Must the layers of the Prismatic Wall spell be destroyed in order?

Must the layers of the Prismatic Wall spell be destroyed in order?

You’d think this is answered read-as-written on page 269 of the 5e PHB:

The wall can be destroyed, also one layer at a time, in order from red to violet, by means specific to each layer.

However, I know I’ve seen multiple people around the internet talk about strategies of intentionally destroying certain layers of their own wall like the red, orange, or indigo layers, and holing up inside the wall and shooting or casting spells through the wall without those layers blocking them.

It also begs a question like “If a fireball hits the wall as part of battle and does 25 damage, does it destroy the blue layer individually?”

I just want some validation that you really do have to hack each layer down from red to violet as opposed to losing middle layers intentionally or unintentionally as part of the overall battle.

Does Antimagic Field suppress all layers of True Polymorph simultaneously?

Imagine that you have cast True Polymorph to turn a medium object into a Helmed Horror with Spell Immunity to Antimagic Field. If you were to then True Polymorph that Helmed Horror into something else, it would lose its Spell Immunity, as its entire statblock is changed. However, what exactly would happen if the newly True Polymorphed creature walked into an Antimagic Field?

I see two plausible outcomes:

  1. Both layers of True Polymorph are suppressed simultaneously and the creature is immediately turned back into a medium object.
  2. The most recent layer of True Polymorph is suppressed, at which point the Helmed Horror’s Spell Immunity kicks in and prevents the next layer from being suppressed.

I think each of these interpretations has a decent argument:

  1. This is how Dispel Magic works. If the Helmed Horror also had Spell Immunity to Dispel Magic, the underlying layer of True Polymorph could still be dispelled, because Dispel Magic reads (emphasis mine):

For each spell of 4th level or higher on the target, make an ability check using your spellcasting ability. The DC equals 10 + the spell’s level. On a successful check, the spell ends.

Both layers of True Polymorph are on the target, and so both can be dispelled, and there is no reason to suspect they aren’t dispelled simultaneously (before the Spell Immunity could ever kick in).

  1. Antimagic Field reads (emphasis mine):

Any active spell or other magical effect on a creature or an object in the sphere is suppressed while the creature or object is in it.

And while there are two layers of True Polymorph on the creature, it seems reasonable to assume that only the most recent is “active”, due to the rules for Combining Magical Effects:

The effects of the same spell cast multiple times don’t combine, however. Instead, the most potent effect–such as the highest bonus–from those castings applies while their durations overlap, or the most recent effect applies if the castings are equally potent and their durations overlap.

This (plausibly, I think) could be interpreted to mean that the most recent layer of True Polymorph is suppressed first, at which point Antimagic Field would suppress the next layer if not for the Spell Immunity.

Is one of these two interpretations unambiguously supported by Rules as Written?

Training of two 3×3 convolution layers vs training one 5×5 convolution layer

I’m not 100% sure this is the right stackexchange, please feel free to redirect me to another one.

I know that two 3×3 convolution layers can be equivalent to one 5×5 convolution layer.

I also know that in many cases (maybe all of them), the training of both options is equivalent (in terms of result, not speed or optimisation)

Suppose I have a dataset on which I KNOW that no information whatsoever can be derived from any 3×3 window, but it can be from a 5×5 window.

If I wanted to, I could train a network with a 5×5 convolution layer, then manually transform it into two 3×3 convolution layers.

My question is: could I train a network directly with two 3×3 convolution layers ?

My reasoning: What could the first convolution learn apart from overfitting the training dataset? Supposedly nothing. So, the only layer that could some generalization of my data is the second one. But then, my problem sums up to comparing ONE 3×3 convolution versus ONE 5×5 convolution.

In the general case, the 5×5 convolution would perform better. But with my restrictions on the dataset, maybe it becomes equivalent to the 3×3 convolution? I which case I could train 3×3 convolution.

Thank you !

Invisible layers in Sketch when using Material UI as the basis for UI component library

I’m trying to use the Material UI Theme Editor which spits out a Sketch file as the basis for our component library. I’m then exporting the components to Zeplin for developers to get specs on.

