What if Deflektor had some elements from Mindbender

[Holger]

I have experimented a lot with the new proposed "slope" style element for the last week, which is in fact more complicated than I had expected, with quite a number of "special cases" that still have to be detected and treated to have them work correctly and "as expected".

Here's what I already have:

df_slope-3.png (39.59 KiB) Viewed 1136 times

And here's what's still not working correctly (plus some more "special cases", which I haven't shown here):

df_slope-problem-1.png (34.02 KiB) Viewed 1136 times

df_slope-problem-2.png (34.32 KiB) Viewed 1136 times

df_slope-problem-3.png (33.51 KiB) Viewed 1136 times

As mentioned, it's fun to play around with this, and I hope I will be able to manage and handle those cases which still don't behave as they should.

[Eizzoux]

Yeah, I was expecting it would be weird for you to work around with, but pretty cool to see me giving you encouragement to actually replicate new elements with their unique behaviors. And I guess if it was actually fun to make, then good to hear as well! I imagine you'd expand MM just as much as R'n'D in few updates as well, and could be quite a lot of ideas doable in MM engine, it might all be yet to see in the future, of course.

To be real, with new unique elements, this could pretty much be it's own big update overall, coming out more as 4.3.6 instead, but that's just me speculating anyways. And since you're into it already, then good luck with the progress, I'm looking forward into it!

[Eizzoux]

Speaking of, there's also quite a lot of weird cases that could possibly be accounted for the beam reacting to the block corners...

heck1.png (31.43 KiB) Viewed 1007 times

heck2.png (33.7 KiB) Viewed 1007 times

This one especially can complicate things quite a lot.

heckomg.png (98.52 KiB) Viewed 1007 times

[Eizzoux]

For a while I felt uncanny from the way DF mirrors were done for MM. For pretty much all of the angles other than the vertical and horizontal ones the sprites were rotated in sort of lazily rotated fashion, making them look kind of... jagged. Especially their outlines are really strange.

Original DF mirrors up close

gsdhfsdhgdfh.png (25.11 KiB) Viewed 978 times

And since at first I just recolored mirrors for the Reflektor color palette, I just didn't like what I've done and wanted to changed them up a tiny bit for RF, and I don't mean just the shadow/outline.

Old | New

kfbhdksfbdsbfdsk.png (244.27 KiB) Viewed 978 times

[Holger]

To be real, with new unique elements, this could pretty much be it's own big update overall, coming out more as 4.3.6 instead, but that's just me speculating anyways. And since you're into it already, then good luck with the progress, I'm looking forward into it!

Yes, it will most probably be released as a major release version 4.3.6.0. But I'm afraid it will still take some more time, as there are far more edge cases than I first expected (especially after testing around with some of your examples where the beam is reflected by a single "steel wall" block edge, which really adds even more weird cases ).

Also crazily complicated are all those cases where the beam hits exactly between a slope element and a steel wall element, for example (like in your first example above). Maybe I will not cover them all, and just let the engine do some stuff that looks inconsistent or feels unexpected here and there... :-/

BTW, your redesigned DF style mirrors look cool indeed!

[Eizzoux]

But I'm afraid it will still take some more time, as there are far more edge cases than I first expected (especially after testing around with some of your examples where the beam is reflected by a single "steel wall" block edge, which really adds even more weird cases ).

Also crazily complicated are all those cases where the beam hits exactly between a slope element and a steel wall element, for example (like in your first example above). Maybe I will not cover them all, and just let the engine do some stuff that looks inconsistent or feels unexpected here and there... :-/

I think in cases where the beam is pointed against such corners it should just start overloading. If you're going to program full functionality of the corner between slope and vertical/horizontal surface, that might lead to such broken cases where the beam might just redirect horizontally/vertically right between the tiles where it would be almost a dead end with no way of returning aside from using another slope on other end or rewriting the game to allow you to place bigger tiles out of their natural grid, which is not only going to be abysmally long process of performing that, but also ridiculously cursed and broken for any of the engines.

asdatgadfsdfdsf.png (53.65 KiB) Viewed 806 times

[Holger]

I think in cases where the beam is pointed against such corners it should just start overloading.

Yes, I agree, especially after looking at your example!

I think such "bad angles" do not happen with my current test version yet, but I will check again. Apparently supporting such "non-standard" reflection angles will do more bad than good...

