News Corporation, a media and information services company, creates and distributes content for consumers and businesses worldwide. It operates in six segments: Digital Real Estate Services, Subscription Video Services, Dow Jones, Book Publishing, News Media, and Other. The company distributes content and data products, including The Wall Street Journal, Factiva, Dow Jones Risk & Compliance, Dow Jones Newswires, Barron's, and MarketWatch through various media channels, such as newspapers, newswires, websites, applications for mobile devices, tablets and e-book readers, newsletters, magazines, proprietary databases, live journalism, videos, and podcasts. It also owns and operates daily, Sunday, weekly, and bi-weekly newspapers comprising The Australian, The Weekend Australian, The Daily Telegraph, The Sunday Telegraph, Herald Sun, Sunday Herald Sun, The Courier Mail, The Sunday Mail, The Advertiser, Sunday Mail, The Sun, The Sun on Sunday, The Times, The Sunday Times, and New York Post, as well as digital mastheads and other websites. In addition, the company publishes general fiction, nonfiction, children's, and religious books; provides video sports, entertainment, and news services to pay-TV subscribers and other commercial licensees primarily through cable, satellite, and Internet distribution; and broadcasts rights to live sporting events. Further, it offers property and property-related services on its Websites and mobile applications; online real estate services; and professional software and service products, which comprise Top Producer and ListHub. News Corporation is headquartered in New York, New York.

News Corp.

newscorp

The latest developments in the Bruce Lehrmann case are an expression of poisonous fallings out and the reckless tabloid culture in Sydney media.

