It doesn’t take me long to conclude that Read Mana is attached to something in the center of my back, far enough down that it doesn’t respond to either my shoulders or hips twisting. I’d guess that this means I have another “casting point” attached to my spine or ribcage, like I have on my fingertips, so that spells can anchor to “me” instead of my hands or head. That’d also explain how Read Mana gets a signal to end when I stop casting without it being tied to my fingers.

When I start trying to cast Shield for One, I’m suddenly presented with another thought. How am I casting this spell while walking? I experimentally wave my fingers through Read Mana to produce some ripples, then step to the side while the ripples are still moving, the same way I’d tested Read Mana. The ripples follow me too! They’re even using the same smoothing function that Read Mana is, staying perfectly aligned with the cube as they move in synchrony. Or, more likely, the mana I’m using and Read Mana are sharing the same coordinate frame, and they stay locked together when that coordinate frame moves. Then the coordinate frame is what’s actually attached to that spot in my torso.

Of course, that would mean that spells that are stationary with respect to the environment, like Lightwrite, would have to have a way to subtract out the transform from the world to me so they’re stationary in the world frame instead of my frame. Or maybe there’s a way to just choose which frame a piece of magic uses?

Also, I think that it’s time to define some terminology. My fingers are “casting points”. I am “casting” when they’re active and making ripples. My “origin” is the spot in my torso or spine that anchors the coordinate frame that I cast in and that spells like Read Mana can attach to.

Hmm. I should probably figure out what Read Mana does with a successfully-cast spell before I try using it. I begin casting Find Spellcraft.

As I’d expected, the purpose of the gestures is to create ripples which constructively interfere to build a framework of stable red mana. Interestingly, these shapes, once produced, do affect further ripples. I can see why the book says that mana is like but not actually a fluid; the ripples aren’t behaving exactly like I’d expect waves to, but it’s close enough to a first approximation. For example, in places I can see ripples being pulled in to curve around the surfaces of shapes rather than running into the shape and reflecting. More than a few shapes turn out to be scaffolding for the creation of other shapes, presumably shapes that’d be difficult or impossible to get right with nothing but hand-poses.

I want to say that I’m looking at a sophisticated glider synthesis in a Life-like cellular automaton, where you fire small self-propagating patterns at each other so that their interactions produce useful objects instead of random junk. In the context of exotic models of computation, a glider synthesis is the equivalent of a ten-man circular firing squad all pulling the trigger simultaneously and the bullets colliding in the middle and fusing together to make a wristwatch. But this clearly isn’t a simple cellular automaton. For one, there aren’t any visible cells; if space is discretized it’s at a sufficiently low level that I can’t see it. For another, velocities don’t appear to be fixed to particular angles and they’re not going faster in some directions than in others, both of which would be the case if space was discretized into any reasonable volumetric tiling.

Todo: See if I can increase the resolution of Read Mana until it shows me what’s happening at this system’s equivalent of the “molecular” level.

I get to the final parts of Find Spellcraft, paying very careful attention to every spot I see ripples intersecting. I’m particularly focusing on figuring out the “aspected” mana that the grimoire talked about. The ripples coming off my fingertips have all appeared to be scalar-valued, since Read Mana is displaying them along a single colormap, and that’s not enough for the complex discrete system that the aspects were described as. The construction of the spell is hard to follow sometimes, especially when ripples interact with scaffolding and reflect or slide off.

The final result looks like an x-ray view of a Minecraft world with all the settings cranked up to max. Irregular, organic-looking blobs sprout off of or intersect with other blobs, interspersed with the occasional spherical reflector, small plane, or regularly curved segment. There’s a surprising lack of surface roughness given that I essentially sculpted this with my hands. Maybe there’s some smoothing going on? Maybe it’s just a result of the ripples spreading out as they move. Either way, I’ve produced an incredible little piece of modern art. If I could stabilize this I’m sure that I could donate it to a museum or something.

Of course, there’s so much going on that I predictably, if annoyingly, miss some of the important final interactions.

And it’s easy to see that some of those final interactions were more important than the rest! I pause, a few gestures before the end, when part of the sculpture is suddenly full of life, a stream of tiny blue chips of light blasting through the tumorous mass and then spraying off in every direction, like angry laser confetti. There’s so much going on that I can’t figure anything out at all. Maybe some parts are denser or looser, but whenever I try to lean in to get a better look the whole sculpture moves away from me.

I grumble and do the next gesture, which makes my hands pass through the sculpture where I can’t see what’s going on. I notice while doing that that the vibrating sensation I’d noticed while casting indicates that my fingertip is passing through one of the more focused streams of lights. Not too surprising. When I finish the gesture, the blue chips of light are joined by another shower of green chips, this one looking like it originates on the other side of the sculpture. I also see a few flying particles of other colors, but they seem to be almost completely random.

It’s clear that I’m not going to get anything by inspecting this mess, so I keep casting, each gesture adding new and exciting colors. Eventually the whole thing is a whirling mass of technicolor motion, like some kind of science fiction dynamo. It’s much more organized now, with clearly identifiable streams of colored light being funneled around and intersecting with each other to produce other colors, but it never entirely stops throwing glitter in every direction.

