When I added the code to sort the results table on any reasonable column, I noticed that the priority column sorted the tasks backward. Hmm. So I fixed that.

Surprisingly, the analysis code did not totally explode, though some results were a bit strange. (Most analysis code assumes that the tasks are sorted so the higher priority tasks are at lower indexes.) I fixed that. Now sorting the output table and sorting for analysis use different sort methods.

Also, 16 turned out to be way too high a bound for exponential search. I'm still experimenting with different cut-off values, but it will probably settle on something like 8.

Since there is a difference between having a heuristic fail to find a good priority assignment and having an exhaustive search fail to find one, I have them generate different messages. One says there is no feasible priority assignment. The other says say a feasible priority assignment cannot be found.

I've been thinking about a good way to describe a system in a way that will let the applet compute total blocking times. Up to yesterday I was working on something that replaced the cost term in the current input format with something using lots of parenthesis to express cost, cost while holding a lock, cost while holding another lock, nested locks.... Everything with enough power was ugly. I'm trying a new approach that looks more like pseudo code.