Inside the ‘dirty strategy lunches in media city’ | The Saturday Paper

Crypto bros, rejoice! 2024 might as well be renamed the year of the crypto now that the US Securities and Exchange Commission has opened the floodgates into crypto investment with the approval of a spot bitcoin exchange traded fund, causing the world's most popular cryptocurrency to hit the highest it has ever been since April 2022. And it's not even the end of January! The approval of the ETF, which had been a hot topic for years and had recently picked up steam after the likes of Blackrock and had on the US SEC's door seeking its nod, is now being described as a watershed moment for bitcoin and cryptocurrencies as a whole. Grayscale knocked again But the writing was on the wall though. In fact, a recent poll conducted by betting platform Polymarket showed the odds in favor of a bitcoin ETF being approved by the US Securities and Exchange Commission by the 15th. According to a CoinDesk , an overwhelming 85% of market participants believed the ETF could be approved later this month. Meanwhile, the good folks at Bloomberg, (yes, Bloomberg!), had pegged the chances at over 90%! report At HackerNoon, however, readers were a lot more pessimistic as only 46% thought such a thing would be possible while another 31% thought it might happen, just not this month. HackerNoon CEO David Smooke was one of the beliebers believers (unlike * *HackerNoon editors). some Now that the ETF is approved, it does raise some questions about exactly how valuable bitcoins and/or cryptocurrencies actually are. Most will tell you that the magical number is now in sight, and while that may or may not happen, there are the likes of data provider CryptoQuant which had previously that the ETF approval would trigger a sell off. $100,000 warned As of publication, it doesn't look like bitcoin has actually gone past the $50,000 mark, but that's no reason for the hodlers and diamond hands to worry as optimism surrounding cryptocurrencies , with more and more people pointing at a to drive home the fact that maybe, just maybe, cryptocurrencies will finally live up to their promise (whatever they might be). persists flurry of events And so, it just might be: 2024, The Year of the Crypto™. 👋 You’re reading HackerNoon's Tech Company News Brief, a weekly collection of tech goodness that combines HackerNoon's proprietary data with internet trends to determine which companies are rising and falling in the public consciousness. ubscribe to receive the newsletter in your inbox every Tuesday. S here complete Microsoft, OpenAI Hit With Copyright Lawsuits Ah. What edition of HackerNoon's Tech Company News Brief be complete without a mention of artificial intelligence? I mean, if it's not crypto, it's definitely AI. Well, it looks like and are getting into trouble for profiteering off of other peoples' work by building AI products using content available on the internet. The big one, in case you missed it, was this the New York Times filed against Microsoft and OpenAI for allegedly copying and using millions of its works to build generative artificial intelligence tools that rely on so-called large-language models. Microsoft OpenAI lawsuit But funnily, while the NYT is taking the lawsuit route now, it had been in negotiations with OpenAI on licensing its content. For one reason or the other (maybe it was the money?) negotiations fell through, and the NYT decided to file a lawsuit. But while that plays out in court, the NYT itself reported that other large US media organizations are already in talks with the companies to license their work. We're talking about the likes of Gannett, News Corp and IAC, who are deciding the pricing and terms of licensing the work they produce to Microsoft and OpenAI to build or enhance their AI products, like ChatGPT. In fact, a now says CNN, Fox and Time are also discussing terms with the ChatGPT creator. new report If news organizations licensing their content alarms you, it's important to note that The Associated Press already sells its content to OpenAI and so does Germany's Axel Springer, but that doesn't mean Sam Altman just wants to keep everyone that comes knocking to his offices. In fact, only recently the OpenAI CEO tried to make a case for continuing to use copyrighted work to train AI by arguing it under fair use. That is essentially the that Mark Zuckerberg's Meta plans to make if any of or 's users try to sue the company for using their posts to develop the company's own AI. fell same argument Facebook Instagram But think about it, though. If something like ChatGPT can learn how reporters write stories, could it then replicate seemingly believable stories to fool the masses into one action or the other? The future looks bleak if one were to believe so. It seems like jokes about AI manipulating people by controlling the media and the news are no longer going to remain jokes. 🤫 Microsoft ranked #14 on HackerNoon's Tech Company Rankings this week. New Year, New Layoffs Looks like we're experiencing a repeat of 2023. While the extent of job cuts isn't as severe as the one we saw at the beginning of 2023, just announced that it would be laying off hundreds of its employees across multiple teams, and damn*!* Google Last year was a tough time to be working in tech as almost every major tech giant announced job cuts affecting thousands of people. We're talking , , , Google, - you name it, and they were there. The rationale was that these companies had hired way too many people during the COVID-19 pandemic and didn't really know what to do with them anymore now that demand for their products was tapering off (and like, why wouldn't it, now that people could go outside and touch grass!). Meta Microsoft Amazon Twitter/X While things at Google might not be pleasant, they're downright nasty at videogame software provider Unity Software which is of about a quarter of its workforce (roughly 1,800 people). The company recently courted some controversy after introducing a fee system for developers using its software to create games; the move was not popular, forcing the company to . letting go revert the changes To top things off, Amazon too is of several hundred people at some of its divisions marking yet another tumultuous start to the year for tech works. letting go Google ranked #2 on HackerNoon's Tech Company Rankings this week. Amazon was on the #11 spot. In Other News.. 📰 Apple shares the most popular podcasts of 2023 — via Apple CES 2024: The weirdest tech, gadgets and AI claims from Las Vegas — via TechCrunch Snapchat to let parents decide whether their teens can use the app’s AI chatbot — via CNN Microsoft overtakes Apple as world's most valuable company — via Reuters Exclusive: Even Bill Gates was surprised by ChatGPT — via Axios And that's a wrap! Don't forget to share this newsletter with your family and friends! See y'all next week. PEACE! ☮️ — Sheharyar Khan, Editor, Business Tech @ HackerNoon

