Prototyping Printed Circuit Boards(PCBs) is now cheaper and easier than ever before. We are seeing a huge increase in the number and quality of small batch custom boards, like DefCon badges.
In its simplest form, a two-layer PCB consists of three layers- a copper layer, a substrate like FR4, and another copper layer. The substrate is generally a lightweight, stiff material. High-quality PCBs have strong adhesion between the copper layer and the substrate.
Desktop CNC’s allow milling PCB’s in-house. When milling your own PCB, an engraving bit is used to mill away the copper on the top to leave the insulating substrate, thus making electrical traces. A thick layer of the substrate is left to provide structural support, as shown below.
The substrate is quite stiff and does not allow bending at most temperatures(although, it has been done). But the copper is quite malleable. So if we mill away most of the substrate, a point bend can be made as shown below. The resulting cut in the PCB can be soldered to maintain the 3D shape as well as to keep electrical conductivity.
Note that this technique is very different from using flexible or semi-flexible PCBs. For one it is much easier to prototype and it results in rigid 3D structures.
I started out exploring shapes afforded by bending PCB material. Specifically, I explored making Miura fold which might allow warped 3D shapes from a single piece, and making a semi-dodecahedral lamp.
After spending some time with the material and process I had a better idea of the possibilities, and it became clear that making the entire toy using this technique is unfeasible. So as a next step, I made 2.5D mock-ups of the final design in 2.5D by stacking paper, first hand-cut then designed and laser-cut.
Prototyping Printed Circuit Boards is now cheaper and easier than ever before. This is because of automation of the production process in a large-scale bringing down cost of manufacturing. And, the availability of free tools and tutorials to design circuit boards.
The container offers affordances in the form of layers of materials sandwiched together. The layers are silkscreen, soldermask, copper, FR4, copper, soldermask, silkscreen. From the perspective of a designer, the material offers more affordances than can be easily counted, so I shall focus on affordances to the user of a produced PCB. Here are some of the affordances and great examples of their use to create good content:
The copper layer affords passage of electricity. All the other layers provide the anti-affordance of blocking the flow of electricity. e.g. PCB motor
The copper layer provides the anti-affordance of blocking light from passing through, all the other layers provide the affordance of letting light pass through to varying degrees. e.g. PCB visual Art
The copper layer affords sticking to solder and soldermask affords the anti-affordance of repelling solder. Together these afford sticking components on the board strongly and precisely. e.g. PCB Braille signs
Response to in-class questions:
What is content? What makes “good content”? What makes “viral content”? How are they the same and how aren’t they?:
Good content satisfies needs, viral contents satisfies wants.
How is it useful or detrimental to think in terms of content/container? Is there another way?:
Container/content model separates the users from the designers i.e. designers make the container, users use the container to make content. Granted there is often overlap in the people acting as designers and users i.e. a person can be both a user and a designer. Nevertheless, there is seldom an overlap in the thought process of acting as a designer/user i.e. a person seldom acts as both at the same time or in the same mindset. So the divide is valuable. At the same time, I feel like better design emerges from a narrow divide between designer and user, and both should inform and influence each other.
Can the same “content” be translated across different containers?:
Yes, definitely! For example, a painting can be made in any medium, a piece of music can be played with many instruments, The same spice/texture palette can emerge from different ingredients. But also, synesthetic nature of humans means that similar emotions can emerge from different sensory inputs.
Can the same “content” be translated across different containers?:
I put container first, more often than not. In fact, I see the world in terms of simple components e.g. I see a toaster as a heater connected to a timer rather than a means of generating toast(the content). My process is to explore the perceived affordances of a container, drill down to the real affordances and think of which affordances are invisible. Then, make them visible! For example, the copper layer of a PCB affords bending, but this is seldom visible because the FR4 layer affords stiffness. I am currently exploring making bent PCBs to surface this previously less visible affordance.
Can the same “content” be translated across different containers?:
Yes, users deserve compensation from the content they create. Even if it is non-monetary compensation in the form of reputation/credits..
Reflections on the reading:
Two forms of bad writing that I hate are (a) jargon-heavy text with runaway sentences (b) writing that makes universal claims without citation or context. The reading Rethinking Affordance- An Introduction seems to combine the two into a perfect mix of misery. After reading the piece three times I still don’t understand what the author wants to convey other than that they think that ‘affordance’ needs to be redefined for the digital age. I have more issues with the text than can be easily listed, so I shall not attempt to list them.
I have worked as an engineer for a few years now. And since the very beginning git was part of the workflow. I started out by pushing back. I forced people to send me zipped files and did the version control myself. This works for small projects but it’s terrible for maintaining anything larger than a text file.
I reluctantly accepted git and GitHub; slowly. I started out learning git for maintaining personal projects, with low stakes. I maintained my CV in git, maintained a few Arduino codes here and there. No big deal. For the most part, git was just a way back machine rather than and serious version control. No fancy branching or issue tracking, just a backup for old code, in case I need to get back to it. When my laptop was stolen I doubled down on using GitHub for backup. I now have well over 50 repos, most of them horribly maintained spaghetti code with the last commit made a year ago. They’re not meant for collaboration.
I’ve made a few blunders while using git. I’ve published IP protected code publicly without a license. I’ve had ‘detached head’ errors and ‘merge conflicts’ and while fixing them I almost destroyed repos. But I’ve learned a few lessons from all this. Here’s what I’d tell my past self:
It takes less time to learn git than to do manual version control
Don’t be ashamed to publish your code. No one’s gonna read it anyway.
You can’t push and pull too often
When in doubt don’t just read a forum post and use the -f flag. Talk to an expert or think about it.
There’s no shame in using a GUI to interact with git. GitKraken is beautiful!
I’m still learning and I’m still making mistakes that seem ridiculous with hindsight. But I’m happy that they’re at least public.
Reflections on readings:
I connected to the message of inclusiveness in multiple readings. Whether it be by simply being polite to n00bs or going out of your way to include diverse groups often at cost to the project. I think more transparency and diversity leads to better projects. That said, the old adage ‘A committee should consist of three men, two of whom are absent.’ also springs to mind. We should be aware that inclusivity comes at the cost of overheads and slower decision making. We should not be quick to forget the fruits reaped by grace of the BDFL’s.
Personally, I will be more mindful and sensitive towards feelings of anonymous people on forums. And practice polite discourse or refrain from commenting.
I use my laptop mainly in the office so generally it would be connected to the charger all of the time. WIRED seems to think that this will kill my battery and that I should instead cycle between 80% and 40% capacity. Very well, I can’t change my habits but I sure can make stuff to have the same effect.
This project is a Bluetooth relay controlled by my laptop. I plug in the laptop charger into the output of the relay. A bash script monitors my laptop charge in the background and switches the relay on and off to maintain the mythical 80-40% charge level. I’m not fully sold on the idea of cycling the battery vs fully charging it. I’d love to hear from someone who knows their Li-ion battery chemistry.
I published a step-by-step tutorial for this, including the CAD files and all of the code. It got featured at Instructables.
Yupo Jelly is a transparent PET sheet with micro-suction cups on the back that stick to flat surfaces. Since it works using suction, it doesn’t leave any adhesive residue when removed.
It is typically used for putting up signage temporarily. It does not support printing via inkjet or laser printers however and needs one of the following techniques: UV offset, UV silkscreen and UV digital. It is fully recyclable.
Manufacturer Link : The manufacturer is easy to get in touch with and eager to work with artists. Samples are provided freely. Product is available for purchase from the manufacturer easily. Swatches available through application here. Distributors.
I scanned my hand using iPad structure sensor. The sensor had a lot of trouble scanning my hand separate from my body, so I set up a board with a hole to scan my hand as an object instead. Many thanks to Ridwan for helping me with the scan.
Mad props to Ridwan for the help with scanning.
A 3D model produced was processed and reduced in meshmixer. The resulting model is uploaded to sketchfab here. I want to try out wrapX for redefining the mesh.
I then made a shell around the model and split it in two. I had to be careful to split it such the two halves could fit on the model when 3d printed. The separation between the model and the shell is 2.5 mm(1/8″) this is so that I can use 2.5mm(1/8″) acrylic pieces as spacers to ensure the model is centered inside the shell then closed.
I printed the models using Cura and Ultimaker 2+. Since the model is relatively smooth, the layer height does not need to be fine. Even a 4mm layer height with 6mm nozzle is good. There were a LOT of failed prints.
Take care of the orientation of the build. It can have a huge effect on the time taken for the print.
The print needs to be cleaned up before use. I sanded the prints with a Dremel, going slow to prevent melting. Then I covered the entire surface with epoxy. The epoxy adheres to the rough surface well and allows the silicone to be removed easily.
I stuck discs of 2.5mm(1/8″) acrylic on the model, spaced regularly and test fitted the shell. Then I stuck neopixels and a flex sensor on the model. I used silicone wires to connect the sensors so that it sticks to the poured silicone easily.
I followed standard techniques to cast silicone using a vacuum chamber. I’d recommend this video for instructions.
What went wrong:
The silicone I used(EcoFlex-30) has a relatively low working time of ~20 minutes. So there wasn’t enough time to degas the silicone when pouring it into the thin cavity between the model and the shell. The bubbles got trapped in the silicone and ended up in the final solid cast. Also, the silicone is not very durable and tears easily. And the electronics are not flexible enough to allow the cast to be removed from the model without damage.
What I would do differently:
Model in a different way so that the casting is in a perpendicular direction. Having a shorter distance for the bubbles to travel will allow for easier casting.
Design the mold with air escape holes at the bottom
I would use silicone with longer working time.
Use shorter stretches of electronics separated by coiled silicone wire.
Fig wasps and figs trees live in a mutualism ecosystem where the fig tree provides nourishment to wasp eggs and the wasps pollinate fig flowers. This is a mutually beneficial arrangement. The lifecycle of a fig wasp is illustrated below.
So, female wasps pollinate the fig flowers by entering the fig with their pollen laden body. Parasitic wasps have co-evolved to get nourishment from the fig trees without pollinating the flowers. They do this by drilling into the fig from the outside with their long ovipositor.
From an engineering perspective, the ovipositor is a flexible, maneuverable hollow tube with sensing elements at the end. This mechanism has been studied by multiple researchers with recent findings experimentally studying the mechanisms of insertion. The tip of the ovipositor is enriched with zinc, presumably for hardening.
This problem can have interesting direct applications in any number of in vivo experiments e.g. in neuroscience for non-destructive probing of brain cells.
Cable mechanisms consist of a tension(/compression) bearing cable running through a compression(/tension) bearing channel. This mechanism can be used to transmit force over relatively long distances with light and flexible material. Notable examples include Bowden type 3D printer extruders, bicycle brakes, early car throttles etc.
While experimenting with sphincter type channels, I discovered that to get symmetric contraction of the mechanism it is important to apply force through both ends of the cable as shown below.
One end of the cable is anchored directly to the channel. Note the asymmetric contraction.
Both ends are actuated. Symmetric contraction.
I used this mechanism to make Cookie Monster, an evil cookie jar that doesn’t let you take cookies. If you reach in and grab a cookie, it mouth of the jar traps your hand and won’t let go until you drop the cookie back in the jar.
I want to make a robust version of it, keep a delicious cookie or amazing toy inside and exhibit it in a children’s museum. I can just imagine kids trying all sorts of ways to get at the toy and failing. Ahhh.. it’ll be fun to watch that! If you’re someone who can make this happen, I want to get in touch with you!
I also experimented with developing a lightweight upper body exoskeleton to help me do pull-ups. But I gave up on the idea when I realized that a robust mechanism would require at least some metal machining. If you want to develop the concept further, feel free to reach out. I’d be happy to share thoughts and experience.
Most of the effort in making a wearable electronics project seems to go into the mechanical sewing of conductive thread and placement of components. This does not leave enough room for exploration, say changing the placement of components or trying out different ones. This seems reminiscent of early rat-nest electronics prototyping before breadboards came along.
The aim of this project is to make a wearable jacket with pop-in connectors that act as a bridge between the clothing and Arduino pins. The jacket will essentially be a breakout board for Arduino pins on to the body. This way wearables can be prototyped quickly. To further ease the development, the jacket will have wireless connectivity, and power management built-in.
Some projects that can be easily built using this setup are:
Music synth touch controls and wearable speaker
Bike turn lights with touch controls
Personal space tracker
Hands-on multimeter/conductivity tester (I’m most excited about this one!
Gesture based remote control
The project has evolved over the past two months through the Wearables class at Tandon, NYU (DM-GY 9103-I). It started out as a wearable breadboard, with the aim to allow quick and dirty wearable prototyping. After much design and redesign, I realized that a robust and easy to use solution is not possible. So I altered the idea to be a wearable prototyping platform.
I wanted the jacket itself to be very robust against wear and tear, even with rough use. So I opted against sewing conductive thread. I researched a lot for a suitable conductive material, testing dozens of metal ribbons. I finally settled on steel ribbon from Adafruit. But the ribbon is prohibitively expensive at 25$ per meter. In true ITP fashion, I hit the junk shelf instead and realized that the steel sheathing for display cables is an ideal replacement. I stripped around 10 meters of the stuff and was on my way to fabrication.
I designed the PCBs in KiCAD and then manufactured them with an OtherMill. The jacket was purchased and the female snaps added by hammering in place. Steel mesh cables were added with insulated covers and a layer of lining was added for comfort.
I’m looking for collaborators to make projects with Sez. I’d love to hear from you if you’re interested!
As a first step to designing a business card, I made a map of adjectives that best describe me. Here is the mind map that resulted:
The common theme that is relevant to a business card is my self-identification as a maker. Since I have a broad set of skills in electronics, carpentry, design.. a title such as creative technologist seems too constrictive. So I decided early on not to include any title.
I looked through my desk drawer for business cards and found a few samples to draw inspiration from. I particularly liked the simplicity of ny sci business cards. The portrait orientation makes it stand out and the watermarks give a pleasant second read without being too distracting on the first read. Inspired by this, I decided to adopt a portrait orientation and watermarks suggesting my skills. I got a lot of icons for tools, electronics, art and code to use as watermarks from The Noun Project.
I experimented with a lot of styles and combinations on my name and handle (tinkrmind) on paper, before looking for fonts online. I decided to go with ‘Sawasdee’ for the contact information. It is a compact font, which renders will for the portrait orientation. The fact that it is open source is just a cherry on top. For my handle and name I decided to go with ‘Neon 80s’. It is a playful and bold font in contrast to ‘Sawasdee’.
I wanted to develop a logo for my handle as a part of this project. I wanted the logo to represent both art and tech. I made the following two logos. I finally decided to go with the cog and brain logo since I think that the wrench and paintbrush look too much like a construction company logo.
Since I generally rely heaving on grids, I took it as a challenge to work without a grid and go with intuition composing this project. I settled on a fairly simple design, with a symmetrial composition for the back.
Experimenting with changing color palette of an image:
The technique: I converted the image into 5 color palette using the imageMode tool in GIMP(~photoshop) and the used the alien filter to change the palette until I found something interesting.
I used these colors as an inspiration to get the palette below.
I made the following compositions with this palette in Processing – original image files and source code on github:
I really enjoyed changing the palette of the leaf image above, so I made a tool in p5.js to convert any image into my palette. (code on github and alpha editor) Here are some interesting compositions that resulted from playing around with this code:
You know how when you get your boarding pass the teller highlights the boarding time and the gate? The fact that they have to do it is a testament to the abysmal design of boarding passes.
Through this assignment, I got interested in the design history of boarding passes. Turns out, the machine-readable component of boarding passes is regulated by the International Air Transport Association (IATA). But, as far as I can tell, there is no regulation on the human-readable elements. So, the only reason that we have these awful boarding passes is that some engineer who was only interested in the boarding pass barcodes made up a template for boarding passes which has been followed out of inertia since. Here’s an example from the ‘Recommended BCBP document layout‘:
Why then are we still suffering from awful boarding passes like this:
Here is an attempt to redesign this monstrosity:
Font used: Bitstream sans
Side note: If you’re interested in what information is contained in the barcode on the boarding pass, it’s easily decoded with a simple app. It is PDF417 format data-matrix code and contains mostly the same information as typed on the pass. But it can contain a lot more information as well, including the sequence number and frequent flyer number. More details in the IATA implementation guide.
It is not a pretty sight when a drill bit rotating at 20,000 rpm gets caught on a work-piece. Even a light foam core model then seems like a deadly projectile, especially if it’s headed straight for your eye. This happened to me a few months back. What makes it worse is that I had my safety glasses right on my forehead! I had pulled them up to examine the part for a minute and forgot to put them back on. I want to make sure that this doesn’t happen again. Hence, PopEye.
PopEye is a safety glass add-on that detects when you take your glasses off and cuts power to the tool you’re using. It seems like a relatively straightforward idea, and for the most part, it is. One can use a Bluetooth low energy beacon placed on the glasses. The beacon is switched on whenever the glasses are put on. An accompanying power relay will go between the mains supply and the power tool. It will listen for the beacon packets and will switch off the power if it can’t detect the beacon. This entire setup can be completed quickly over a weekend. I explored IR and other wireless protocols as well, but the simplicity and extensibility of BLE make it the best choice. So, the major challenge is to detect when the glasses have been put on correctly.
Before proceeding much further, it is important to narrow down a bit and focus on a particular type of safety glasses. Safety glasses come in at least half a dozen different configurations. I chose to focus on over prescription glasses type, with an elastic band instead of stems. The elastic band will ensure direct contact with the face, making detection easier. I chose to go with 3M 91252-80024 glasses(image below). It is a cheap and effective choice. Although I will guide the design to have high dimensional tolerance so that it will work with any similar glasses.
Back to the problem at hand, detecting when the glasses have been put on. When I put on the glasses, right away I noticed a few points of detection. The forehead/brow, arch and cheekbone regions make a tight direct contact with the skin when the glasses are put on correctly. I would imagine, based on facial anatomy, the arch and cheekbone contacts can be interchanged for support at the bottom, but the forehead must necessarily make good contact with the glasses. More importantly, if I move the glasses up to my forehead, there isn’t a strong contact. So, we have a long (~10cm) strip which makes good contact only when the glasses are put on correctly.
To detect contact in this region, a light touch micro-switch easily comes to mind. It can be placed on any part of the glasses which come in contact with the skin only when the glasses have been put on correctly. But it can get false positives, e.g., if the glasses are packed inside a bag.
Using skin conductivity is the next best idea. Two conductive pads can be placed on the top of the glasses such that they contact the forehead when the glasses are put on properly (see image above). If the glasses are not put on properly, the circuit will be open. A microcontroller can then read the resistance between the pads and only switch on the BLE beacon when the resistance is in a sensible range, say 1K to 1M ohms. This will prevent accidental switching on if the glasses are packed in a box, even if placed along with metal tools -thus shorting the pads. It is unclear if a microcontroller is needed, a clever analog switch should be able to achieve this while saving cost and -more importantly- power.
Copper pads connected to probes of the multimeter, in 20M resistance measuring mode.
The multimeter shows overloaded resistance even though glasses are placed on properly.
Note that the left contact is in the air, making no contact with the skin.
After doing some experiments with this, I found that this method works most of the time. Of the six people I tested on, only one person had a facial structure that prevented the pads from making contact with the skin. (Image above.)
NFC can be used to quickly pair safety glasses to a tool. So, in a shared workshop, people can sign punch in and punch out tools by simply waving their safety glasses close to the power terminal for the tool.
The next steps for development are as follows:
Test different pad configurations: length and separation of pads, material: conductive cloth, pogo pins, copper tape, spring contacts..
Explore analog circuits for switching power to the BLE beacon.
Procure and test BLE and beacons, and RFID/NFC tags.
One of the guiding principles of good design is functionality. When pertaining to signage, one of the most important features of functionality is visibility. The following examples show examples of signage that are visible from afar.
The gate numbers at Port AUthority bus terminal are angled so as to be visible from afar.
The NYU flags are a common theme binding buildings in a scattered campus.
I love this example of wayfinding in Imaginarium a 3D printing facility in Mumbai, India. The different departments are color-coded and lines on the floor lead visitors directly to their destination. I love how the lines go all the way up the doors and end in markers that jut out of the door, marking the end of the line.
Signs at Port Authority bus terminal are very inconsistent in marking the ticketing locations. There are plenty of signs marking the gates, but the signs seem to include ticketing windows only occasionally. There is a distinct lack of directions to ticketing booths at the key decision points, of which there are plenty thanks to the boxy nature of the building. What’s worse, on the second floor of the south wing of the terminal, there is only a single sign marking that the ticketing booths are on the first floor. It is a poorly accessible sign right in front of the escalator going down.
The sign on the far right points to the ticketing booths beings to the left but the sign on the left does not mention ticketing at all.
At a key decision point, there is no mention of ticketing.
The ticketing booths are on the first floor, going down from the escalator. But the sign right next to the escalator makes no mention of the fact.
This is the only sign on the entire floor that suggests that the ticketing counters are on the floor below.
This problem can easily be fixed by including markers for ticketing in the existing signage, as in the image below.
If one wants to take a look at bad signage one does not need to go far from Port Authority it seems. Right in front of the ticketing booths, there are signs pointing to gates 1-34 along the length of a 150 meter long corridor. The gates are arranged in an orderly fashion on the lower floor, but on the upper floor, there is no mention of which escalator should be taken for which gate. I was led on a sprint through the terminal only to end up right under where I started after about a 250m dash.
Further notes on Port Authority signs: Sign design throughout the terminal is consistent and simple. They use Arial font and simple icons, most of which are available for free on the noun project. The color scheme of green for streets and subways, black on yellow for gates and ticketing, and yellow on black for restrooms, information, police etc. is consistent throughout the terminal. Except for the lack of signage at places and ill-advised placement of some signs, I really like this system.
“ITP space habitat” is a sound walk where you as a recruit get a tour of the habitat by a veteran astronaut. The space theme was inspired by the Applications class, and it allows us to work with some weird sound effects.
(Elevator door opens)
Hello! Welcome to the ITP space shuttle. I’m Gwen, your guide. Come with me…
(pause, walking, sound of cane and footsteps)
This is the Earth room. We are currently test driving simulations of different locations on Earth, for recruits to go in when they are homesick for our home planet.
First, we are prototyping a rainforest: (Sound of rain, rainforest animals) …
As well as a field in the countryside (Sound of crickets, running stream)…
And the ocean: (Ocean sounds)
We also are working on a big city: (Sound of New York City, people walking on sidewalks, cars beeping, etc.)
Next, I’ll show you where we are building the dormitories… Follow me.
(pause, walking, sound of cane and footsteps)
(Sounds of construction start happening in background)
This is where you and the other recruits will be sleeping. We are currently installing the oxygen tanks and testing the pressurization of the cabin.
Now we will go to the meeting room, where you will meet the other recruits…
(pause, walking, sound of cane and footsteps)
(Sounds of chatting)
This is where we will be holding the training.
I was once a recruit, too, you know. All of us senior members here were. Now I am a veteran due to an accident I had onboard the ship… (Sound of explosions) (People running)
Quick! Duck! There’s been an explosion in the dormitory!
(Sounds of commotion fade out…)
The gist of the article as I read it is: Some of the greatest art would not be possible without plagiarism. Copyright laws are there to “promote the progress of science and the useful arts” and not to protect the creator’s self-interests. Great art is by definition a gift and it cannot exist in as a commodity. It is therefore important to allow art to go into the public domain and to promote second use. While doing this it is vital for creators to acknowledge their sources and pay homage to their roots.
This article forced me to articulate some of my own thoughts on piracy and ownership. I believe that great art is measured not by any objective quality, but by the number of lives it has touched and affected. Star Wars is made great not by the directors and the film crew, but by the teenage kid who goes to Comic-Con dressed as a Wookie. Should that kid be sued for stealing copyrighted material, or monetarily compensated for advertising the franchise? If the franchise is made successful by the public, shouldn’t the public own it? In a world of digital rights management, when your entire book collection can be wiped by someone at Amazon; what does it even mean to own something?
“If I have seen further, it is by standing on the shoulders of giants.” the quote by Issac Newton may have been meant as a quip to Hooke, but that makes it no less valid. The greatest enterprise of humanity, The Scientific Revolution is built on free sharing of ideas. The ill conceived notion of equating creative property, with physical property -which can be stolen- is detrimental to progress. I firmly believe in free sharing and open-source. It is much better to allow everyone to work on the same, shared knowledge base than to stifle progress with artificial constraints.
The woman in the red jacket will now forever be a part of my memories of Central Park. Starting out, I was a bit skeptical about the value of the sound-walk. I am an avid podcast listener and I thought that sound alone can create a landscape in the mind. So, I listened to the sound-walk on my way to Central Park from New Jersey. The things I missed on the first listen and the experience of having a moving scenery as I listened were simply staggering! The timing of cricket noises just as I pass by a bush, the subtle reverb in sound along with architectural changes and the melding of noises to a point where I couldn’t really tell if there was someone behind me added up to an unforgettable experience. Even though I didn’t relate to the story very much I shall not forget the woman in the red jacket.