Splat ruler: Newbies think full height is for walker, incl/excl ambiguity

[mobius]

I understand the arguments but IMO these are all corner cases that don't warrant over-complicating what should be a very simple tool.

And it's not like I'm totally against complicated design. For example, I find the "instert-replay mode"  to be very complicated and also probably problematic for new users, yet I've stated else where that it really is a god-send for use on certain situations, and not uncommonly so.

[WillLem]

I've now uploaded Version 4 of the NeoLemmix 12.13 Experimental, which includes a new fall distance ruler for testing purposes.

It now only shows the actual fall distance - no potentially confusing markings or additional measurements. Plus, we'll give the colour cycle a spin and see if we like it (I understand Simon's concerns regarding this, so I'll look at either making the colour cycle optional or only cycling through non-problematic colours).

I'll also be adding the Climber ruler as a separate graphic, callable via its own hotkey and/or by Shift-modding the regular hotkey (currently I), but that's a job for another day.

Please do check it out and offer feedback as the features in this experimental are likely to make it into NeoLemmix, so it would be better if they've been fully tested and discussed prior to final implementation.

[Simon]

Here are two candiates.

Design criteria:

• The main bar is 3 pixels wide, like the NL 12.14 ruler. I think that's fine, but I have no strong feelings about the width.
  
• No 1-pixel-high horizontal marking lines. Horizontal features, e.g., the triangles at the ends, are several pixels high. Reason/goal: Prevent the inclusiveness/exclusiveness ambiguity.
  
• You must stand the ruler onto the ground. Don't wedge it into the ground. Reason: This is the only natural way when you have no visible horizontal cut-off other than the very ends of the image itself.
  
• The ruler measures only one distance, the regular walker's length. It doesn't measure the other two distances.
  
• The ruler has two-colors, black and white. Reason: See that you're measuring correctly even when you measure from bright/dark ground. You won't stick the ruler into the bright ground by accident.

Next livestream, I'll play with one of these.

Please shred these ideas apart. What sucks with them? Do you want the old ruler as a user option?

-- Simon

[WillLem]

The default ruler probably does need to be updated.

We can provide a new default ruler alongside the old one and either ask the user to rename the images according to preference (the old one would have "_alt" appended to it in the file system) or, at a push, have a toggle for it in the F2 menu.

Simon's triangle-ended ruler is best. I'm personally not too keen on the stripe, but if others like it then sure, we can go ahead with that design. I wonder: would black & white (as opposed to rainbow) colour cycling suffice here instead?

Colour cycling would always be optional either way:

[Simon]

Hmm, I'm not sure how good the black-white is. The ends of the ruler should come in at least two colors (design criterion #5), the central bar doesn't have to be.

I'm okay making the central bar thinner (make it 1 pixel wide) and keep the white/black near the ends only. Or make the white/black into white/medium grey. Many things to try.

The color cycling is fundamentally problematic. NL color-cycles the gadgets in clear-physics mode, I don't like that either. With the glaring color cycle, the designer is telling me: He has no clue what colors are best, therefore let's cycle through all colors in hope that, sometimes, non-clashing colors are in the cycle. But now the designer has dodged the design problem at hand. His job was to make stuff look distinct. Instead, he distracts me with animation/motion, and there will be clashing colors in the cycle, i.e., stuff looks temporarily the same. Exactly what his design should have avoided.

I wouldn't buy a hammer that is good for hammering nails throughout most of the day, but fails between 09:30 and 10:00 in the morning because during this half-hour, it transforms into a flashing light so I can see it in the toolbox.

-- Simon

[Proxima]

The color cycling is fundamentally problematic. NL color-cycles the gadgets in clear-physics mode, I don't like that either. With the glaring color cycle, the designer is telling me: He has no clue what colors are best, therefore let's cycle through all colors in hope that, sometimes, non-clashing colors are in the cycle. But now the designer has dodged the design problem at hand. His job was to make stuff look distinct. Instead, he distracts me with animation/motion, and there will be clashing colors in the cycle, i.e., stuff looks temporarily the same. Exactly what his design should have avoided.

 

[WillLem]

The color cycling is fundamentally problematic ... there will be clashing colors in the cycle, i.e., stuff looks temporarily the same. Exactly what his design should have avoided.

Whatever colour we make the ruler, it will always clash with any tilesets in that colour.

If we make it blue (for example), any blue tileset will clash 100% of the time. The colour cycle means that tilesets of any colour will not clash 90% of the time.

