Lighting Engines

Studio Design I Project IV

Jenny L
Jenny L
Nov 6, 2018 · 25 min read
Nov.1st Class notes


We started a new project today by thinking about how do we experience the world. One major way we perceive the world is by seeing, where lighting is an indispensable part.

various amount of light sources occur in our daily lives, they all have different properties and functions and, moreover, they make us feel in a certain way. I noticed that lights are tightly linked with emotion. For the next class, I will be walking around the campus at night time and take photos of different artificial lighting equipment, their effects on objects and reflect on the emotion these lights can bring to us.


Looking at different artificial lights

Here are some of the photos that I took. I analyzed them with an ‘emotion wheel’. I noticed lights guide me to emotions in the “happy” section of the wheel generally have warm colors (yellowish white). On the other hand, cool tone lights that are too dim for me to see the surrounding makes me feel lonely and sad. If cool tone lights are too direct (harsh), these illumination equipment lead me to feel threatened and anxious.
Also, I noticed the lights are either direct or indirect. Here is an example of indirect light: lights in Margret Morison hallway shoot light up to the ceiling; the light lit up the ceiling and light bounce from the ceiling to lit up the whole hallway. Such diffused soft light created can be considered as an example of “ambient light”. Linking back to the emotion wheel, such lighting doesn’t give me much emotion. Example of direct light is the kind of lights we have in the studio — lights shoot directly from lights hanging on the ceiling to our desks, thus, helping us to see when we work. When been exposed to this kind of light, I feel a bit insecure since I feel I have been put on a spot (lack of privacy).

emotion wheel and lights in the physical world around me

More photos + Discussions

The photos we took and Stacie’s class notes

As a group, we shared our photos. We use a huge range of vocabularies to describe different lights. This light is been described as a “fairy light”. The light isn’t very intense and can barely light up the environment to a level that my eyes can see clearly. The main function would be to decorate the working space. The light is warm-toned and the light source is quite small, making the light actually quite harsh.


The purpose of this light is very different from the former one. This small red light informs the status of the washing machine. The color of lights usually contains extra information. To illustrate, red lights are usually linked to messages such as “in progress” or “emergency”. While green lights inform people information such as “passed” or “charged”.


This is the studio light. Since the main function is to light up the workspace (so can be put into the WORK category on the whiteboard) It is very bright, direct, harsh. Normally we don’t look directly into these lights, but when the lights shine directly into our eyes, the eyes hurt badly. (gives me a bad headache :-( ). However, the workspace in the studio is well lighted and allow drawing, modeling, and discussing activities.



Playing with light and paper

On a rotating plate, with a light bulb, I tried how different papers, molded into different forms. I experimented with Bristol paper, yellow card stock, newsprint (light grey color), black origami paper and cardboard. Here are some of my experiments I found interesting.

Bristol paper > hard paper, not wacky, easier to work with
Yellow card > light pass through yellow paper, producing a really warm light

By doing these playful explorations, I noticed a few things:

  1. Different kinds of paper have different “transparency” (different amounts of light can go through different materials). For instance, no lights can go through a single layer of cardboard, while lots of light can go through multiple sheets of origami paper.
  2. The traces of folding/ sticking (with tape) that are merely detectable when lights are turned off, but become very obvious when lights are turned on. I have two distinct options here to deal with these traces: 1. to minimize places of folding and sticking 2. to use the pattern created by sticking and folding (make it into a feature of the lamp )
  3. I also noticed the paper look very different when lights are on and when lights are off.
  4. When rotating the wheel under the light, the forms of paper shells also vary drastically (in a 3D space, changing viewpoints → form is also changing).
My first thoughts of the context given to me

The context I got is “Eating a meal with a few friends in a home”. Since a significant component of my context, I visited a few family dining places around campus to see the lighting there. I plan to look at:

  • What are the forms of the light?
  • Except for decorative function, how does the form inform its usage?
  • How to turn it on/off?
  • What quality of light does it produce?
  • How do people that are dinning react? (Actions? Sayings? Emotions?)

Here are the eating places I visited during and my notes:

Taiwanese Bistro Cafe 33
  1. Thin Cylinder from (lights directly goes up and direly shin down to the table however the lights goes up are much more harsher, while lights fall on tables are more defused.)
  2. Since the overall form is pretty simple, the main function of the light is to provide sufficient light (rather than attract people to view them)
  3. Warm, soft, indirect & direct, cozy.

