A couple of additional thoughts:
I'm still somewhat surprised the SE/30 Finder Shutdown doesn't toggle the line, but if the PDS slot maintains 5V, I guess that's not totally shocking. I'm not entirely sure that pin is useful to an add-on card anyway, since hard power offs were a fact of life... So maybe the docs are just wrong in this regard. Or reflected original intent that didn't pan out.
As I think about it, the soft power gestalt probably only matters for preventing systems with soft poweroff from powering off rather than the inverse. The problem would be, not all machines soft power off the same way (not all machines even have the 2nd VIA that the SE/30, IIx have), so just claiming it has soft power doesn't really tell it how to soft poweroff. There's something else somewhere that knows how to soft power off a particular machine. This probably isn't determined by shutdown-time evaluation of the machine gestalts, if it was, the Wish I Were app would change your poweroff behavior, which could potentially be detrimental considering how hardware specific the shutdown is.
So anyway, there's apparently a bunch more stuff involved with the shutdown process. I'll try to read more to understand how it determines the shutdown procedure for a specific machine. Ideally, convincing the shutdown procedure to use the IIx's shutdown instead of the SE/30's seems like a good idea. Short of that, being able to hook in to the process at a safe point would work(hopefully making sure any existing block cache is flushed, etc. before actually toggling power).
I'm still somewhat surprised the SE/30 Finder Shutdown doesn't toggle the line, but if the PDS slot maintains 5V, I guess that's not totally shocking. I'm not entirely sure that pin is useful to an add-on card anyway, since hard power offs were a fact of life... So maybe the docs are just wrong in this regard. Or reflected original intent that didn't pan out.
As I think about it, the soft power gestalt probably only matters for preventing systems with soft poweroff from powering off rather than the inverse. The problem would be, not all machines soft power off the same way (not all machines even have the 2nd VIA that the SE/30, IIx have), so just claiming it has soft power doesn't really tell it how to soft poweroff. There's something else somewhere that knows how to soft power off a particular machine. This probably isn't determined by shutdown-time evaluation of the machine gestalts, if it was, the Wish I Were app would change your poweroff behavior, which could potentially be detrimental considering how hardware specific the shutdown is.
So anyway, there's apparently a bunch more stuff involved with the shutdown process. I'll try to read more to understand how it determines the shutdown procedure for a specific machine. Ideally, convincing the shutdown procedure to use the IIx's shutdown instead of the SE/30's seems like a good idea. Short of that, being able to hook in to the process at a safe point would work(hopefully making sure any existing block cache is flushed, etc. before actually toggling power).