Map My Crop

That’s a short warning to all Tensorflow users working with visual content. Short notice: don’t use any functions! tf.image.resize I was rewriting codebase of our — Let’s Enhance to make it ready for bigger and faster models and API we are working on. As we work with image generation (superresolution, deblurring, etc) we do rely on a typical image processing libraries like OpenCV and PIL. I always had suspicions that it makes sense to use Tensorflow image processing capabilities — in theory, they should be faster. So, I decided to stick to native Tensorflow image preprosessing and dataset building tools using dataset.map() to keep all operations in tensors all around my code. neural network image upscaling service The problem was quite awful — my new and shiny code for superresolution wasn’t able to reproduce not only any state-of-art networks but even my own code that I wrote 4 months ago. And the ugliest part was that results of superresolution itself were pretty good sometimes, the network was working, although not reaching target PSNR and having strange visual artifacts sometimes, like doubling of small lines. Let the debug begin What initially was looking as minor bug became 60 days of struggle and sleepless nights. My faulty logic was simple — there’s something wrong with network definition or training process. Data preprocessing is definitely fine, as I am having meaningful results and visual control over image processing in Tensorboard. I tweaked everything I was able to find, defined network using Keras, Slim, raw TF — nothing, looked for changes in TF 1.3->1.4->1.5 and different CUDA versions, paddings behaviors. I am ashamed even to tell you about my latest suspicions, which involved defects in GPU RAM and statics. I was tweaking perceptual losses and style losses looking for a reason. And each iteration took days to retrain until some meaningful result… Yesterday I found when looking into Tensorboard. It was almost subliminal feeling that something is wrong with the image. I disregarded network output and just overlayed target image and input image(that is a downscaled target image) in Photoshop. Here’s what I got. The Bug, Looks strange, somekind of displacement happening here. Totally againts any logic, this just can’t be true! My code is dead simple. Read image, crop image, resize image. All in Tensorflow. Anyway, RTFM. tf.image.resize_bicubic has a parameter — “align corners”. How in the world would you like to downscale image and not have corners aligned? You can! So there is a very weird behavior of this function known for a long time — . They can’t fix it as this can break lots of old code and pre-trained networks. read this thread Our function isn't even reflection equivariant. It would be lovely to fix this, but I'd be worried about breaking old models. tf.image.resize_area This code does actually displace your image by one pixel to the left and top. Thread suggests that even the interpolation is broken in TensorFlow. It’s 2018, people. Here are the actual downscaling results with TF. Stick to Scipy/OpenCV/numpy/PIL whatever you prefer for image processing. The second I changed it my network worked like a charm (next day actually, when I saw training results).