When I see other people are photographing food, I realized I have to look at the lighting effects on the food and people that are eating. In this restaurant, dishes are properly lighted and look quite inviting. Shadows are present but not very heavy nor very sharp. The level of lighting supports activities when eating (see one anthers’ action, read the menu …).


Another component of the context is “home”. I went t IKEA to look at some “dinner table lighting” and “home lighting”. Some findings are:

  1. Most lights hang from above
  2. There is usually a series of lights that can light up a whole long rectangular table, making sure all the seats are equally lighted.
  3. Lights usually directly shine from above. Some go to side-ways and go up.
  4. Lights produced will be quite soft (light source defuse from the light bulb and spread downward, making sure all the seats are equally lighted)
  5. Warm color lights.

Inspirational quotes I found when reading some articles about lighting + eating

Enlightening: Sara Manuelli samples the lighting design at three different eating and drinking establishments

“LIGHTING design is largely responsible for creating the right atmosphere in restaurants and bars. From classic to ethnic, lighting designers work closely with architects and designers to conjure up a mood and reinforce the identity of the space. (Manuelli)”

LIGHTING: Light meal

“ There’s a real trend for lighting to support the design and be largely invisible, except for feature pieces,’ says Filskow. ‘It’s not about flooding the room with light, but about creating intimate, homely spaces.(Lorenz)”

ROUND #1 - First Ideas + Sketch Models

Ideas on paper

The Problem: It is hard to draw light effects. And the light effect in my imagination might not be the same in the reality.

Evaluations of sketch models: What’s working? What stops it from working? What can be further developed?

Variation #1
  • Lights are successfully delivered both upward and downward. The lower stream shine light to the ground is the major stream. left and right sides seem to be too dim.
  • The light produced is soft, while bright enough to light up the environment. (dining rooms at home will have ambient light, so this won't be the only light source in the environment)
  • The form is not very pleasant to look at when the light if off.
Variation #1
  • I modified the proportions of the last model, so light can be guided in a way that spread to a wider range at the bottom. The lights also spread to surroundings. Comparing to previous models this makes the whole room more lighted up.
Variation #2
  • The idea of a very flat cone seems to be working (spread light to a quite large range). This can be used in other models.
  • the outer most layer is a very flat cone, and each inner layer has slightly more slanted side, narrowing the cone of light.
Variation #3
  • From the outer layer to inner layer, the materials are: Bristol paper, pallet paper and tracing paper. This arrangement helped me to make sure major stream of light goes down, while the brightest part gets lighted up in the middle. (it is going to be placed in the middle of the table, so the center of the table where the main dish is can be well lighted and each person sit around the table can receive the same amount of light ).
  • I placed the very flat cone shape up-side-down.
  • the upper part is not covered — light too harsh. For all subsequent models, I should make sure the bulb is fully covered.
  • From the middle image, it's obvious that the mainstream of light goes up. This is not suitable for a dinner table.
Variation #4
  • The shape doesn’t quite inform the placing of the light (if the base shape is a circle, users know the light should be placed in the middle of the table).
  • The form of the lights gives the static, hard and straight impression (straight edge). This is not consistent with the context “eating with friends”, which should be related to happy, relaxed, unity, bringing together.
  • This kind of paper performs better than Bristol paper and tracing paper. (not much texture, allow a suitable amount of light to pass through)(YUPO paper)> But what are the original purposes of YUPO paper? The look like plastic?? Are they not truthful enough as papers materials?
  • It is the scale that allows this light to look like a dinner table lighting equipment.
Variation #5
  • This is a spiral shape I haven't tried in previous models. The folds are very nice when the light is on. And when the lights are off, the form is more pleasing and interesting than previous models. I can detect certain movements in the lights; there is a certain rhythm in the form that makes it less static.
  • Lights go more vertical rather than horizontal. This might result in an insufficient amount of light for everyone around the dinner table.
Nov.11th Class Notes

Things I would consider + bring to the next stage of making prototypes:

  1. Be truthful about the materials: They are papers. So consider what properties paper have. What does paper do well?
  2. The ‘wholeness’ of the light: not be seen as separated parts (eg. the stand, the lift, the diffuser…).
  3. the form should inform its usage and positioning.
  4. Consider the environment that the light will be placed in what kind of table? what kind of room? View the design in the context.
  5. The ‘small’ to ‘large’ form
  6. The idea of kicking lights mainly downward, slightly sideways.
  7. Provide light for everyone
Ideas for round #2 on paper

Use semantic differential to Evaluate current sketch models in round #1

My semantic differential

Stacie shared an idea-generating + evaluation tool called semantic differential. It is a set of scale. There are two opposite adjectives on two ends of a scale. (eg. big vs small, fast vs slow). With these, I can “measure” my context and my models on the same set of scales and see if some properties are off. I realized, in terms of properties I have on my scales, some of my sketch models differ greatly from my context. This is not a bad thing since at least I know I should not go into these directions and I can generate some new shapes to work with. There are some current ideas can be brought to the next stage.

Rough models don’t work well: What can I change in the way of making models?

Another significant thing I learned is: There are a different kind of models and each kind have different functions. Quick sketch models able me to quickly transfer a form on paper into the physical world. But in this case, for the lights, if my edge works are crapy (bad craftsmanship), I can’t see the form of the light properly. Also, if parts of the lights are not put together properly, lights leak from sides of the light, prohibiting me to see the idea effect of the light. I certainly can’t evaluate the light produced if the light produced is not what I want it to be like (this is quite complicated to phrase ).

Overall, in this case, models need to be accurate rather than quickly generated (accuracy>speed). So moving forward, I will be making more refined models.

ROUND #2 — Making more refined models of different variations: Problems with current models

I generated more refined models. After making them, I reviewed some of them with Steve & Stacie in class and others on my own. Here are my evaluations of the modes (Problems+Things could be brought to the next stage):

Variation #6
  • The light doesn’t have much movement to it. This light feels stagnant and too mechanical due to its form.
  • The form of the four sharp corners is too pointy and aggressive for a dinner table.
  • Light for dinner table should inform users of ‘socialness’.
Variation #7
  • The form just reminds audiences of a floating disk/spaceship.
  • The form feels too mechanical. The form doesn’t have much energy in it. Therefore, the form doesn’t fit a dinner table, where people move quite a lot and actively communicate talk to one another.
  • The up-side-down cone doesn’t really look like its made of paper. Not truthful of the material. (What is YUPO paper?)
  • The thing attractive here is the folding technique of the two cones, rather than its essence as a light. In this case, the ‘making’ distract users from the form. Noe the piece is about the folding technique, not about the light.
  • Something I could work on is the proportion of the folds.
  • The very flat form makes it look like a hanging light.
Variation #8
  • Again, the form here it too static. The geometry is too regular and even. It might be usable and even useful, but not desirable for a dinner table.
  • The transformation of a polygon top to acicular bottom is too subtle to inform viewers the ‘enlargement’ movement want to communicate.
    Although I could adjust the top and bottom proportion to solve the problem, the overall inherited form is still boring and stagnant.
  • the inherited form of the light is kicking lights out horizontally, providing a sufficient amount of light for the table beneath the light
  • There are two movements communicated with the form of this prototype: ‘enlarging’, and ‘twisting’. This model has more energy than all other variations. There are more ‘socialness’ in this variation.
  • The rotation angle of each layer now is pretty random. Viewers can’t tell if the layers are all supposed to rotate in the same angle, or they adopt certain other patterns. Since, for this model, I placed the layers randomly but they seem to be following some pattern, viewers will feel something is off with this light.

While all of the more refined models in round #2 could be improved in some way (either form, selection of material or size), I think it is time to focus on one of them to develop further. I decided to develop this one further. With all the knowledge that I have learned/discovered in making all previous variations, I evaluated this model in more depth. Here are some problems I would like to solve in my iterations for this prototype:

  1. The pattern of rotation is not clear. (equal in rotation angle? enlarging angle? )
  2. The enlarging movement within the form is not strong/ active enough.
  3. The height & base proportion make the overall shape of the light doesn’t speak it is a hanging light (Currently, it could be seen as a lamp sitting on a table).
  4. The light could be brighter and the color could be cooler (because the light not only needs to illuminate the dishes but also people’s face to see their facial expressions and body movements).
  5. Lights coming out of the small corners goes both upward and downward. The lights shooting up could be dimmer to inform the light should be hanging. And the mainstream of light shoots downward.

