Chips should remember materials used when created

[Mordenkainen]

Chips should remember what materials were used when they were created.

Currently if you create a very compact circuit by using rcactivate to specify different IO materials to achieve density, then have to do an rcreset (or worse rcreset /all) things can get very ugly.

If the chips remembered what materials where specified when they were activated, then a re-activation of the chip should work with no special handling in most cases.

[Mordenkainen]

After looking at the other issues, this seems related to issue #37

[eisental]

Yeah, that's exactly what I had planned for solving issue #37. Can you give an example, though, for when it helps using different io materials? I thought changing the chip block type would be enough in most cases.

[Mordenkainen]

Sure.

In this first picture, you can see a delay on the left built out of nether brick. It has to be moved out one block because it's layout would run it right up against the output blocks of some receivers. Even if the IO blocks were changed it would attempt to use the receiver output blocks if re-activated.

http://dl.dropbox.com/u/53001205/counter%201.png

In the second picture, we have the same circuit, with the "ideal" placement of the delay. It is using different chip material and IO blocks. This works fine, but on a reactivate it forgets that the IO blocks were changed and malfunctions.

http://dl.dropbox.com/u/53001205/counter%202.png