The Year of the Crypto

ক্রিপ্টোর বছর

Das Jahr der Krypto

El año de las criptomonedas

L'année de la crypto

क्रिप्टो का वर्ष

暗号通貨の年

암호화폐의 해

O ano da criptografia

Год крипты

Kripto Yılı

Năm của tiền điện tử

加密货币年

All about the OTT App Market The best over-the-top, or OTT, apps provide an essential turnkey to offer customized solutions. find this a potential opportunity to launch a scalable application in just a matter of weeks! For example, Netflix or Hulu over-the-top services have turned the entertainment sector into a profitable industry, especially for content owners, social media marketers. OTT platform providers These apps enhance popular video streaming services with essential features to stream digitized content in smart TVs, laptops, personal computers, androids, & many more compatible devices. This helps them facilitate elevated user engagement within their best OTT platforms, providing customers with a convenient and ad-free experience. According to , the global video streaming market is expected to have an impactful compound annual growth rate of about . grand view research’s report 21% from 2021 to 2026 Outlining Predominant Market Players: 1. Netflix The ever-growing top of by Marc Randolf and Reed Hastings. Initially, they pioneered the very concept of renting DVDs through the mail by charging viewers a monthly fee. Later, they came up with TVshows, movies directly streaming into , they grew as top distributors and producers for both of the streaming genres mentioned earlier. OTT platform Netflix was started in 1997 customers' PCs. In 2012 To top it all, the movie streaming apps in both media formats, Android and iOS apps, have an impressive collection of content libraries which include original series, movies, and programs captured in distinctive languages, making it well-diversified. Undoubtedly, that states the reason why Netflix carries . paid customers spanning 190 countries Some of the amazing factors that pick up attention are: Netflix largely captures a 63% market share in the U.S. OTT video streaming industry is expected to see a hike of $73.9B by 2024 In the majority, 60% of Gen-Xers watch their streaming videos on a daily basis Watching Pay-TV is fading, and users allot less than 30 mins a day 2. Hulu The pioneer OTT video platform of Hulu, based in the U.S., provides video streaming subscription services. and was initially a joint venture between NBC, News Corp., Province Equity Partners, and later The Walt Disney Co. but is now majority-owned by Disney. Hulu was launched in the year 2007 This OTT streaming solution facilitates users with unique and unfettered access to new episodes, and in 2017 they launched Hulu with Live TV offering over-the-top IPTV services that featured linear television channels.  As of 2021, Hulu carries 41.6 million subscribers. A look into some exceptional facts: As per records, Hulu is known to invest more than $2.5 billion each year on content Surprisingly, Hulu Ads are found to be 2x more effective than TV Ads Hulu OTT TV app allows registration via email as well as social media They offer 43 different video ads to their users every month Hulu is one and only available in two countries: the U.S. and Japan 3. HBO Now HBO Now is one of the biggest inspirations you can use to build your best OTT app development company. It was , and this American subscription video-on-demand streaming service became one of the favourite amongst consumers. launched officially in 2015 The surprising factor catches our attention when we come to know that this 5-year old standalone Subscription-based OTT solution provider facilitates access control to the platform through HBONow.com, Google, or  Apple TV to their users by effortlessly getting switched to Max no extra charge.. Interesting facts that hold and attracts views are: HBO Now has more than 10,000 hrs of content at launch Home Box Office (HBO) was originally named the green channel HBO carries around 44.2M subscribers as of the fourth quarter of 2021 In the early 1970s, they distributed signals using microwave Its first live event coverage was 1973’s Pennsylvania Polka Festival So far, we have seen the industry's pioneer leaders of OTT streaming solutions and their growth over time. Now we will get into its types and check how to create an exclusive OTT app! What are the Types of OTT Apps? There are two major types in which you can capture your audience’s attention when you plan to have your own application specifically designed via OTT app development. Primarily, they are developed where the majority of your audience resides to watch your content. And, that is in mobile or television. Both are broadcasted in audio as well in video formats. You can have your dynamic OTT apps in a categorical manner, suiting its niche preferences. 