How Tensorflow’s tf.image.resize stole 60 days of my life

BROMSTEAD AND MY FATHER

The Two Shark-Faced Creatures Were Dragging at my Arms and Legs

Las dos criaturas con cara de tiburón arrastraban mis brazos y piernas

Les deux créatures au visage de requin me traînaient les bras et les jambes

दो शार्क-मुंह वाले जीव मेरी बाहों और पैरों पर खींच रहे थे

サメの顔をした 2 匹の生き物が私の腕と脚を引きずっていた

As duas criaturas com cara de tubarão estavam arrastando meus braços e pernas

Hai sinh vật mặt cá mập đang kéo cánh tay và chân của tôi

两个鲨鱼脸生物拖着我的胳膊和腿

Teaching my old iPhone new tricks

Here is some advice for anyone studying abroad. Just do it. It is true when they say that your college years fly by. Here I am in my second semester of my junior year and I can’t seem to comprehend the fact that I am supposed to have a career plan in a year from now. Hell, I’m still waiting for my letter from Hogwarts. Although there’s one thing I do know and it’s that I’m so glad I took the opportunity to travel abroad last June of 2015. When I look back to my college years in the future, my intent is that I never have to think about the “ what if’s”. Last summer, I was blessed with the opportunity to visit a little town in the south of France called Pau. You’ve probably never heard of it, but if you have, let me know because I want to meet you. Pau is surrounded by the breathtaking Pyrénées Mountains that also extends to the border of Spain. Pau also has it’s own , along with other rich historical architecture around the city. I experienced all kinds of bipolar weather, but the most being rain. : I can proudly say that I had one foot in Spain and the other in France when my host family and I drove up to the mountains to visit a small market place. château Fun fact Anyways, I could really go on and on about my experience abroad but I wanted to focus on a few things here. If you are a college student like me, and the thought of studying abroad crossed your mind even for a second, I strongly recommend it and here’s why: Say good-bye to your comfort zone You may be used to having the same routine every day that keeps your life on schedule. Living abroad, especially in Europe, time doesn’t exist. Looking back, I realized how much Americans are obsessed with time. In France, if someone invites you to dinner at 9 p.m., you better not show up until 10 p.m. Have an appointment after class? What’s the rush, let’s just go grab an afternoon coffee and talk about nonsense. Taking the train? Yea…you might want to be on time for that. La langue d’amour I put myself in a homestay to force myself to learn French in the quickest, most natural way possible. I kid you not, all those five years of French I took in high school and college, I relearned in three weeks. By five weeks, I was able to hold a conversation with my host parents at dinner about what I did during the day and what I’m going to do the next day. If you know French, you know that speaking in the past is not easy because you have to think about the or . Did it happen once? Or did was it something I used to do? There’s your French lesson of the day. l’imparfait le passé composé Also, what could be better than home-cooked French cuisine every other night? Absolutely nothing! Coming from a Cape Verdean background, I felt right at home. For others, one might experience something called culture shock as defined below: The feeling of disorientation experienced by someone who is suddenly subjected to an unfamiliar culture, way of life, or set of attitudes. To overcome this horrid curse, we must become open-minded to new experiences, ideas, and cultures and be willing to try new things. So, I’ll start with something we all can’t live without: Food. If you are a picky eater and you don’t like to try new things, then you will definitely struggle eating abroad. I ate a variety of food in France ranging from Vietnamese to Lebanese to Moroccan to Mexican (granted I ordered a cheeseburger and fries, but it was the best damn cheeseburger and fries I’ve ever had, and I was feeling slightly homesick) and even Italian. Oh, and of course I had to have sushi a few times, and let me be the first to tell you that there is no all-you-can-eat :’( Paella from San Sébastien, Spain. My point is that everything food-wise is just better. You really don’t have to worry about GMOs and processed food because everything is already fresh and organic. My host parents bought their fruit, veggies, meat, fish, and cheese from the local farmer’s market every Saturday. I am completely surprised that I did not gain 100 lbs because I ate bread and cheese at almost every meal and dessert after every dinner. I ate at Georgio’s, an ice cream truck, about four times a week. I also had a box of macarons with me at all times. Even when I went out to restaurants, we ordered an abundant amount of food because we wanted to try everything. Also, the fact that I could order a £6 ($6.67) wine bottle for myself at dinner was game over for me. Les amis If you are wondering what this means, it simply translates to “friends”. Again, if you don’t like to meet new people, I suggest skipping this part. But if not, I invite you to stay and read about why I think making friends abroad is important. I met some of the nicest people in my study abroad program and even the friendliest locals. : Want to know another fast way to learn a language? Date a local! The whole “French people hate Americans” is just a stereotype. What is true, however, is that the French think we are loud — that I have to admit, is true. We can all recognize that cultures have their differences, but for the most part acting as French as you can without going overboard will definitely get you in with the locals. Another tip Almost every night you could find us at The Galway. Any opportunity I was given to either see a movie, go out to the bars, shop at , or just to sit down with a friend and drink some , I wouldn’t miss. Every day is a new adventure being abroad, and it is even better when you have friends by your side willing to explore with you. I went to Spain with a Spanish speaking friend from my university and she taught me how to order a drink in Spanish, which still is the only phrase I remember how to