Steensgaard vs Andersen's analysis

[ahorn]

Thanks for sharing your technical report.

I am curious:

Your experimental results (from 2016) seem to suggest that a Steensgaard-style pointer analysis achieves comparable precision compared to Andersen's analysis, at least for the data sets you considered. Two years on now, I am wondering whether you happen to have collected or encountered any more data points that may have changed your outlook on the trade-offs between precision and performance.

Thanks for your time!