I finish the spell, tapping a point in the middle of the mass with my fingertip, and the spell… does nothing except continue spitting out flecks of light in every direction. I must have done something wrong.

I dismiss the spell and start over, this time trying to ignore the colors and focus on getting the gestures right.

It takes me three more tries, but Find Spellcraft’s cone of light finally springs out. Even better, I now have some idea of how it behaves! The end result is the same scintillating whirlpool as before, but now it’s attached to the end of my finger, and it’s throwing out a uniform mix of pale blue, dark blue, and gold dots in a cone that perfectly matches the extent of the display.

Interestingly, now that the spell is working, I’m starting to see some of the limitations of Read Mana. In particular, while Find Spellcraft is rigidly fixed to the casting point on my right index finger, the streams of light lag behind the sculpture! When I wave my hand around to test the lag, the display separates out by color, different… kinds of mana lagging by different amounts? I do know that this means the dots aren’t going at “actual speed”; it’d take multiple seconds for one of the dots to reach forty-five feet out but Find Spellcraft reacts much more rapidly than that.

Read Mana must be even more complicated than I thought it was, then. Could be anything - simulating the trajectory of a bit of mana, dropping a marker into the mana and letting it drift the same way you might drop a bit of paper into a stream, tracking the path of a single “mote” of mana and then replaying it slower, there are plenty of ways to produce that effect. The different colors lagging differently makes me suspect a replay, though, especially given the way they pass through the red shapes sometimes.

Everyone is curiously looking my direction, dazzled by all the activity going on in front of me.

I quickly pull out the lightmaker trinket, hand it to Bob for activation, and point Find Spellcraft at it. I soon see a trickle of green sparkles returning, a thin line terminating in the dot on Find Spellcraft’s display.

Liv apparently gives in to her curiosity. “Figured anything out yet?”

“Not really,” I say. “Find Spellcraft constructs three different kinds of mana, then flings them all out into the world,” I explain, thinking out loud. “Some of that mana eventually comes across a spell, returns, and is converted into a dot on the display. I’m not sure precisely why it’s radiating three different kinds of mana and getting a different kind back.” Hmm. Brainstorming time. “Sense, measure, return, and then something to combine with the return mana to display? A time-of-flight measurement using one kind of mana, a mana for sensing, and something to combine them into the return mana that handles display when it times out? Or, maybe it’s just two different potential fields and a single mana that does all of the out-and-back logic and Read Mana just displays it differently for some reason.”

Nobody seems to know how to react.

“Eh, don’t worry about it, I’m just rubber-ducking at you,” I say absentmindedly. Maybe some or all of the colors are actually byproducts of the process that creates the functional mana?

“Rubber ducking?” Liv asks, baffled.

“…What? Oh, oops, sorry,” I say, tearing myself away from my questions. “A lot of the mistakes you make in software engineering are because you don’t actually understand the problem you’re working on, you’ve gotten confused and made an incorrect assumption or missed an interaction or just not defined the process correctly and done something invalid. Like, if you’ve ever found yourself absentmindedly trying to brush your teeth with your hairbrush?”

No, don’t ask me why that particular example comes to mind so readily.

“I sort of know what you’re talking about,” Heather says. “A form that looks applicable, but the last space asks for an ID that you don’t have. On further inspection the form is for a completely different problem.”

“That’s not a bad analogy,” I tell her. “Except this kind of issue can be incredibly insidious because you can’t track the misalignment down to a particular box, and every time you try to go over what you’re doing to see where it might have gone wrong you make the same mistake.”

Heather winces.

“Well, the only real way to deal with the roadblock is to force yourself to re-examine the situation from a different angle,” I continue, “where the incorrect assumption doesn’t come into play and can’t confuse you again. One way is to take a problem that you’ve been working on with paper and pencil and explain it to someone else out loud. It’s hilarious how often you spend two hours on something, finally give up and go bother a coworker for help, get two sentences into your explanation of the situation, and realize exactly what you’re doing wrong.”

Agnes reacts to that one, surprisingly. “A sight not uncommon in seminary,” she laughs. “A novice who raises their hand and answers their own question with their second sentence.”

“Yes! The interesting thing is, you don’t actually need another person for this,” I say excitedly. “You just need to force yourself to reframe the problem. So a lot of programmers keep stuffed animals or these adorable little rubber ducky toys next to their computers and monologue at them when they’re having trouble with something.”

“No wonder yer so scatter-brained,” Bob realizes. “Y’gotta pay attention ta ever’ single thing that goes through yer head jus’ in case it’s the thing yer’ wrong ’bout.”

“And then you get off-topic because you keep following digressions, yeah. I mean, some of that’s just my style,” I sheepishly admit. “But you’re not wrong! Just, like, stop me if I start really going off on a tangent?”

“I can do tha’,” Bob says, amused.

Nobody else has any questions for me and I think I’ve gotten what I can out of Find Spellcraft, so I move on to Shield for One.

Having Read Mana up while I attempt to learn Shield for One is surprisingly helpful, mostly because it’s easy to tell when I’ve messed up a gesture and totally failed to produce a new shape. I’m producing plenty of colored light, and some of it’s even bouncing around merrily inside the shapes, but none of it’s doing anything particularly organized.

That said, the complexity of the spell is apparent, and I’m still having trouble when we reach Calfort.