I’ve been experimenting to detect cross-BIP interactions and behavioral divergences between Bitcoin Core versions.
In early testing, it flagged a few interesting patterns. One example:
Finding: Different error signaling for invalid BIP‑68 relative lock‑times between v0.16 and v28.0.
- Legacy nodes reject quietly.
- Modern nodes return a specific mempool‑conflict error (-26).
This could theoretically be used to partition mempool visibility between older and newer nodes.
My question:
-
Is this behavioral divergence already known / documented?
-
If so, is it considered a protocol risk or just a policy cleanup?
-
More broadly: does systematically testing cross‑BIP interactions with automated hypothesis generation sound like a useful direction for security research, or is it mostly noise?
I’m not claiming a 0‑day—just trying to validate whether this approach surfaces meaningful signals. But I found 100+ more of these. Happy to share more details or logs if of interest.
