Earlier this 12 months, college students from throughout the world put their ardour, ingenuity, and dedication to work crafting Swift playground tasks for the WWDC20 Swift Student Challenge. Students from 60 totally different nations and areas introduced their abilities to the desk, utilizing the problem’s three-minute restrict to stretch their imaginations and discover Apple’s frameworks and applied sciences. The ensuing 350 Swift Student Challenge winners have created AR experiences, tasks powered by machine studying, instructional materials, digital musical devices, 8-bit video games, and a lot extra.
Swipe to unlock
In 2019, after having studied simply one month at the Apple Developer Academy in Porto Alegre, Brazil, Henrique Conte submitted his first Swift playground challenge. While he didn’t win a WWDC Scholarship that 12 months, the expertise drove him to proceed exploring and refining his code.
“This year I felt that, after reading and studying about so many different Apple frameworks, I should try to accomplish the WWDC19 main phrase: ‘Write code. Blow minds,’” he informed us. And he did. His profitable submission, a three-minute sport designed for MacE book Pro’s Touch Bar, excels in each its technical acuity and creativity.
Within the playground, gamers have to assist Eleanor, a younger developer, escape from a cave. The twist: The “cave” stage is solely positioned inside the Touch Bar. “I chose to use unusual frameworks to show that it is possible to do amazing things with them,” Conte stated. “I feel that [the Touch Bar] has so much potential yet to be discovered, and I wanted to demonstrate some of its capabilities.”
Though he had by no means beforehand developed for macOS, Conte instantly took to the problem. Like studying any new improvement idea, he bumped into a few preliminary obstacles — “when I received the message “No such module ‘UIKit’ [after trying to add it to my project] I noticed things would get interesting,” he joked — however he shortly picked up the fundamentals of Mac programming, utilizing AppPackage and SpriteKit to construct a absolutely-interactive expertise, together with faucets, slides, keyboard integrations, and multi-display screen storytelling. He additionally paid shut consideration to the design, one thing that has develop into more and more necessary to him.
“I am definitely not a designer,” says Conte, “But in the past few years I noticed how essential it is to follow Apple’s Human Interface Guidelines and to provide a great experience. There is no point in creating a perfect code if people will have trouble using your application!”
You can discover extra of Conte’s work on the App Store: In the final 12 months, he’s created 4 apps, together with one to assist youngsters with autism talk. He’s at the moment at work on his subsequent challenge, which he’s constructing for each iOS and macOS, that addresses the downside of meals waste. And — he’s completely satisfied to report — the macOS model will incorporate the Touch Bar.
The world turned upside-down
Louise Pieri, 21, fell in love with laptop science at a younger age in her native Lyons, France. She’s since gone on to review at École 42, the programming college based by French businessman Xavier Niel. Pieri’s profitable challenge, Meep, drew inspiration from an article she learn in a scientific journal about the risk of parallel universes, in addition to her personal private journey as a transgender girl.
“Meep is a game with two levels: the first is a level where everything is reversed and upside-down and the second is where everything is normal,” says Pieri. “The story is about a little blue transgender monster who wants to reach the final level and turn pink… it’s a beautiful metaphor for what happens in the life of a trans person.”
Though Pieri had by no means beforehand used SpriteKit or AVKit, she knew she wished to construct a sport for her Swift playground submission. She spent two days brainstorming concepts earlier than touchdown on the idea for Meep and diving into the frameworks and interface. Initially, she’d hoped to point out each of Meep’s universes on-display screen at as soon as earlier than deciding on a multi-stage expertise, together with a complete stage upside-down.
In addition to designing and coding the sport, Pieri additionally created her personal 2D paintings for every stage in Adobe Illustrator and a story for the sport — all in below two weeks.
Pieri has been tuning into WWDC20 this 12 months from France, and is very eager to be taught extra about the way forward for Apple platforms. “I can’t wait to get to know iOS 14,” she says. That will turn out to be useful for her subsequent challenge — bringing a model of Meep to the App Store.
A robotic of one’s personal
Devin Green’s love of improvement stemmed from a lifelong fascination with pondering machines. “I have always been in awe of artificial intelligence,” he informed us. Out of that concept, the 18-year-previous’s profitable challenge — an AI bot named Stanny — was born.
“With everything that is going on in the world right now, I thought people stuck in isolation might find it beneficial to their mental health to talk to a capable AI companion,” he stated. Green, who will attend Stanford this fall for laptop science and engineering, took about a week to construct his playground — most of that time devoted to refining machine studying fashions that created Stanny’s ‘intelligence.’
“The model was trained on a data file made up of all the things you could possibly say to Stanny,” Green stated. After researching how others had skilled chat bots, Green created his personal mannequin in TensorFlow, then introduced it to his Xcode playground by means of Core ML Converters.
While no stranger to experimenting with machine studying fashions, Green used this challenge to get to know extra of Apple’s ML choices. “I wanted to make it as simple as possible to go from data to usable Artificial Intelligence,” he informed us. He labored with NSLinguisticsTagger to construct a working mannequin, then created a generated Core ML mannequin to foretell the particular person’s intent from their question and architected his playground in SwiftUI.
Green sees Swift as the way forward for machine studying and AI functions. “Swift is not only a really simple and easy to use language, it’s also really expansive,” he informed us. “It can be built on in such a way that [it] is capable of doing just about anything.”
Stanny isn’t fairly as full-featured: His joke-loving AI is restricted to solely 63 totally different intents. But Green has huge plans — and he can’t wait to include a few of the expertise introduced at WWDC20. “The Natural Language framework is astounding,” he informed us. “If I had any doubts about using Swift for machine learning, they quickly disappeared while watching a natural language processing application understand text with about five lines of code… projects I’ve created, like Stanny, are about to get 100x better!”
Code as design
For first-time winner Renata Pôrto, the problem gave her a probability to confront her personal self-doubts. “As a designer, I have always felt insecurity regarding my ability to code more complex ideas,” she stated. After two unsuccessful Swift playground submissions in earlier years, the 21-year previous pupil from the Universidade Federal de Pernambuco in Recife, Brazil had as soon as once more determined to submit. But after a few days of labor, she scrapped the challenge.
“I was not satisfied with my own decision to continue with a ‘safe’ idea,” she informed us. Instead, she started contemplating ideas she’d wished to find out about however hadn’t but explored — together with generative artwork. “One of the things I always thought was fantastic about programming is the possibility of transforming lines of code into visual and interactive experiences,” she stated. Just six days later, she emerged with Polar Patterns, a Swift playground that helps folks be taught extra about mathematical roses and generate their very own visible artwork.
“With SpriteKit, I was able to convert the polar equation of mathematical roses into SKShapes, transforming the results from the formula into visual elements,” she stated. Pôrto designed simply two photos herself, counting on her algorithmic code and a few UIKit parts to create the total visible expertise.
Designing a wholly programmatic artwork interface was a departure for Pôrto. “I am very used to prototyping before programming,” she informed us. This challenge, nevertheless, concerned creating an expertise that might dynamically shift and alter relying on what actions somebody took inside the playground, making it necessary for Pôrto to visualise and experiment continuously to good her interface.
The consequence brings the magnificence and complexity of polar roses to a Swift playground — and landed Pôrto a profitable submission. “It’s a great joy learning about development, even though I am a design student,” she informed us. “Studying design helped me to practice my empathy for [people] and to know what resources to use and how to work with them to achieve better results.”
That empathy continues by means of her work with a native developer group, creating instructional instruments for brand spanking new builders in her group. “I always try to pass on my design knowledge to developers, and my developer knowledge to designers,” she stated. “And hopefully one day they will make products that make a difference in other people’s lives.”
Learn extra about the Swift Student Challenge winners.