[Eizzoux]

Just out of curiosity, what will be tokens for the fixed DF mirrors and slopes? I'm asking just because I'm in progress of expanding canvas and reorganizing game sprites to fit in the new elements. Pretty much getting it ready for the update whenever it comes out.

dsfkdnsfndsnfkls.png (343.2 KiB) Viewed 727 times

Also, I might have given this idea before already, but will there be possibly active states for more elements? For example, DF refractor and MM prism or MM pushable blocks? I added animations for active states for some of them which can only be seen in element info screen for now.

dkasjdhnakdj.png (161.57 KiB) Viewed 727 times

[Holger]

Just out of curiosity, what will be tokens for the fixed DF mirrors and slopes?

It will be "df_mirror_fixed_1" to "df_mirror_fixed_16" and "df_slope_1" to "df_slope_4".

Also, I might have given this idea before already, but will there be possibly active states for more elements? For example, DF refractor and MM prism or MM pushable blocks? I added animations for active states for some of them which can only be seen in element info screen for now.

This would indeed make a lot of sense, yes! I will try to add some more active element states in the future!

[Holger]

OK, the new slope game element for the Mirror Magic game engine is finally released to the public!

In fact, I think it is the most complex game element of the MM engine now; at least it required quite some effort to implement it and especially to work around lots of potential problems and "special cases". And there are still a number of known (and probably unknown, or yet undiscovered) edge cases where the laser beam could be reflected in a way where no mirror can be hit anymore (because the beam does not hit the game elements in the exact middle or at the "right places" anymore). But in general, and in most cases, it seems to work quite nicely, and the crazy edge cases seem to be limited to those cases shown above where the laser beam gets reflected from the corner of a steel wall (as far as I can tell for now), which creates different and more "complicated" beam positions just as shown and described above.

Please have a closer look at the new game element, and let me know of any major glitches I might have overseen so far.

[Eizzoux]

Yeah, for the most part, they do their job just as they are intended to, and yeah, as expected, some special cases are just way too special to work around. But still, good job on actually implementing that! Although I just asked if that's a possibility, but... I guess that answers my question :Р

ohwow1.png (83.05 KiB) Viewed 416 times

Although there's one thing which actually "kinda" breaks the game actually by misaligning the beam just so it would be able to reflect off single block's side, which is not possible with any other elements except slopes.

ohwow2.png (31.93 KiB) Viewed 416 times

And honestly, sometimes I just like broken stuff because it's kind of fascinating to look at that.

ohwow3.png (13.61 KiB) Viewed 416 times

Also, sorry for being kind of rude, but the fixed DF mirrors look kind of... not really nice. I feel like they would look better if they instead had sort of an outline around the mirror (sorta like the glue on my fixed mirrors) instead of the mirror being recolored by itself.

ohwow4.png (2.25 KiB) Viewed 407 times

[Eizzoux]

Click on the image...

[Holger]

Although I just asked if that's a possibility, but... I guess that answers my question :Р

Well, many "special cases" (using single steel wall reflections) from your previous post should work more or less fine (resulting in "correctly" reflected beam which can later correctly hit a mirror again), but what indeed makes trouble is the beam hitting the very edge of the slope at the top towards the slope or at the bottom, barely touching the slope (lie shown in your screenshots), plus those situations where the beam gets caught in a "45° slope corner" (as also shown by your screenshots above). These either cause an "off by a few pixels" situation (like can be seen with being reflected by the side of a single steel wall block) or totally breaks the beam (both as you just demonstrated).

I have already played around with some potential workarounds, but they don't work that well. Detecting these cases and just causing the beam to stop and overload may be the better "solution" here, but sometimes also look weird/wrong.

But I think we can live with that.

Just let me know if you should manage to cause weird behaviour when the beam is *not* reflected by a single steel wall block. That at least should be more or less solid (I hope).

And honestly, sometimes I just like broken stuff because it's kind of fascinating to look at that.

Yes...

Also, sorry for being kind of rude, but the fixed DF mirrors look kind of... not really nice.

No problem with that.

I am also not too happy with the current graphics choice... IMHO it should better be like the MM graphics, having colorful graphics for the non-fixed mirrors and gray graphics for the fixed ones. But the all-gray non-fixed DF mirrors are already there, so I decided to use the colored ones for the fixed variants, which probably not the best solution.

[RF] Surprise Orb showcase

Looks cool -- a little bit zonk-like!