say. La Place Clémenceau café au lait Becoming friends with the locals will really make you feel apart of the culture. I went out and played pick up soccer as often as I could and picked up (no pun intended) on street soccer terms like “ and “ ” to call for the ball. For whatever reason there is, making friends with the locals is beneficial to learning the slang and non-formal part of a language. vas-y, vas-y!” ouais Ma famille d’accueil I couldn’t have asked for better host parents. They picked me up and dropped me off whenever I needed it. They fed me breakfast, lunch, and dinner whenever I was home. They threw me a mini birthday party with the best homemade tiramisu cake I’ve ever eaten in my life. They bought me a dozen of my favorite macarons. They drove me to their favorite spot in the mountains. They gave me a local cell-phone with my own French phone number so I could contact them and my local friends. They brought me to their garden to show me all the crops they grew. They gave me sweaters to wear when I was freezing because I only brought tank-tops and T-shirts. And they even asked me how I spent my day every single night. It really couldn’t have gotten any better than that. I miss them dearly and I try to stay in touch every once in awhile via email. Aren’t they just the cutest? The most frequently asked question I get is about being in a homestay is awkwardness. Here’s my answer: it’s only awkward if you make it awkward. My host parents encouraged me not to stay in my room when I was home and to talk to them about anything so I could practice my French. I hope that you will have a family to make you feel as comfortable as mine did to me. If you are thinking of being in a homestay, I strongly encourage it because I built a connection so great to the point where I had tears in my eyes when I had to say good-bye for the very last time. Transportation Traveling within another country does sound terrifying in itself. However, the only way to learn your way around is to become familiar with the area by actually experiencing the city. People don’t want to look like tourists carrying around a map, but sometimes it’s completely necessary. I had to take the bus from my homestay to school every day and it only took me two days to understand the bus routes. Granted that Pau is a small city, but the only way I learned was by making a mistake. Story time. The first day I took the bus home from school, I got off one stop too early and I ended up walking around downtown for a good, solid hour. I was panicking in my head but I was trying not show it on my face. I was too proud and inexperienced in the language to ask for directions, so I just tried to figure it out on my own. I also was noticeably holding open a map like a full blown tourist…but no one stopped to help me. The good thing was that this happened to every one else in my program. We all at some point walked around aimlessly trying to find our homes. Eventually I made it home in one piece. I’m not exactly sure how, but I made it a point to never get off on the wrong stop again. Here is my lesson to you: familiarize yourself before you blindly hop on a bus and think you know where you’re going — you don’t. For traveling short distances outside of the country, I recommend bussing it as often as you can. It may take longer, but it is way cheaper than a train or an airplane ticket. I had learned about “Bla Bla Car” from my host parents, so if it’s coming from them, then it must be legit...right? This is also a cheap way to travel and is very similar to über. All you do is look up people who might be driving to a certain direction you need to go, hop in their car and enjoy the ride. This method is best for traveling to popular cities. Classes Fortunately, I was only taking two classes during the summer. Learning a language isn’t easy, but engrossing yourself in the culture will make it 10x easier. I almost forgot I was in Pau for summer school because it felt like a 5-week vacation in Europe. All I can say is as long as you do the homework, stay engaged in class, and show up, classes will be a breeze. Also, do not be afraid to speak the language. Fake it ’til you make it. You are only as immersed as you allow yourself to be. Go out and explore. Say yes to coffee and crêpes. Be prepared to get little to no sleep and to walk . Take a road trip to a neighborhood country. Visit that museum you walk by every day. Snapchat all your meals. Go wine-tasting and then buy your favorite bottle. Try that Lebanese restaurant every one keeps talking about. Keep a journal to document your favorite memories. Take pictures of everything, they probably already know you’re American when you walked through the door anyways. Lastly, be safe and have fun. a lot Traveling is truly exhilarating and reverse homesickness is real my friends. One last thing for your flight back to the U.S….yes, that’s correct. You might be thinking, “ ” but you never know what could happen. Check and double check your flight information and make sure you have a reliable mode of transportation. Do not depend on the front desk clerk to tell you that the shuttle only takes an hour to get to the airport when it really takes two and a half hours because there’s loads of traffic and then in the middle of the shuttle ride, you realize that you read your itinerary wrong and have to admit to yourself that you’re probably giong to miss your flight. Then, when you show up to the airport, you only have five minutes to board the plane…but wait! You still have to check in your overweight luggage and wait by the check-in services because TSA shut down the gates due to a bomb threat and then you have to call your mom at in tears. Please, just don’t do that. : arrive at the airport at least when traveling internationally and you will avoid all of these issues. Be on time well, duh… 1 in the morning her time Moral of the story here four hours early To conclude, here is a quote from my dear friend and mentor, Barbara Lane: “Travel enriches us…not only seeing beautiful places, but in the experiences we have. Our world views from language and food, from culture, is deepened. We learn about ourselves, our strengths and weaknesses.” Tu me manques, Pau ❤