1. Video When you plan to have an effective interactive streaming platform or an app in the OTT industry, you want to engage markets, as most service providers do, with high-quality video players and content with clickable user interfaces to entice all kinds of demographics on the move. 2. Audio You can set your niche brand by having audio-based OTT platform apps where your users can listen to your curated pieces of OTT content, such as live audio, engaging podcasts, and radio broadcasts. Now, in the next segment, let’s take a look at: Building a Streaming App for OTT Business 1. Wherever your target audience is, be there Due to sophisticated streaming capabilities, providing an engaging viewing experience with high-end quality is the backbone of your OTT development. As a result, you can quickly reach out to every target base through devices like smart TVs, Android, iOS, mobile or web apps;  smartphones; tablets; desktops, etc., and observe user behavior to serve as per their preferences. 2. The Key to Engagement is Personalization Only by providing new streams of material within your white-label OTT platform, your audiences will remain continually engaged. For example, if you broadcast a big movie every Thursday, people will devote their entire schedule to watching it within the OTT branded apps on the go. So, pick which form of personalization will catch their interest and stick to it for a while. 3. Each Screen Has Its Own Viewing Experience Every device has its own unique OTT TV solution viewing experience that is feasible and versatile for your users worldwide. Also, with streaming analytical capabilities, you'll be able to target and analyze every device user who can now effortlessly use your platform and make any subscription transaction. 4. Help Audiences Find Your Content In Seconds Is your information categorically placed in a way that allows easy searchability? If you've already done it, the best method to talk about your popular content is to incorporate a theme around your OTT platform development, such as movies, music shows, kids, news, games, and so on, so that people may find your content under these categories while they're surfing the web in general and have no specific preferences. 5. The way to scale your OTT TV platform success Through analytical tools and feature-rich services offered under a single white-label OTT solution, you may constantly or periodically update your over-the-top software's functionality and control its efficiency. Furthermore, you can now effortlessly acquire insights like audience tracking data, live custom analytics, browsing insights, reviews of your OTT app platform, and more while on the go! Highlighted Features to Determine Your OTT App Revenues 1. Easy App Onboarding Improve your user’s entire OTT streaming app experience by making registration, verification, and profile creation as simple as possible. Make it possible for them to sign up through social networking platforms as well. 2. Search Feature It’s a must for over-the-top app builders to ensure your users have smooth navigation where they can find their favorite content. It's a quintessential function that you may customize with genre, language, and other options. 3. Multilingual Content Definitely, your subscribers might want to enjoy streams globally. They need to understand your content in the language they prefer and make it more relatable. So, your OTT streaming app should be flexible in streaming content in many languages. 4. Watchlist The watchlist integration lets you know whether your viewers want to save items for later viewing. In addition, this feature will aid in creating an in-app notification strategy and monetize new streams that are introduced using their own white-label OTT solution. 5. Payment Gateway Integration Whether you're using Google or Apple as your operating system, you can add your own payment gateway service. In addition, in-app purchases in advanced mobile frameworks are available and made easy for your potential customers. 6. Push-Notifications Use pop-up messages to remind your platform members about their membership expiration dates or promote new popular content available for purchase. This has the potential to boost app engagement and reduce retention rate too. 7. Social Feature You may also freely promote your developed OTT apps by including a share button on social media handles to increase your chances of attracting new users through online marketing. 