Solving Problems

My plan to solve the problem

Some of these issues have multiple solutions ( for instance, to enhance the enlarging pattern in the form I can either increase the thickness of each layer or change the size of the base shapes), so I will be making multiple iterations of models to test these solutions physically and see whether the change enhanced the design or do I have to think of another solution.

ROUND #3 —Prototypes of Iterations

For each iteration, I will be evaluating them, looking at their problems, as well as ideal properties that could be brought to the next stage.

Iteration #1


  • When the lights are turned off, the overall form of the light seems to be quite rigid, aggressive and boxy. (feels like conference room lighting)
  • The square form has a sense of seriousness to it. However, eating with friends feels like a more relaxed and informal event.
  • The size is too big. Now the light seems more like a light for a conference room.

Good points

  • the height — base proportion make the positioning of the light obvious (viewers: oh the light should be hanging rather than sitting on the table)
    Users can detect a ‘rhythm’ in the form of the lamp. The rotation pattern is very clear, especially when the light is turned on.
  • Adding a layer of tracing paper to the top of each layer make lights shooting up dimmer. Therefore, the mainstream of light travels downward, proving light for all friends around the dining table.
Iteration #2


  • The Bottom layer is too thick, making the form seem to be boxy. The thickness also lowered the illumination level of the bottom layer. (since the bottom layer is a major part of the whole form, overall, there will be less light produced by this iteration).
  • The rotation movement becomes less apparent.
  • The interspace between layers might be too small. Less light can pass through the small triangles. Not only the illumination level is lowered, but also the light produced is as interesting and stunning as other iteration.

Good Points

  • the process of enlargement is very obvious.
  • The form is still interesting to look at when the lights are turned off.
  • The base shape and the rotation make the form speak an active, vivid and alive feeling, which is consistent with the mood of eating with friends.
Iteration #3
Plan of the top view of this iteration


  • The core idea of this iteration ends up of being not very obvious or not even noticeable by viewers. The design idea is: the shape of each layer changes (base shape for each layer have one more side than the upper layer). I intend to make the shape less rigid by doing so, but viewers can hardly notice I have changed the shape when viewing from a lower angle.
    Since I have changed the shape for each layer, the rotation pattern became quite confusing and messy. They look like unintentional mistakes or just a stack of shapes randomly stacked together > no rhythm and pattern in the form.
  • Like the first iteration, the scale might be too big. Shrink the size down a little is likely to make the form have more ‘socialness’ and ‘life.’
  • The lower three layers obviously become dimmer than the upper layers, but The significant steam of light should travel downward, providing a sufficient amount of light for everyone around the table.

Investigating YUPO paper

I Used YUPO paper in one of my models and I like its especially smooth texture. But since its too smooth and the light produced is too even, it reminds people of plastic, rather than paper. I feel the material is not truthful enough, and give an inorganic feeling. I wonder what these papers are? (Material, usage?)

about Yupo Paper and Synthetic Paper

Now I see why YUPO paper is like plastic and doesn’t give an organic feeling. They are made of petroleum rather than trees. Plastics (called polypropylene) are heated, compressed and become YUPO paper. Knowing these facts, I feel YUPO papers feel even more “lifeless” than before. (they should be called plastic!) I now have a deeper understanding of being ‘truthful’ with the materials we are working with and I feel this is crucial to me. I’ll work with papers that have texture in them and I’ll use them in ways that are suitable for paper’s nature.

ROUND #3 My next Iterations based on evaluations

The idea on paper
The prototype


  • The top three rings seem to be quite decorative. The rings feel too thin and weak, while the rest of the light seems to be bold and strong. So the top rings look like added-on pieces, rather and forming a whole form with all bottom layers.
  • The light looks quite complicated, especially when viewing from a bottom angle since it has so many layers and each layer is too thin.
  • the light produced could be brighter to support ‘eat with friends’.
  • Although there is a quite strong movement going on with the form, viewers can’t clearly figure out what the movement is and what’s the pattern within the movement. This also makes the light looks complicated and ‘heavy’.

