Keynote timing unravelled: it’s not good, but it’s a reason

Finally figured out the Keynote nonsense. Apparently, the timer doesn’t start counting until the slide and everything on it–and possibly the next slide as well?–is completely loaded. The transition seems to start (n) seconds after the “Ready to Advance” indicator in the presenter view flashes. I can see where that makes sense, actually, in an overbearing DWIM, CYA sort of way. It guarantees that the content on the slide displays for at least the expected amount of time, no matter what.

The problem, of course, is that load times are impossible to predict. They depend entirely on system configuration and load. And while CYA is nice, I’ll take repeatability over CYA any day of the week. Also, it’s not what I M.

Fortunately, builds seem to actually be timed from the transition as expected. So the screwy transition timing can be circumvented by building out at the appropriate time and transitioning automatically after the build.

Best of both worlds? Maybe. It seems backward to me, though. I’d expect the transition delay to accurately time the interval between transitions. If anything were going to wait to make sure content is ready, I’d expect it to be the build timings, which actually time, well, content display.

Either way, it works, and I guess that’s what’s important.