Why Studying Abroad Changing My Life

Any views expressed in the below are the personal views of the author and should not form the basis for making investment decisions, nor be construed as a recommendation or advice to engage in investment transactions. We are all speculators. Every instant we exist in this universe is filled with uncertainty. In our attempt to wade through an unpredictable existence, our brain constantly constructs a probabilistic map of our environment. Our actions are based not on facts but on perceived probabilities across various outcomes. A simple example that applies to my existence is the risk of triggering an avalanche while skiing. The most enjoyable powder runs in the backcountry are located on pitches 35° to 40°. This is also the perfect gradient for an avalanche. Before dropping in, my guide assesses the probability of an avalanche, given the observed snow and weather conditions. My guide also relies on recent observations of other guides who have skied in the same area. If the risk is too high, we don’t ski. A more quotidian example is whether to take the elevator or the stairs. The former is faster than the latter. However, elevators are mechanical devices that sometimes fail, and failure could lead to severe injury or death. You assess the expected value (probability * outcome), whether you are conscious of it or not, of the time and effort saved by taking the elevator up 30 floors, given your belief about the chance of injury or death, vs. walking up stairs which is a less risky mode of travel but takes far longer and is more tiring. You are gambling with your life every second of every day. It’s not a bad thing; it just is the essence of humanity’s inability to perfectly forecast the future. And what an awful existence that would be if we knew precisely how the future would play out. I prefer our imperfection. The story you tell yourself about certain actions informs your perception of their risk. I will call this the narrative. For social beings like humans, the narrative is primarily created through the “wisdom” of the crowd. For better or worse, the most powerful stories are the ones in which everyone believes. The narrative is also created by objective facts. The facts, in most cases, are discrete events that point to the riskiness of certain behaviors. It is a fact that avalanches are more prevalent on a 40° pitch. It is a fact that people have been injured or killed while riding an elevator. Common chatter and objective facts combine to create the narrative. While facts are great, it is challenging as an individual human to know the exact number of fatalities on an elevator over the total number of rides ever taken. It is also challenging to know the number of skiers who died in avalanches on a 40° pitch over the total number of runs taken on a similar slope. In the absence of our ability to ascertain the exact actuarial data, we rely on others. It goes something like this. Avalanches: I know avalanches in this type of aspect are more prevalent. But my guide, who has extensive experience and training on how to determine which slopes are avalanche-prone, believes this particular line is safe. Safe doesn’t mean there won’t be an avalanche; safe implies the probability of an avalanche is low enough to be acceptable. Therefore, due to my trust in his training system, which evolved from the experience of thousands of mountaineering guides over time, I will follow him down this slope. Elevators: I know that riding the elevator is more dangerous than taking the stairs. But everyone else is taking the elevator. If everyone else is taking the elevator, then it must be safe. Everyone can’t be wrong all at the same time. Additionally, there are building codes that were created using the experience of trained engineers, and this elevator is certified as safe. Therefore, trusting the expertise of engineers I will never meet and the crowd's wisdom, I feel safe riding the elevator. The way we assign probabilities doesn’t hinge on the facts or the technology but on our perception of the facts and how good the technology is. These perceptions are based on what other humans, who we assume know more than us due to their training and experience, say are the facts or say is good technology. Crypto: To tie this into crypto, consider the following. Suppose a new project claims to solve a problem using novel technology. The problem they claim to solve is well-known, and the tokens of other projects attempting to solve this problem are highly valued. You believe the project’s engineers are smart and talented enough to solve this problem. You believe this because other engineers who have launched successful crypto projects are advising them. You are also confident in the team because they have graduated from well-regarded tech-focused universities and have prior work experience at successful tech companies. Because the narrative is strong (story + tech), you invest. But digging deeper into your thought process, what is more important: the story or the tech? The Story. The story is more important than the tech. Your perceived probability of success is based on what other people say about the problem and what other people say about the team's technical competence. In very few cases do you have the ability to evaluate the tech at a fundamental level solely. That’s why you trust others you believe are better informed than yourself to signal whether the tech is highly likely to solve the problem. While your tech skills are usually inadequate to properly evaluate the project, you can easily understand whether a story is good or not. A good story is one that more and more people tell each other. Of course it’s great if the story is spoken about in a positive manner. For example, “All retail traders are going to switch from CEXs to DEXs in this cycle.” But even if the story is spoken about negatively: “There is no way that retail traders are going to leave CEXs for DEXs.”, the story of the migration of trading volume from CEX to DEX is still being spread. I don’t care if people believe the story; I just want them to tell a positive or negative variant of it. Because you make more money being long than short, optimism will win over pessimism during the cycle. It’s just how the human brain is wired. What Is My Job? While my official title is Chief Investment Officer at Maelstrom, I should alter it to Chief Story Officer. I tell stories. The better and more concise the story, the faster it will spread. The greater the story's virality, the more tokens adjacent to that story will appreciate in value. The tech does not matter! The financial professionals at Maelstrom all graduated with finance degrees from the Wharton School of Business at the University of Pennsylvania. I didn’t plan it that way; it just happened. While we understand the potential applications of cryptographic and blockchain technology, we are not cryptographers, distributed network specialists, or have deep technical