Problems pointed out by peers

Nov.20th Peer review
Class Notes Nov.20th

learning from others’ prototypes

Steve and Stacie discussed a few different lights with the whole class. Some problems they pointed out on others’ models also appear on mine. Here are some key issues I want to stress+reflect on:

  1. Design flaw vs Crafting flaw: some models I saw today are more on the sketch model side rather than on the prototype side. They are trying to carry out rough ideas that the designer has. But the craft might be too dull to demonstrate the concept to others. In this case, the crafting flaws have hindered viewers to see if there are any design flaws or not. There are two solutions to the problem. One is to improve the craftsmanship, while the other is to simplify the design.
  2. Make symmetrical shapes symmetrical: if a form is intended to be symmetrical, then make it perfectly symmetrical. Viewers can easily pick up on flaws in symmetrical shapes. Even if an angle is just slightly off, the whole form ac become unconformable to look at. And viewers’ full attention will be on the flaw, therefore ignoring the form of the light.
  3. Asymmetrical shapes should come together as a whole: asymmetrical forms I saw in class today are often made of various different pieces. In many of these forms, the pieces don’t resonate with one another. This means that the light looks like an assembly of small paper pieces.
  4. Designing a 3D object: lights appear in the three-dimensional world, meaning that they have to be viewed from various different angles. For good designs, no matter viewing from which angle should remind viewers of the same object. There should not be a ‘front’ side and a ‘back’ side to the form. Views from different angles should resonate with one another.

Solving problems and why

  • Reduce the number of layers: only have three layers. Since I simplified the design, I’ll encounter fewer problems in crafting it.
  • Align the bottom layer and top layer; only rotate the middle layer: This way, I can keep the ’movement’ in my light and the movement is still dramatic. And since I have made the movement simpler, it will be easier for viewers to see what is actually going on.
  • Increase the thickness of each layer. As pointed out by my peers, they can see the light bulb directly when the layers are too thin. To block the blinding light, I’ll increase the thickness.
  • the thickness of layer increase from top to bottom. I have tried this before in a previous iteration. The enlargement pattern looks neat and is pretty clear for viewers.
  • Use tracing paper to enclose the very bottom layer. Current paper choice has made the light produced too dim for eating with friends.

ROUND #4 the final prototype: problems+ solutions

the final prototype: on — detail — off

It works quite well except the adhesives are quite distracting when the lights are turned off. (shiny part of the tape!)

Evaluating different adhesives

In all of my previous iterations, I intend to hide the adhesive places. However, I haven’t figure out a way to hide them completely. So I experimented/examined different attachment methods to see which method is more ‘invisible’.

Left to right: shine-finish tape, glue dot, matt-finish tape, and hot glue

Shine-finish tape: Light can pass through the tape. This means that it won’t block the lights. However, when lights are turned off, its shine finish has a texture that is so different from the blotting paper used on the body part of the light. As I discovered when making the final prototype, this will be very distracting.

Glue dot: it is very small compared to other adhesive agents. It is almost invisible when viewed from a distance. However, the adhesive isn't strong enough to hold the whole structure. Although not visible, it is not functioning well.

Hot glue: it is easy and efficient to work with hot glue. And the adhesive is pretty strong. However, it seems to be quite messy when I applied too much. Spots where I applied hot glue is not very obvious when the light is turned on, but very obvious and distracting when lights are turned off.

Matt-finish tape: The tape is not fully translucent. When lights are turned on, areas, where I put tapes, will be slightly dimmer than parts without tape. When lights are turned off, the part becomes not very obvious. The texture is also quite similar with blotting paper (both Matt-finish).

Decision made after evaluation: Overall, after evaluating all these methods, I’ll be using Matt-finish tape to attach the bottom polygon to sides of each three layers. Because they won’t be too shiny and, therefore, distract users from enjoying the form of the light when the light is turned on. Also, some light can pass through the tape part when the light is turned on. I will also be using hot glue in my final light engine to attach the layers together. Such adhesive is the strongest adhesive tool among all tested. If used carefully, and only apply the glue from inside of the light, it could be almost invisible. I realized it is almost impossible to join papers together ‘invisibly’ and separated pieces of papers are not meant to be joint together (It is part of paper’s nature). So’ll fold the paper strip and connect the beginning and the end with matt-finish tape since its the lest obvious method.

