The issue is "bigger" than I thought in a sense. I...
# ballast
r
The issue is "bigger" than I thought in a sense. I found another bug caused by LIFO where the system was receiving an input, updating some state (in this case to close a popup), and then before it could go do the other things it needed to do, the processing of the input was cancelled. Which basically means that when doing LIFO, the programmer must be very careful to defer state updates to the end of the input processing -- this isn't something most people will intuitively do, I think. Could the default behavior of LIFO be to do LIFO as now, but not cancel any already ongoing operations?