computer science knowledge. When we do deals, we outsource tech due diligence to others with those skills. The others might be the lead VC  firm or qualified angels in a pre-sale token round. Some might be the project’s highly respected technical advisors. In the absence of these types of validators, we might get comfortable with the tech because the founders have launched successful projects, and by successful, I mean applications that were used by many crypto projects in the past. Our job is to ascertain which project has the best probability of success given the story vertical. Success depends on a macro and micro-story spreading wide and far. You make the most money on a token attached to a story to go from being perceived as “never” to “maybe” happening. I would rather invest in a token with a perceived probability of success of 0.01% that has a story in the growth phase of virality than a token with a perceived probability of success of 50% but has a story that has reached the common knowledge phase. If the probability of success rises from 0.01% to just 1% because the story quickly infects the minds of many, I 100x my money. But the only way to 100x my money with a token that has a perceived probability of success at 50% is that the actual results, in whatever format is relevant to that project, are so amazing that growth comes from observed results rather than a rising perception of future success. The macro story speaks to an observed trend and how this project will capitalize on it. It’s a story more than a trend because we are taking a small movement and extrapolating its impact much further into an uncertain future. Macro Story: “Retail derivatives trading volume is shifting from CEXs to DEXs.” BitPerp is building a perpetual (perp) swap DEX. BitPerp’s token will pump because its macro story is currently not well known but has the viral potential to infect many minds. The micro story speaks to why this particular project will be the best in class out of all those competing within a particular macro story. Micro Story: “BitPerp is advised by Arthur Hayes, who helped invent the perpetual swap.” When others hear that Arthur is involved, they assume the project will get some kick-ass advice to help them best all other competing projects. This blog is usually about macro stories. Most of the time, I’m telling stories about thieving central bankers and politicians who are destroying the value of time and human labor by printing fiat money. I’m telling the story of how Bitcoin and the crypto ecosystem are an antidote to this organized theft of human dignity. But given I run a trading book, I also tell micro-stories about trends within crypto and how my chosen coins and tokens will rise in price as more people believe and hear about the story in question. I don’t write in-depth micro-stories that often about specific tokens outside of Bitcoin and Ethereum. But, it’s bull market time bitches. I laid the foundation of the significant forces that will drive crypto usage and adoption; it’s time to shill my bags. Results Do results, and by that, I mean growth in trading volume, total value locked, number of unique wallets, etc., even matter? Yes, they matter, but their importance to the price of a token differs depending on what part of the hype life cycle in which you are investing. When investing in a story/trend you believe will go from “never” to “maybe” happening, the importance of a project’s traction is low. The market doesn’t expect much because the market believes this token is attached to a trend that is unlikely to grow in the future. Therefore, even mediocre results are trumped as groundbreaking because the expectations are so low. When investing in a story/trend you believe will go from “maybe” to “definitely” happening, the importance of a project’s traction is high. The market’s expectations are high because they believe in a bright future. What would be considered mind-blowing results in the previous phase are considered mediocre in this phase. Amazing results are not enough; in this phase, a project must be genuinely revolutionary in order to meet expectations. Shitcoin Story Time The point of this essay and thought exercise is to provide readers with a view into the conceptual framework that guides Maelstrom. Over the coming months, most of the essays I write will focus on specific tokens we hold and their macro and micro-stories. These tokens have launched or have upcoming public launches, and thus, I’m trying to spread the story wide and far. I do not care if you buy or sell any tokens spoken about. I care that I present such a provocative story and supporting arguments that you discuss it in a positive or negative manner with others. I know I have succeeded when I read the following on social media:\ “Did you read Arthur’s latest essay? Man, that guy is a fucking dumbass. There is no way that Pendle will be the biggest derivatives trading platform in crypto, surpassing Binance. I don’t even know what an interest rate swap is, nor do any other degens in crypto.” Or “Did you read Arthur’s latest essay? Fuck me, none of us own enough of Ethena. Tether is going down, and Ethena is definitely going to be the top USD-pegged stablecoin.” Below is a rough idea of the macro and micro-stories I intend to tell over the coming months. Retail derivatives trading volume will shift from CEXs to DEXs. Projects in question: dYdX, GMX, and possibly another challenger The launch of ETH staking will spark a surge in interest rate swap trading volumes across DeFi Project in question: Pendle There is a way to use tens of billions of dollars worth of low market cap shitcoins to power DEX quanto derivatives trading volumes. Project in question: Krav DEX on-chain liquidity will be provided by middleware that disintermediates the current crop of market making firms. Project in question: Elixir As DEXs become the primary venue for price discovery, on-chain oracles that provide prices for settlement and liquidation will surge in importance. Project in question: Flare Why Tether and any stablecoin uses a TradFi bank for custody of fiat will face pressure and we can create fiat pegged stablecoins without relying on TradFi. Project in question: Ethena How to solve cross-chain bridging of assets without building a bridge. Project in question: Axelar Harvest Right now, the energy and attention is on the astonishing amount of Bitcoin the US-listed spot ETFs are accumulating. This, along with a global fiat debasement orgy, will drive Bitcoin to unfathomable heights in fiat currency terms. And the upcoming US-listed Ether ETF will drive Ether prices higher as well. I’ve got my Bitcoin and Ether. I might buy a bit more, but by and large my focus is shifting to shitcoins. Are there tokens that I can purchase that will outperform Bitcoin and, secondarily, Ether? This is Maelstrom’s hurdle rate. We accomplish this by becoming as knowledgeable as possible about certain projects and telling amazing stories.

Chief Story Officer

প্রধান গল্প অফিসার

Director de historia

Chef de l'histoire

मुख्य कहानी अधिकारी

チーフストーリーオフィサー

최고 스토리 책임자

Diretor de História

Главный директор по истории

Baş Hikaye Görevlisi

Giám đốc câu chuyện

首席故事官

Many years ago, I first wrote up working with the . I find myself returning to it again and again, and this weekend, I built a fun little game I think you may enjoy. It's called "Guess the Decade". my experience Marvel API Marvel's art style has changed over its long history. Back in 2018, I shared a that demonstrates just how much variety you can get just by looking at covers. So for example, Spider-Man in 1962: drastically demo Versus 1988: And then 2018: Given that there's such a variety of styles, I thought it would be fun to build a demo. If you want, you can just right to the , but here's how I built it. game The Backend For the backend, I built a serverless function with . I've really been enjoying playing with them lately (see my ), and I figured this would be the quickest way to get the API running. Cloudflare Workers earlier posts Let me share the code, and then I'll explain it. const IMAGE_NOT_AVAIL = "http://i.annihil.us/u/prod/marvel/i/mg/b/40/image_not_available";

const getRandomInt = function(min, max) {
    return Math.floor(Math.random() * (max - min + 1)) + min;
}

async function getCover(pubKey,priKey) {

    //first select a random year
    let year = getRandomInt(1950, new Date().getFullYear()-1);
    //then a month
    let month = getRandomInt(1,12);

    let monthStr = month<10?"0"+month:month;
    //lame logic for end of month
    let eom = month==2?28:30;
    let beginDateStr = year + "-" + monthStr + "-01";
    let endDateStr = year + "-" + monthStr + "-" + eom;
    let url = "http://gateway.marvel.com/v1/public/comics?limit=100&format=comic&formatType=comic&dateRange="+beginDateStr+"%2C"+endDateStr+"&apikey="+pubKey;

    // add hash
    let ts = new Date().getTime();
    let myText = new TextEncoder().encode(ts + priKey + pubKey);

    let hash = await crypto.subtle.digest({
        name:'MD5'
    }, myText);

    const hexString = [...new Uint8Array(hash)]
      .map(b => b.toString(16).padStart(2, '0'))
      .join('');

    url += '&hash='+encodeURIComponent(hexString)+'&ts='+ts;

    let result = await fetch(url);
    let data = (await result.json()).data;

    if(!data.results) {
        throw('No results available.');
    }

    let resultData = data.results;

    console.log('initial set',resultData.length);
    // if(comic.thumbnail && comic.thumbnail.path != IMAGE_NOT_AVAIL) {
    let comics = resultData.filter(c => {
      return c.thumbnail && c.thumbnail.path !== IMAGE_NOT_AVAIL;
    });
    console.log('now we have ',comics.length);
    let selectedComic = comics[getRandomInt(0, comics.length-1)];
    //console.log(JSON.stringify(selectedComic,null,'\t'));
    //rewrite simpler
    let image = {};
    image.title = selectedComic.title;
    for(let x=0; x<selectedComic.dates.length;x++) {
      if(selectedComic.dates[x].type === 'onsaleDate') {
        image.date = new Date(selectedComic.dates[x].date);
        //rewrite nicer
        image.date = `${image.date.getMonth()+1}/${image.date.getFullYear()}`;
      }
    }

    image.url = selectedComic.thumbnail.path + "." + selectedComic.thumbnail.extension;
    if(selectedComic.urls.length) {
      for(let x=0; x<selectedComic.urls.length; x++) {
        if(selectedComic.urls[x].type === "detail") {
          image.link = selectedComic.urls[x].url;
        }
      }
    }

    return image;
}

export default {
    async fetch(request, env, ctx) {
        const PRIVATE_KEY = env.MARVEL_PRIVATE_KEY;
        const PUBLIC_KEY = env.MARVEL_PUBLIC_KEY;
        let cover = await getCover(PUBLIC_KEY, PRIVATE_KEY);

        return new Response(JSON.stringify(cover), {
            headers: {
                'Content-Type':'application/json;charset=UTF-8',
                'Access-Control-Allow-Origin':'*',
                'Access-Control-Allow-Methods':'GET'
            }
        });

    },
}; Starting at the bottom, the method there will be fired when a request comes in for the function. It uses two credentials I've set up as secrets in Cloudflare and then calls the main function. In the response, I return the result and use CORS headers as my demo lives on another domain at CodePen. fetch getCover is code I've shown before, it's what drives my Mastodon bot, but the idea is to basically select a random year, random month, and then ask for 100 comics from that period. I filter out for comics that don't have a cover and then return a random selection. I do a bit of manipulation on the result to return a very small part of the comic. getCover randomcomicbook You can see this endpoint yourself here: https://randomcover.raymondcamden.workers.dev I want to note two things. In my testing, I'm seeing the function run really slow sometimes. I'm certain that's not Cloudflare but the Marvel API. I need to confirm this with some testing, though. And in maybe one out of like twenty or so calls, I'm getting an error with the results, which is also something I'd blame the Marvel API for, but I don't handle it well in my code. All in all, I'll just say - expect imperfection. And speaking of that, here's an example result from the call above: {
    "title": "Conan the Barbarian (1970) #54",
    "date": "9/1975",
    "url": "http://i.annihil.us/u/prod/marvel/i/mg/9/70/646cc9d27a80a.jpg",
    "link": "http://marvel.com/comics/issue/72071/conan_the_barbarian_1970_54?utm_campaign=apiRef&utm_source=fe877c0bf61f995fc8540d9eac4704f1"
} Notice how the title includes a date that doesn't match the date the Marvel API returned? In this case, it's the same decade, but in many examples, it will be the next one. I'll talk about how I handle that in a bit. Here's an example of that issue: {
    "title": "Doctor Strange (1974) #66",
    "date": "8/1984",
    "url": "http://i.annihil.us/u/prod/marvel/i/mg/3/d0/621679f213e8d.jpg",
    "link": "http://marvel.com/comics/issue/20149/doctor_strange_1974_66?utm_campaign=apiRef&utm_source=fe877c0bf61f995fc8540d9eac4704f1"
} Now, let's turn our attention to the front end. The Web Game For the front end, I made use of my favorite framework, . My game has three states. The initial state simply tells the player what's going on. Alpine.js After you click the button, I get an image and render out buttons for guessing. I also show your current stats. When you click, I will let you know how you did: Let's start with the HTML first: <script defer src="https://cdn.jsdelivr.net/npm/alpinejs@3.x.x/dist/cdn.min.js"></script>

<div x-data="app" x-cloak>
    <div x-show="intro">
        <p>
            Welcome to the Marvel Guess the Decade game. In this game, I'll show you a random comic cover selected anywhere from 1950 to the 2020s. You need to guess the decade the comic was released. 
        </p>
        <button @click="start">Get Started</button>
    </div>
    <div x-show="!intro">
        
        <img x-show="imageSrc" :src="imageSrc" class="cover">
        <div x-show="!imageSrc">
            <p><i>Loading cover...</i></p>
        </div>

        <div x-show="guessReady" class="buttonRow">
            
            <template x-for="i in 8">
                <button @click="guess(1950+((i-1)*10))"><span x-text="1950+((i-1)*10)"></span>s</button>
            </template>
        </div>
        
