So I was writing a new test and checked another test and sth weird happened: a test run via lein test ... :only ... failed, but did not fail when run with others, which led me to discover a relatively benign fault in Clojure/CLojureScript Matrix: the default value for a Cell pulse was zero, the same as the starting pulse of the Matrix. Egad! That means it is current, when in fact it is not. Fortunately, in normal Matrix app flow we start with models, not standalone cells, and new MX objects (models) force their Cells awake, so this is not a problem. But still. And for one, standalone Cells are allowed, so I guess they should work. And tests should work. So now...
• cells start with nil as the pulse, pulse-observed, and pulse-last-changed;
• tests run OK as a suite, but I should try them individually; and
• I got carried away and pushed to main and CLOJARS.
It was only a couple of lines of code, but...from now on I will adopt proper semantic versioning.
OK, all tests have been updated if necessary to self-initialize individually (what makes me think I could have done that in one place if I had RTFMed the deftest doc? 🤣 ) and tests still run OK. I have made one final push on the legacy tiltontec/matrix/4.1.8-SNAPSHOT/matrix-4.1.8-20230109.164728-8.jar branch. Next up, when things settle down, a stable 4.1.9.