Nice work.
As to the users, I can only answer for myself.
I’ve been using CLIPS since 2007, both as a teaching tool (in an introductory AI course) and for my own research.
I’ve developed a pattern-based solver of Constraint Satisfaction Problems (with applications to familiar logic puzzles): CSP-Rules.
It has thousands of rules, most of them generic (i.e. valid for any finite CSP) and some application-specific.
It can be used at a totally elementary level of puzzles solving.
But it is also a research tool. In particular, I have used it for large scale studies (unbiased statistics of Sudoku, filtering of 3-digit patterns that require T&E(3), classification of the recently discovered puzzles in T&E(3) …)
What I can say about CLIPS is, it’s a really wonderful tool. I have sometimes run it for a month to solve millions of puzzles, without noticing any memory leakage. I have also run extremely hard (and rare) cases that required up to two millions of facts and 50 GB of RAM, still with no problem.
I take this opportunity to thank and congratulate Gary for his commitment to developing and maintaining it. The few times I’ve found a bug, Gary has corrected it quite fast.
I have also used JESS (the Java version of CLIPS), but it is now much slower and no longer maintained. Originally, CSP-Rules was compatible with JESS, but I have given up with this because it implied too many restrictions on the use of CLIPS.
I have also quickly searched for other projects using CLIPS on GitHub, but I haven’t found any large-scale one.
If you can find this (large-scale CLIPS projects), I would be very interested in having a look at them.
Best
Denis