8. Subtitles This functionality lets your users engage with non-localized material and boost the engagement of your films in apps created by your in-hand OTT developers and are streamed globally. 9. Ratings and Reviews Allow your viewers to evaluate and rate videos, movies, or shows that they just saw to collect their opinions or thoughts and share them with other users via rapid clicks. Popularize your content in social forums & similarly have user personalized feedback. 10. Admin-Panel Manage your app's functionality and growth in real-time with an all-in-one live dashboard to see exact view counts, subscribers, and other metrics and improve the user experience with new features. How Much Will It Cost to Develop an OTT Application? 1. Platform Genre I am sure the type of platform with which you plan to skyrocket your revenue will be outstanding! Besides, you can have two choices of building a SaaS platform or else go for SaaP venture. This purely depends on your brand objectives and cost factors. 2. Features & Tech Stack Choose the ideal list of features, technology-driven capacities for you to launch an enticing business based on for setting business cornerstone with these top emerging (OTT) services. OTT app development OTT App Support Any Industry 1. Content Owners They are the foundation of building any OTT platform that is used to broadcast a creation. These content creators have a variety of opportunities to promote their work, including VOD, movies, virtual events, seminars, and more, all of which rely on the power of streaming technology. 2. Health & Fitness Many fitness enthusiasts & health activists find this a key source for staying healthy and fit, allowing fitness professionals who stream their valuable workouts through OTT services that provide expert advice and supervised programs. 3. Enterprise Owners Every organization has now become touch-friendly. Also, pairing to streaming technology has made visual presentation more effective, resulting in a significant increase in the revenue ladder. Entrepreneurs have a tremendous influence over their target market with the strategic exhibition of ideas that give practical answers. 4. Schools and Universities The advancement of technology has provided great opportunities for connecting and catching up on learning! Curriculum tutors and university chairpersons can now use the e-learning mobile OTT platform to hold live sessions, webcasts, and other events using the power of smart applications. 5. Sports Clubs Now, sports personalities or athletes may easily broadcast their must-see events or tournaments via mobile connectivity, courtesy of launching their own sports-based OTT tv apps, allowing every sports lover to enjoy the classic replays or best shots via video on demand. 6. Personal Trainers Create a bespoke OTT app with leading streaming solution providers and tailor your training platform with live streams, one-to-one personalized video conference sessions, physical fitness courses, and more. 7. Charitable organizations As charitable or NGO activists, you can reach the poor much more quickly than you might think, just as celebrities using real-time streams, motivational on-demand videos, and vocational streaming courses to raise funds. Top OTT App Monetization Strategies With different video monetization possibilities that revolve around learning & apps, a personalized library of increasing income spikes is always worth it. Here are a few examples: how to create OTT platforms 1. SVOD A subscription-based video-on-demand package, also known as SVOD, can be used for a set period of time, such as monthly, quarterly, or annually. Establish a loyal customer base and regular revenue for your streams. 2. AVOD Advertising-based video on demand monetization, which is popularly known as AVOD gives you an ad-supported streaming app. Build an OTT platform with third-party ad integration apps to captivate large audiences and instantly gain increased viewership. 3. TVOD Transactional-based video on demand, often known as pay-per-view, is a great way to generate income for your premium video streaming. Showcase top-rated VOD and one-time watch videos by charging per content and quickly increase revenue. Last but not least: Hey folks, we have now come to the end of this blog, where you have gotten a detailed picture of how to start and how to proceed with an OTT streaming business. Our list of guides can help you outline deciding points for shaping the finest OTT app/platform to construct a worldwide user base. Also, please let us know what you think in the comments section below for more information on how to use the potential of OTT right now!