ROUND #5 the final light

Takeaways from the project

Designing is about experiencing

In this project, a step that is very crucial for me is to walk around campus and observe the lights in real life around me. And I paid close attention to the lights at places where I eat with my friends. While experiencing my given context, I started to question myself: what’s the mood of the light? How does it aid me and my friends’ dining experience? Are these lights warm/code? harsh/soft? …. All these questions that I can only gain answers through personal experience have helped me to understand the given context, therefore, helped me to generate a wide range of appropriate variations for dinner table lighting.
If I got this project at the very beginning of this term, I would go on google and search what do dining table lights look like and start drawing with that information gained. Photos taken and posted online by others are there because of specific reasons (to advertise, to explain…). This means that they intend to guide audiences to think in a certain way, to get a certain message. In other words, they are not truthful. However, if I walk around campus and see the lights myself, I can interpret, analyze and feel the lights truthfully. Since I can base my designs on truthful observations, I think I’ll be more likely to start in a good direction.

Originality and creativity: See designs in contexts

While it is fine to look up paper light designs online and even reproduce them, how I use other people’s ideas is crucial. Try to be a hundred percent original is hard and unique designs are not necessarily good designs. For me, looking up existing paper lights and reproduce them helps me to understand the properties of papers that I am working with and mature techniques of making paper into forms. However, completely copying a design is first, not ethical, secondly, boring and thirdly, not suitable for my given context. Existing paper lights were designed to solve their problems, and I have my own set of problems associated with eating with friends. Why use a solution for problem A to solve problem B? Problem B probably won’t be solved well.

know the material that I am working with well: both physical aspects and the contextual aspects

One of the shocking discoveries I made in the project is that Yupo papers are actually more like thin pieces of plastics since they are made of petroleum. I then realized how this material has lied to me and I don’t want to work with it anymore. I then searched how tracing paper and blotting papers are made and their initial usages. I feel I am more connected to these two kinds of papers that I am working with. And I can use them and shape them in ways they are capable of and good at.
Overall, personally, understanding the physical properties of materials and their process of making/history/original uses is fun. Knowing these just make me enjoy using them more.

The relationship between drawings and designing 3d objects

Drawings are 2d. Even if I am doing perspective drawings, the drawings themselves are still 2d, letting viewers see a specific angle of the object. Since I am used to drawing, I am very used to think in 2d: what do the front and the back look like? what’s the view on both sides? It is so different when it comes to designing in 3d spaces. there are a countless amount of viewing angles and viewers’ viewpoints will keep on changing as they move around the object. In this project, good light designs should remind viewers of its consistent form no matter where viewers are.

Functions of different types of models

So far in studio class, I have used two types of modeling: sketch models and prototypes. Although they are both ‘not the real thing’, but they have their own properties and serve distinct functions. Sketch models can be made very quickly. They function as tools to quickly actualize designers’ ideas and test a certain part of the final real product. On the other hand, prototypes are made with much more detail. They should resemble the final product as close as possible. Which one should I use all depends on the situation and the reason behind making them. To illustrate, If I am trying to show my design concept to Steve and Stacie, I’d make highly finished prototypes so that they can clearly understand my design concept rather than being confused by crafting flaws. If I just testing the properties of papers, I’ll make a bunch of rough sketch models since this solution is fast and effective. If I am brainstorming ideas, I would draw. Because drawing can help me to get ideas out of my mind and help me to think.

What is decorative and what is not?

It is sometimes tough to judge whether a property of an object is decorative or not. (although sometimes decorations are easy to identify, for instance, a box of strawberry has a strawberry pattern printed on it.) regarding this project, If a pattern can strength a lighting engine’s form and can resonate with the form, it can be considered as not decorative. So I questioned myself: how can I justify if the shape resonates with the form? Should I explain it with reasoning or with my feeling? Actually, both of them could be used. The problem is: I am trying to find a highly simplified summary/statement as my design principle, but Well-done designs can be very different and good in their own ways.
Going back to decorations and functions, there might not be any written materials stating how are decorative features different from function features. And it is almost unnecessary to clearly differentiate one from the other. In the end, I couldn’t even think of a reason to completely separate one from the other. As humans, we naturally have the ability to be both logical and emotional at the same time. The key is to find a balance.

The end ~

Jenny L

Written by

Jenny L

More stuff →

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade