maximum flexibility.

So, share how to collaborate with the project / interesting issues to tackle, etc?

[NA] Yes.
The second point aims to provide some flexibility for one-off situations. If a conference is not initially included in the list or a new one emerges, having a statement can simplify the decision-making process.

[bhess]

The list of conferences currently in #276 seem the most relevant to me. I would also keep an eye on an upcoming NIST PQC Standardization Conference. There isn’t one announced yet, but there has typically been a conference following each round of the standardization process, and Round 2 of the signature on-ramp is likely to close soon. All NIST submissions are usually represented at these, which could help attract new contributions.

In that realm, may it be an idea to document (as paper or otherwise) what OQS does in the space and how it compares to the state of the art (@bhess)?

That's a good idea.

[RodriM11]

As discussed in yesterday's TSC meeting, short abstracts have been submitted for three conferences with open submission programs. The short abstract goes as follows:

"The Open Quantum Safe (OQS) project was founded with the mission of supporting the transition to Post-Quantum Cryptography. On its way, it became one of the most relevant open-source post-quantum cryptographic organizations. The subsequent publication and adoption of the first PQ standards by major cryptographic projects helps the cryptographic community in the adoption of this required type of cryptography. The mission of the OQS organization remains being at the forefront of PQC research. This talk provides an update of what these efforts look like in the era of the first PQ standards. It will include lessons learned along the way, challenges encountered when handling a PQ Open Source project and future directions."

Note that this is intended to be just a brief overview of what we would want to convey, derived from the discussions in both this issue as well as a number of meetings. If the proposals are accepted, we would still have to expand and go into much more detail, so a lot of room left for members to contribute and help shape the message.

[baentsch]

Thanks for sharing @RodriM11 . Here's a thought regarding this message and pertaining to efficiency, namely the question whether OQS may want to strictly adhere to the mission of "being at the forefront of PQC research", of making available PQ algorithms that are novel and still in the running for some standardization.

OQS already does prune algorithms that are no longer "en route" to standardization. The only exception is if someone external to the core OQS team takes up the baton and keeps supporting them within OQS.

Now, what about extending this logic also to algorithms that have been standardized?

They are no longer novel, are available via really well supported crypto packages and only cost time maintaining in OQS for no incremental benefit to the FOSS community as that already is able to select now many other packages containing those algorithms.

From an external messaging perspective, this would also make things really simple: OQS is about making novel things available via various integration/evaluation routes. OQS does not compete with other FOSS focusing on standardized stuff. OQS focuses its scant resources to deliver things no-one else does. OQS does not have to handle many arbitrary (AI?) "security reports" about code that anyway is not meant to be used productively: OQS could simplify its processes and focus on (making) new things (readily consumable).