However, I’ve been unable to get the ripples to show up in Zeplin, and they don’t even show up when one UI component is created from the base sympbol in another part of the the Material UI file (here’s ours:

This makes no sense to me…how could the “RIPPLE” that I can see so clearly on the base component disappear from this instantiation of it? I think the fact that I can’t see it in Sketch is the root of the trouble, and why I can’t see it in Zeplin.

Base Component

Instantiation of Component in Design

Maintaining consistency with loosely coupled business and data layers

Take the following sequence of events:

  1. Business layer requests data x and y from data layer.
  2. Data layer returns version 1 of x and y.
  3. Business layer starts performing logic based on data x and y.
  4. Another (concurrent) operation updates data x to version 2.
  5. Business layer instructs data layer to save new data z based on logic at step 3.

The data z saved in step 5 has now been saved based on inconsistent, or “stale” data. The data x became stale during the business transaction. Say for example, data x holds a flag indicating whether the creation of z data is permitted.

In past I’ve seen this issue dealt with by:

  1. Tightly coupling business logic with data operations. For example, business logic in RDBMS stored procedures, or application code co-mingling logic with persistence-aware data operations. or;
  2. Ignoring the issue because the likelihood x impact = too low to be concerned with.

My question: is there a viable 3rd option? Something that ensures consistency while maintaining a healthy separation of concerns between the business and data layers.

Is this something that ORMs address? I know they’ll deal with optimistic concurrency for data writes, but I’m not wanting to update data x, just ensure that nothing else has updated it during the course of the business transaction. Or more specifically, that the “is z creation permitted?” flag hasn’t been updated.

For any platform-specific answers or comments, I’m working with C# and Postgres.

Plotting relief data from tif file with other layers using ggplot

I’m (very) new to R. I’m trying to produce a map of the world with the following features:

  1. ISEA3H grid (with a spacing of 6 miles between hexagons (I don’t have enough memory to generate one with 6 mile hexagons; the closest I’ve gotten is 24 mile hexagons)).
  2. No political borders (i.e., only “natural borders, like coastlines, lakes, rivers, etc.) I’ve got this down as well.
  3. Elevation/relief data. This is where I’m having trouble.
  4. The cell number printed inside the cells. I haven’t gotten this far yet.

Specifically, the problem I’m having is I can’t seem to “mesh” the data from my tif file with my map of the world.

I’m using the dggridR package, which produces an ISEA3H grid of the world with variable spacing. I’ve gotten pretty far.

Here’s my code thus far:

library(rgdal) library(sp) library(dggridR) library(mapdata) library(rnaturalearth) library(rnaturalearthdata) library(raster)  # I've increased the spacing to 500 miles so the map doesn't take so long to render between iterations dggs <- dgconstruct(projection = "ISEA", aperture = 3, topology = "HEXAGON", precision = 7, spacing = 500, metric = FALSE)  # Save a fairly large image at the end of it jpeg("mymap.jpg", width = 9000, height = 6000, quality = 95)  # I'm using the physical data from  # and the raster data from  world <- readOGR("/home/mario/Downloads/data/10m_physical", "ne_10m_land") rivers <- readOGR("/home/mario/Downloads/data/10m_physical", "ne_10m_rivers_lake_centerlines") ocean <- readOGR("/home/mario/Downloads/data/10m_physical", "ne_10m_ocean") lakes <- readOGR("/home/mario/Downloads/data/10m_physical", "ne_10m_lakes") coast <- readOGR("/home/mario/Downloads/data/10m_physical", "ne_10m_coastline")  rel <- raster("/home/mario/Downloads/data/NE2_HR_LC_SR_W/NE2_HR_LC_SR_W.tif") rel_spdf <- as(rel, "SpatialPixelsDataFrame") rel <-  # use grid of the entire earth sa_grid <- dgearthgrid(dggs,frame=TRUE, wrapcells=TRUE)  # actually plot the data p<- ggplot() +      # what should be the relief layer     geom_tile(data = rel, aes(x = "x", y = "y")) +      # the world     geom_polygon(data=world, aes(x=long, y=lat, group=group), fill=NA, color="black") +      # lakes     geom_polygon(data=lakes, aes(x = long, y = lat, group = group), fill = '#ADD8E6') +      # more lakes     geom_path(data=lakes, aes(x = long, y = lat, group = group), color = 'blue') +      # rivers     geom_path(data=rivers, aes(x=long, y=lat, group=group), alpha=0.4, color="blue") +      # coastline     geom_path(data=coast, aes(x = long, y = lat, group = group), color = 'blue') +      # hexagonal grid     geom_polygon(data=sa_grid, aes(x=long, y=lat, group=group), fill="white", alpha=0.4) +      # hexagonal grid     geom_path(data=sa_grid, aes(x=long, y=lat, group=group), alpha=0.4, color="grey") +      # some necessary r code that I don't understand     coord_equal()  # change from flat map to globe p+coord_map("ortho", orientation = c(41.0082, 28.9784, 0))+   xlab('')+ylab('')+   theme(axis.ticks.x=element_blank())+   theme(axis.ticks.y=element_blank())+   theme(axis.text.x=element_blank())+   theme(axis.text.y=element_blank())+   ggtitle('World map')  # finish writing to jpeg 

