Improving Software Quality in Cryptography Standardization Projects
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
The NIST post-quantum cryptography (PQC) standardization project is probably the largest and most ambitious cryptography standardization effort to date, and as such it makes an excellent case study of cryptography standardization projects. It is expected that with the end of round 3 in early 2022, NIST will announce the first set of primitives to advance to standardization, so it seems like a good time to look back and see what lessons can be learned from this effort. In this paper, we take a look at one specific aspect of the NIST PQC project: software implementations. We observe that many implementations included as a mandatory part of the submission packages were of poor quality and ignored decades-old standard techniques from software engineering to guarantee a certain baseline quality level. As a consequence, it was not possible to readily use those implementations in experiments for post-quantum protocol migration and software optimization efforts without first spending a significant amount of time to clean up the submitted reference implementations. We do not mean to criticize cryptographers who submitted proposals, including software implementations, to NIST PQC: after all, it cannot reasonably be expected from every cryptographer to also have expertise in software engineering. Instead, we suggest how standardization bodies like NIST can improve the software-submission process in future efforts to avoid such issues with submitted software. More specifically, we present PQClean, an extensive (continuous-integration) testing framework for PQC software, which now also contains “clean” implementations of the NIST round 3 candidate schemes. We argue that the availability of such a framework-either in an online continuous-integration setup, or just as an offline testing system-long before the submission deadline would have resulted in much better implementations included in NIST PQC submissions and overall would have saved the community and probably also NIST a lot of time and effort.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it