After some computer issues, I'm back in business!
On the subject of the number of frames for moving: I can't believe I didn't find that, but then I was using a binary search to find when I could next move, assuming the region where a move could happen is continuous and consistent (it's not).
I just tested it out myself now, and found these results:
- Gaps: 9,6,6 - 4 moves done
- Gaps: 9,7,{6,7,8} - 3 moves done; last move eaten
- Gaps: 9,7,9 - 4 moves done
From these results, it looks like if you don't move again on the first frame, you can't continue with the 7-frame movement pattern, and have to restart with the 10-frame. At least it's an easy pattern to TAS with, and the memory watch you found will really help :)
I'm just knocking up a lua script to overlay that memory address to make it easier to see when moves can be inputted; I'll edit it in when I'm done.
----
Edit:
Basic lua script to watch the memory address found. It unfortunately is more a "are pieces moving" address; it doesn't work to see when a move can be input following a shape complete animation. I'm looking for one of those now.
I've found an almost-there address: 0x5410. The value returns to zero the frame
after a move is first possible again following a join, which isn't ideal.
Update! Found a better address: 0x6A88. This byte has the value 2 while a shape is joining; otherwise 0. Updated script now linked.
----
Another edit:
Updated Lua! Automatically does the best input based on the memory addresses mentioned above; press Ctrl+I to open a form to type the solution into in the format used by qqwref's website.
(Script moved to gist for easier versioning, etc. Can be put in a github repo if we feel like it.)
https://gist.github.com/adituv/6bcc89d7d125e604b0b8
----
Should I just double post?:
Using your new awesome single-join routes, and actually measuring the time taken, I've improved Stage 1 by a further 60 frames, now complete at 13669. The only real improvement now that I can see is a single-join 1-16.
http://tasvideos.org/userfiles/info/25419665607784644
At this typical distance between nucleons (i.e. neutrons and protons), the attraction of the strong force outweighs the repulsion of the electrostatic force between the two protons, leaving the nucleons in a lower energy state compared to being a bit further apart.
If you were to move the four nucleons much further apart, they would gain potential energy due to moving against the field, in a similar way to gravitational potential energy gained by something when you lift it up.
Now imagine two systems: one with the four nucleons in the 4He nucleus, and one with the four nucleons "infinitely far" apart (that is, negligible force exerted on each other). There is a difference in energy between the two systems: the nuclear binding energy, or the energy required to split a nucleus into its components.
Because mass and energy are equivalent (E=mc2), the systems have different masses, with the difference in mass (Δm) proportional to the difference in energy (ΔE):
Δm = ΔE / c2
This happens to be a great enough quantity for 4He that it is greater than the difference in mass between two protons and two neutrons.
