2.00 Bomber mechanics in NeoLemmix V2.00

[Nepster]

namida has recently opened up the discussion for small physics changes for NeoLemmix V2.00, so I will start by discussing one pecularity of bomber mechanics.

It concerns bombing in miner shafts: The bomber hole slightly depends on the direction the lemming is going! See attached picture. Upper row shows the two choices of bomber holes for lemmings going down, the lower row does the same for lemmings walking upwards. The bomber holes for ascending lemmings is actually one pixel further to the left that for descending lemmings.
This small difference has one rather big consequence: Only ascending lemmings can create bomber holes that turns other lemmings around - descending lemmings cannot!

I suggest changing the bomber placement for descending lemmings into the one for ascending lemmings. This way it is unlikely to break any existing levels (as turning lemmings in such bomber holes was already possible in original Lemmix and NeoLemmix up to now), but makes the behaviour uniform.

[mobius]

I can confirm this and I agree; I like the mechanic of being able to create a wall with a bomber in a mining tunnel.

[namida]

This is a fairly minor side effect of fixing the much more significant bug where bomber symmetry was broken (eg. take the Martian set for example, the red bars, blowing up a climber on one side of them would break all the way through, while on the other side it would not; in this case, the effect on miner tunnels would be similar to what you describe, except based on the direction of the tunnel, not of the lemming).

While I'd love to get this fixed for sure, I'm not entirely sure how to do so without going back to a direction-dependant (rather than symmetrical) bomber, which is a much bigger issue. (Or alternatively, writing code that detects miner tunnels or similar slopes, and applies special, inconsistent physics in such a case.) I remember that around the time I was fixing the bug, I had quite a strong discussion with Simon on #lix, but it's probably buried under millions of years of chatlogs now...

[Simon]

Maybe it's a corollary of the walker physics. But I'd have to examine the NL code before assuming anything here. L1 walkers entered solid terrain when turning, but not when ascending 1 pixel, IIRC.

I'm looking forward to many more strong discussions in IRC!

-- Simon

[Nepster]

Have you considered to make the bomber hole one pixel wider, such that it has odd width and can be properly centered around the lemming?

[namida]

I think such an option was discussed when this earlier bug occurred. The end decision was that it should be taken into account that although the lemming's position, in memory, may be only one pixel, a lemming's body is two pixels wide (it is perhaps worth noting that the digger does not really apply this logic, but at the same time, no complaints about symmetry ever came up with the digger anyway). Perhaps then the solution is to better take this into account in other cases too.

I'm not sure whether the option of widening the bomber mask as such was never considered, or was rejected - I'd have to find the appropriate chatlogs to say for sure (unless Simon can remember anything about it). Certianly, it's an option that could be revisited, especially if we're making other physics changes at the same time (which seems likely).