All right, I updated Chrome and tried this out. Just scrolling around in Haunted1, it looks great, though I haven't done any extensive testing to make sure everything's saved properly and such. B and such are working perfectly. One issue I had is that Ctrl+E in Chrome puts a question mark in the omnibox and this blocks the Grab Event behavior. I suspect you are much more versed in Chrome than I am; how do I turn off this behavior?

One thing I'm personally doing differently than in JCS is the default settings for Layer 3: I'm not sure I've ever once used the 1.5/1.5 speeds, so I initialize it as 256/64 and 1/1, just like Layer 4. If you're trying for 100% accuracy, ignore me, but I find it more useful that way.

I'm not sure how I feel about the up/down buttons for animated tiles. All references to the relevant animations within the level remain the same, so that clicking the buttons functionally changes which animated tiles are used at specific places in the level. I had expected the buttons would only rearrange the animations for aesthetic purposes, that is, allowing the user to sort related animations together without changing the level itself. On the other hand, I'm not sure if either approach is more "right" in any meaningful way? What purpose do you plan on the functionality serving, as it is presently implemented?

Are animated tiles properties not implemented yet, or could I just not figure out how to bring up the window?

I was able to mess up the animated tiles somehow, causing big question mark tiles to appear in place of certain frames, but I'm not yet sure of the exact set of steps. I'll get back to you.

The event selection window looks great. I know from experience that there's some tricky stuff there, although I suppose it's a lot simpler without trying to hack in AGA support, and it seems to match JCS behavior quite well. The only issue I notice -- and it's an "issue" only in that JCS does it differently, not in that it's necessarily wrong -- is that when you hover over a generator event, the description at the top is different than the one JCS offers. You have the event name be the event being generated and include the parameters numerically, whereas JCS calls it a "Generator" but gives the name of the event being generated as the first parameter in place of the actual ID. I don't know if this was a conscious decision or not so I thought I should point it out just in case.