This is as far as I’ve gotten:

The code above is currently chugging away, and has been for the past 4 hours. It’s remained within the bounds of the memory of my machine, so that’s a good sign.

What is the idiomatic way of projecting a geodesic grid over a relief map of the Earth? How could I include the relief data from the naturalearth tif file with my code thus far?

Are then any good open source examples of data access layers?

When it comes to developing DAL’s I always seem to end up tying myself in knots. I trawl example after example looking for the definitive way to write a DAL and often end up more confused than when I started.

Are there any examples online of open source projects that have good DAL’s? I know for example that Netflix have a lot of their codebase on GitHub that I’ve enjoyed perusing but there’s nothing in that repo that I can find by way of serious data access code.

Are there any other ways to travel between the layers of the Nine Hells besides the River Styx?

I’m planning on creating an adventure where the party become trapped in Nessus, the 9th layer of the Nine Hells, and must make their way through the layers of the Nine Hells up to Avernus, the 1st layer, from which they can find a portal and finally escape.

So far, all I have found regarding travelling between layers is this (DMG, pg. 64):

To reach the deepest layer of the Nine Hells, one must descend through all eight of the layers above it, in order. The most expeditious means of doing so is the River Styx, which plunges ever deeper as it flows from one layer to the next. Only the most courageous adventurers can withstand the torment and horror of that journey.


No planar portals connect directly to the lower layers of the Nine Hells, by Asmodeus’s orders. As such, the first layer of Avernus is the arrival point for visitors to the plane.

The River Styx, then, seems to connect all of the layers, but I didn’t want the party to just sail all the way out, going past all the layers without really experiencing and surviving each one, so I’d rather not use the River Styx if at all possible.

Are there any other ways to travel from one layer of the Nine Hells to another, ideally in descending order (i.e. 9th to 1st)?

This will be for D&D 5e, but information from any previous editions is welcome. If the Hells differs between editions, then assume Forgotten Realms (although information from other settings won’t be rejected for not being Forgotten Realms). I also consider the party’s level or what I plan on putting in each layer of the Hells for the party to deal with are outside the scope of the question, although I will note that none of them will have access to plane shift or similar.

As an example of the sort of thing I’m looking for, Wikipedia’s page on Baator says this about Cania:

Many things lie hidden under the ice of Cania: vast lost cities, frozen armies of creatures. The primary portal to access Nessus is through a gaping hole guarded by 9,999 gelugons. At the bottom of the dizzyingly deep pit is an icy body of water and 1,001 fathoms deep in that is a silvery portal to Malsheem.

However, there are no sources that I can see for this paragraph, so I don’t know if any of this is true. But it at least implies that there are portals connecting one layer to another…