On Mon, 2026-10-19 at 19:07 +0200, Antoine Pitrou wrote: > Hello, > > Heads up that I gave a talk about fuzzing Arrow and Parquet C++ last > week at Community over Code in Glasgow (thanks to Steve Loughran for > suggesting that I do this). > > The slides can be read online at > https://pitrou.net/coc-glasgow-2026/Fuzzing.html > > Obligatory social media posts that people might want to upvote: > - Mastodon: https://pouet.chapril.org/@__ap__/117468660137262014 > - LinkedIn: > https://www.linkedin.com/feed/update/urn:li:activity:7517859848663846912/ > > Regards > > Antoine.
Nice talk. Several of the trailing questions are open ones. A lot of Google-provided fuzzing infrastructure has been steadily declining in quality/upkeep for years now (OSS-Fuzz, Fuzzbench, etc.), and it has been hurting the community quite a bit. There's a bit of a classic aphorism in the fuzzing community: the first 95% of bugs that your fuzzer can find will be found within the first 30 minutes of fuzzing. The rest need better harnesses. Y'all's work on harness writing has been really fun to watch from the sidelines, and I applaud this work, especially for integrating some of the less common fuzzing strategies. -- Addison Crump | PhD Candidate CISPA – Helmholtz-Zentrum für Informationssicherheit gGmbH Stuhlsatzenhaus 5, 66123 Saarbrücken, Deutschland Room 2.17 e [email protected] | w www.cispa.de
smime.p7s
Description: S/MIME cryptographic signature
