Oh well. we were told many times not to count on layering working forever. I love the new PSTN change and I have at least a dozen save to fix of my own. It is what it is.
It's fine if a change breaks some saves, but only if the change has more good effects than bad.
The PSTN changes doesn't seem to do anything good, as far as I know. It's just another limitation, like, for example, the 5 particle layering limit (unless you use EHOLE, which can cause problems).
OH!
where was this announced? Is there a tutorial/explanation anywhere?
Anyway, the main problem I have is that PSTN doesn't work when other stuff is layered on top of it, like
"Pure" layered piston still works.
The original stacking method still works:
The only thing that broke was PSTN layered under other elements.
This is exactly why I hate layering.
https://github.com/simtr/The-Powder-Toy/commit/d0cf82de03b6d33fea4e5e5c0418625fc78d46e8
As well as making functionality additions difficult, it also means I can't do nice things like move PSTN triggering to the SPRK update function (which would allow removal of the PSTN update function and hence provide a small speed increase).
This is one of the reasons why I don't do as much work on the official version of TPT as I used to... All the layering and subframe stuff means some saves rely very heavily on the exact details of how TPT works, so it's much more difficult nowadays to make any useful changes without breaking something.
I tell people "don't do X, it may break", then someone goes and does X anyway, and complains when X breaks. :(