        <div x-show="resultReady">
            <p>
            <strong x-text="msg"></strong> This cover is from <span x-text="cover.title"></span> in <span x-text="cover.year"></span>.
            </p>
            <p>
                <button @click="getImage">Guess Again</button>
            </p>
        </div>
        <p>
            You have currently gotten <span x-text="right"></span> correct and <span x-text="wrong"></span> wrong.
        </p>
    </div>
</div> For the most part, this is pretty simple. I think the only 'fun' part is how I handle decades. I simply loop eight times from 1950. Obviously, I could make this work for any date in the future, but I was being a bit lazy. Now let's turn to the JavaScript. This time, I'll break it down a bit. First, I specify my variables with default states: intro:true,
right: 0,
wrong: 0,
imageSrc:null,
guessReady: false,
resultReady: false,
cover: { title:null, year:null },
msg:'', I'm not a fan of how I defined , but I got warnings in Alpine when it parsed the HTML. Now, I knew it wasn't a "real" issue because that HTML wouldn't be displayed till I had a cover, but I wanted the warning to go away. Next time, I'll use two variables, like and . cover coverTitle coverYear The method referred to from the button in the initial state just switches us to the main state and gets the first image. start start() {
    this.intro = !this.intro;
    this.getImage();
}, Now, let's look at , as there's an interesting bit in here: getImage async getImage() {
    this.imageSrc = null;
    this.guessReady = false;
    this.resultReady = false;
    let coverReq = await fetch('https://randomcover.raymondcamden.workers.dev');
    this.cover = await coverReq.json();
    // add .year based on date
    this.cover.year = this.cover.date.split('/').pop();
    /*
    So, I kept seeing comics with WILDLY different years and figured out we could use regex.
    I'm keeping the code above, just in case, but this code here should help.
    */
    let found = this.cover.title.match(/$([1-2][0-9]{3})$ #/);
    if(found) {
        console.log(`changing year to ${found[1]}`);
        this.cover.year = found[1];
    }
    this.imageSrc = this.cover.url;
    this.guessReady = true;
    console.log(this.cover);
}, So, as I mentioned earlier, the Marvel API will sometimes return a comic and say it's from 1989, let's say, but with a 'date' field of 1991. Initially, I just assumed I'd have to live with it, but I then recognized that the comic titles all followed a pattern: Uncanny Inhumans (2015) #8
Journey Into Mystery (1952) #31
Strange Tales (1951) #110
Thor (1966) #430 Given that, I figured I could write a regex to get the year, and amazingly, I got it right on the first try! The final bit is the actual "guess" logic: guess(decade) {
    // so decade will come in as a base 4d year, like 2000, 
    // my year from the API is m/yyyy. so we can easily get y, 
    // and then simply check the first 3 digits
    console.log('guessing',decade);
    console.log('right is', this.cover.year);
    if(decade.toString().substring(0,3) === this.cover.year.substring(0,3)) {
        this.msg = 'You got it right!';
        this.right++;
    } else {
        this.msg = 'Sorry, you were wrong!';
        this.wrong++;
    }
    this.guessReady = false;
    this.resultReady = true;
                
} This basically comes down to checking the first three digits of your guess and the comic's year. I suppose you can say I've got an issue in the year 10000, and you're welcome to come find me then. The last bit I want to share is some CSS. You'll notice that the cover images I shared on top all have a red banner at the bottom that has the date on it. If that showed up in the game, it wouldn't be much of a game. I use CSS to crop it: img.cover {
    width: 415px;
    height: 615px;
    object-fit: cover;
    object-position: top;
    display: block;
    margin: auto;
} This does work properly all the time, but it works well enough of the time. Enjoy! (Note: you could run it in the CodePen below, but it's going to be pretty small. Either hide the source or click "Edit on CodePen" to see it bigger.) not most https://codepen.io/cfjedimaster/pen/ZEVrMoj?embedable=true Also published . here

Creating a Marvel 'Guess The Decade' Game Using Alpine.js

The American Connection.

MAP MY CROP

mapmycrop.com

-- employees

Since n.d.

COMPANY RANKING

MAP MY CROP

EVERGREEN INDEX #7038

Representing this company? Click here to claim and customize this page!

Map My Crop WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

MAP MY CROP

mapmycrop.com

-- employees

Since n.d.

COMPANY RANKING

MAP MY CROP

EVERGREEN INDEX #7038

RELATED COMPANIESRANK

STORIES ABOUT MAP MY CROP

Representing this company? Click here to claim and customize this page!

Map My Crop WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!