So, do we want one colour to clash 100% of the time and all other colours to clash 0% of the time, or all colours to clash 10% of the time?

Anyways, it's something of a moot point tbh. It can be turned off in settings for those that don't like it (N.B. The colour-cycling ruler is only in SuperLemmix, it's not yet been implemented in NeoLemmix CE). Besides, your striped (or dotted) ruler solves the above problem, so perhaps should be the default. The colour cycling is (and always will be) an optional extra.

[Simon]

always clash with any tilesets in that colour.
want one colour to clash 100% of the time and all other colours to clash 0% of the time

Tileset designers make lighter top sides and darker bottom sides. Pure white tops are moderately uncommon, pure black is mythic rare. Even the white-black ruler can theoretically run into tiles with checkerboard patterns of tiny granularity, but I haven't seen that in practice.

I'll accept the risk that the ruler clashes with this uncommon type of tile 100 % of the time.

Still, white, hmm. We always measure from top of a tile (where the fall begins) to another top of a tile (where we splat), therefore the white (main color of my ruler) is indeed a concern.

The other worry with white-black is the stark contrast. I'll report findings from tonight's playtesting.

-- Simon

[WillLem]

It seems worth including your ruler in the next update at least for trial purposes, pending further tweaks you may wish to make in the meantime.

The old ruler will remain available as a toggleable alternative. The colour cycling will also be available as an optional extra.

By all means test your new ruler and see if it could be improved. I trust your judgement, and I find it unlikely that anybody would have any objections to at least giving your ruler a try.

Let us know what (if any) issues you find with the new design during playtesting and we'll see if we can brainstorm some solutions.

[Guigui]

As a newcomer to NeoLemmix, I can confirm the splat ruler is kind of a mystery. Was expecting the NeoLemmix Introduction Pack would explain it to me, but it did not.

So I use it like most newcomers : rough measurment, trial and error.

Adding my 2 cents to this thread, I join a suggestion for a different ruler : show Lemmings walk, fall, walk again after surviving the fall. Mirror all of this to adapt to left or right falls.

Here are 3 versions of it :
* show the floors and walls in flashy green color
* show only the floors in flashy green color
* show nothing but Lemmings, makes it easier to surimpose the "ruler" with the actual setting of the level.

[Simon]

Your lemmings make it clear where you should stick the ruler (onto or into the floor), even though you've introduced 1-pixel-thick horizontal green lines. Yes, interesting take, a new angle of attack against the ambiguity problem.

I still want to write about a second application for a splat ruler that we haven't touched in this topic yet: The ruler is not only for measuring fall distance. The ruler is also excellent for inspecting horizontal offset between two terrain pieces, i.e.: Is terrain X exactly vertically above/under terrain Y? Or, instead, is X a few pixels horizontally offset from Y, so that lemmings can fall from X and land on Y? It's easy to test that when the ruler is a featureless vertical stick.

This will be an argument against horizontally protruding triangles (in my white ruler) or lemmings (in yours). The hunch is to minimize/remove protrusions.

If I had to use one of your rulers immediately, I'd pick your ruler with long green vertical lines, but I'd remove the falling lemmings in the center, or at least slice the fallers vertically and show only 2 halved fallers. The goal is to preserve a long green vertical line without protruding features, for measuring horizontal offset of terrain.

I have a weak preference for a white stick with a triangle, but now I want the triangle smaller than what I've shown before. Haven't slept over it. Must still post feedback from stream playtesting. And I haven't tried your ruler yet.

-- Simon

[Guigui]

You're right that the ruler can also be used to see if terrains are really vertically aligned.

Some designers did one pixel shift between terrains, almost invisible to the eye, so your mind is set to : "Lem will fall here" when the truth is "Lem can build, or not splat, from here". This is rather evil, and had me tear my hair out a lot of time. Usually I discover it by mistake when I let a lemming fall unwillingly.

For sure the ruler can help check this, though only for its own height when sometimes you may need it for longer falls.
Anyway here is a version without the faller in the middle. It was just here for decorative and fun purpose, I love this sprite !

Also asking a quite off-topic question her : on my machine, the ruler is a bit laggy. When I press the key it appears correctly, but it stutters when I try to move it accross the screen. To move it, I better make it disappear, then reappear again elsewhere, not very handy.
Is it just me (running NeoLemmix on Ubuntu + Wine) or all NeoLemmix users ?