How to Create a World Class OTT App for Mobile and Smart TV

Does anyone still remember MySpace? Well, the story about MySpace should be told over and over again. Why? Because you can learn a lot from it. To begin with, they made a movie about called “Social Network,” if I’m not mistaken. Now, if you’re asking me, they should’ve made a movie about the rise and fall of MySpace. That would’ve been one more interesting and far way more dramatic movie. The sweet irony is that Justin Timberlake, who had one of the lead roles in this movie, now owns MySpace. Who could have thought that would be possible? Maybe he got an idea to buy MySpace while participating in a movie about Facebook. I don’t know for sure. What I do know is that MySpace deserves a special place. Here’s why. Facebook Oh Boy, Those Were the Days Back in the day, from 2005 to 2008 MySpace was THE most popular, visited, and used social network in the world. “By 2006, MySpace was being visited more times than Google Search and Yahoo! Mail, becoming the most visited website in the United States. In June of 2006, it was reported that MySpace was responsible for nearly 80 percent of all traffic-related to social networking sites.” That’s something not only Facebook, but any other social network out there can only dream about. Who could have thought that it would be possible to lose $600M in six years? That’s precisely what happened to MySpace. “News Corp bought the site for $580 million in 2005. The company eventually sold MySpace to Specific Media for $35 million in 2011.” So, What Went Wrong and Why? Any online platform with millions of users is a gigantic money-eating machine, plain and simple. Here’s what good old Wiki has to say about that: “Marvin L. Gittelman suggested that the $900 million three-year advertisement deal with Google while being a short-term cash windfall, was a handicap in the long run. That deal required Myspace to place even more ads on its already heavily advertised space, which made the site slow, more difficult to use, and less flexible. Myspace could not experiment with its own site without forfeiting revenue, while rival Facebook was rolling out a new clean site design.” Truth be told, things are a bit more complicated in MySpace’s case. You know what they say. All great empires fall from within, including the digital ones. The virus of decadence is alive and kicking. He obviously didn’t die out with the fall of the Roman Empire. While MySpace was occupied with Facebook was working hard with the modest underdog budget. The rest is the shiny Facebook’s history and the new social networks world order. “the opening of extravagant new offices around the world, ” “According to Businessweek, the pressure to drive revenue, when other startups were using the luxury of venture money to create and explore, stifled the possibility that MySpace would have the breathing room it needed to innovate. Instead, executives were focused on trying to raise advertising profits.” What Does It Mean When Someone Says That You Are the Next MySpace? When someone says that your startup, company, business, platform, or whatever is the next MySpace, you have to immediately stop what you’re doing and take a deep breath. It’s not a curse, but it’s a serious warning. This means that you’re already on the road to Mandalay, but you just don’t know it yet. So, here I am and I am asking this question, Now, Upwork may reply, my dear Upwork are you the next MySpace? what are you talking about? We’re having the time of our freelance lives. The trouble with Upwork, though, is that it just had its “MySpace Google advertising” moment when it made its NASDAQ debut. Meaning, that Upwork took the money, and now it has to live up to its investors’ expectations the same way MySpace had to satisfy Google’s advertising appetites. We know how that one turned out for MySpace. The pressure is going to be enormous and Upwork will have to be very creative in finding new ways to ensure liquidity. Here’s the problem. Upwork already doubled its fees. That wasn’t enough. So, it had to go public. I’m afraid that’s not going to be enough too. Why? Feel free to check Upwork’s NASDAQ situation. While you’re reading this article, its shares are hitting a record low. Punk Is Not Dead, Neither Is MySpace MySpace is still alive, but it isn’t kicking, that’s for sure. The numbers never lie. “Last November, the site got 50 million unique visitors in the U.S. That’s a lot of visitors for a site that most people haven’t thought about in 8 years. Facebook currently boasts 864 million daily active users worldwide.” How do MySpacers feel about it? Well, here’s what they had to say to Business Insider: “MySpace: Yes, Facebook Kills Our Traffic, But At Least We Make Money!” Now, that’s a definition of a positive attitude. Steve McQueen had put it nicely in “The Magnificent Seven.” “ It reminds me of that fellow back home that fell off a ten-story building. As he was falling people on each floor kept hearing him say, “So far, so good.” Well, Upwork, so far, so good, but for how long? Also published here. Vote for this and my other freelance-related stories in Noonies 2022 category: . HACKERNOON CONTRIBUTOR OF THE YEAR- FREELANCING

Is Upwork Going to Pull a Myspace?

¿Upwork va a sacar un Myspace?

Upwork va-t-il tirer un Myspace ?

क्या अपवर्क एक माइस्पेस खींचने वाला है?

Upwork は Myspace をプルする予定ですか?

Upwork vai puxar um Myspace?

Upwork có tạo được Myspace không?

Upwork 会拉动 Myspace 吗？

How would you feel if I told you that making business-related decisions doesn't have to be hard? What if I said that, in reality, it can be as simple as knowing your values? Having strong values is something I take pretty seriously but in a subconscious way. I hadn't given the concept much thought until my interview with ex- -employee-turned-entrepreneur, Sarah Gibbons. Amazon Values, as it turns out, can mean the difference between success and failure in business. Those with strong values have a clear advantage over those who don't because they know what they stand for. This allows them to make better decisions, faster – and they're less likely to waiver when things get tough. Today's newsletter is going to be a brief but powerful one. Let's talk about the importance of values in and how you can use them to make better decisions. business Sarah Gibbons' Take on Values I interviewed Sarah on the Success Story Podcast about a year ago now, and it was one of the most value-packed (pun intended) interviews I've ever done. We only spoke for about 40 minutes, which is much shorter than my usual shtick – but in that time, we talked about: Sarah's pivot from the corporate world to entrepreneurship, The true meaning of fulfillment in business and how it relates to growth, Re-defining leadership, Knowing your values and sticking to them, Spiritual psychology and how to stop yourself from getting triggered, and Other important lessons for entrepreneurs. If any of that appeals to you, I'd highly recommend checking out the interview. It was one of my favorites to do. But for the purposes of today's newsletter, I want to focus on something Sarah said towards the beginning about values: "Once you get clear on your values and you define them on your terms, then you have a very different experience.” Let's dig into exactly what Sarah means by this. From Corporate Work to Values-Based Leadership If you saw the interview, you know that Sarah had an incredibly interesting start to her career. She worked for Amazon's IMDB division for a while, and then she was asked to join MySpace (now News Corp). Can you imagine being part of that team in the early days? As many of us have experienced, though, Sarah soon grew tired of the corporate 'way'; the ladder climbing, the high stress, and the constant grind. She was in search of a new type of fulfillment – and again, like many of us, she found it in entrepreneurship. "I became apparent that I did not want to go back into digital media. The part that I loved was always hearing people's stories, but also getting into people's leadership and really helping them see what's possible beyond what they were currently doing. I ended up starting my coaching business, and with a lot of trial and error, ended up where I am now." Where Is She Now? Sarah is actually something called a spiritual psychologist; she coaches leaders who have become so focused on their success, that they've neglected their personal happiness, growth, and fulfillment. What an incredible role to play in someone's life. "When I got my masters in spiritual psychology, it started to introduce to me a way of leading that used practical tools – particularly around communication – that helped me realize the way in which fulfillment works is through exponential growth, not through doing things on a horizontal line." It was through her training that Sarah began to see just how important it is to connect with and be in alignment with, your core values. "I started to ask, what does it mean to live from my values? What does it mean to really connect with my own intuition, and start to make choices that honor what I want, versus these expectations that I had for myself based on parental and societal influence?" Sarah's training helped her to reach many pivotal realizations about her business, her life, and her way of perceiving the world. "I can create what I want – as long as I get clear, and I'm paying attention to my mindset and my intention, and I'm taking action. What I know to be true is that, then, the universe will meet you at the point of action." Nowadays, Sarah coaches her clients to really get in touch with their core selves and align everything else accordingly. It's a highly values-based approach that helps people make the right decisions – not based on expectations, but on their highest-held beliefs. "I now teach my clients around communication leadership, and so they're having very similar experiences. Many of them are staying in their current roles, but they're experiencing a level of fulfillment and impact that they didn't think was possible." Where Do Values Come In? I was excited to ask Sarah about her views on core values because it's a topic that I'm incredibly fascinated by. What role do our values play in our lives, and how do they shape the way we live and work? Values Shape Your Perceptions “Think about leadership as ‘who you informs what you do’. So if you are disgruntled, overwhelmed, and anxious on the inside, then as you are doing whatever it is you need to do, that's the experience you're going to have.” This was Sarah's insight: values shape the way in which you see and experience the world. If one of your values is compassion, then you're going to be more likely to see the world through a compassionate lens. I find this so powerful because it means that we can actually change the way we experience life. What if I were to change from a mindset of scarcity to one of abundance? Would I then start to see more opportunities? Sarah believes that the answer is yes; her experience has been that, when clients shift to define and live by their values, they start to see different results in their businesses and lives. Values Influence Your Attitude “One of my top values is connection. So as I'm moving throughout the day, I'm asking myself, ‘What value does this align with? Is this in accordance with my values?’ That's a very different experience than ‘What should I be doing?’” Similar to the point above, Sarah believes that our values influence our attitude. If we're constantly questioning whether our actions are in line with our values, we're going to approach life – and work – very differently. For instance, if my top value is freedom, then I'm going to be very aware of anything that constrains my freedom. This might manifest in the way I work – perhaps I'll only take on projects that allow me to work from home, or maybe I'll start my own business so that I can set my own hours. If I value honesty and transparency, I'm going to be much more likely to build relationships based on those values. I might avoid working with companies that have a history of misleading their customers, or I might make an effort to be open and honest in all my interactions. Values Help You Understand Yourself More than anything, your values are a clear indication of the person you are – because values are, at their core, what you believe in. “One of the best ways to get really clear on who you are is to define your values. Everyone has a round 5-7 values; once you get clear on your values and you define them on your terms, then you have a very different experience.” Sarah explained that most of her clients insist they know exactly what their values are. After a bit of digging, though, she soon realizes they're not clear on what their values actually are at all. For instance, they'll list “family” as a value – but when Sarah probes a little further, she realizes that what they really value is connection. Or they might say “success” – but what they really mean is freedom. So, how can we overcome this obstacle? How can we make sure that we're clear on our values? Defining Your Values as an Entrepreneur , your values are the beliefs that determine what your priorities are. They're not simply 'I try to be nice to people or 'I believe in the afterlife' – they go much deeper than that. According to MindTools Some values are personal (like honesty or courage), while others are more universal (like fairness or respect). You might have values that guide your personal life, and different ones that shape the way you work. As an entrepreneur, it's essential to be clear on what your values. After all, you're the one in charge – which means that your values will have a direct impact on your business. So, how can you go about defining your values? I had a look at Sarah's handy ten-minute guide to finding out exactly where your needle points. Here's what I took away from it: 1. Find a list of values. There are plenty of resources for this on the internet; has one. The point of looking at a mega-list like this is to help you start thinking about the values that might be important to you. MindTools It's also interesting to see what counts as a value, and what doesn't. For instance, I would have never considered 'beauty' to be a value – and I thought that 'financial wealth' would be included, but it's not. Here are some of my personal favorites from Sarah's list: Determination Growth Respect Commitment Compassion Wisdom Study the list, and start making a note of the values that stand out to you. I really enjoyed this first step because it made me take the time to consider what I actually value, rather than vaguely understanding myself. 2. Narrow it down to your top 10 values. This is a bit harder than it sounds, but Sarah has a great tip: "You'll know when you hit it because you'll feel it in your gut; you may have memories, feel inspired, or feel a connection." What Sarah actually recommends is to create five core values instead of ten basic values. For instance, if you value kindness, empathy, and love, you might group those under a core value – 'compassion' – which would be one of your five. What this does is it makes you think harder about what each value means to you. It's harder than you think to draw connections between values, but don't overthink it. Your gut will tell you more than your brain ever could. 3. Define each value on your own terms. This is the part where you get to be creative; turn those core values into mini definitions and principles to live by. For instance, if one of your values is 'compassion', you might say: "I practice compassion by being understanding and kind, even when it's difficult." Or if your value is 'growth', you might say: "I commit to continuous learning and self-improvement." Sarah also recommends answering a few questions to help with the process: How might this value shape the decisions you make today? Which value do you live in the most, and what does that look like for you? Which value do you live in least? How can you re-define it so that it feels more natural for you? From there, it's simply a matter of getting super familiar with those definitions. Maybe you read through them as part of your morning routine, or post them on your vision board. Either way, the idea is to keep those values top-of-mind, so that you can make decisions based on them in your day-to-day. Where Your Values Will Take You My interview with Sarah was such a wholesome one, and it really got me thinking about how our values shape the way we live. But what's interesting is that, as entrepreneurs, our values will also shape the businesses we build. And that makes them crucial to success. In fact, that values-based decision-making leads to better business outcomes. It makes sense, really; if every decision is based on the values that are most important to you, then you're more likely to end up in a place that feels good – both for you and for your business. studies have shown Of course, that's not to say that every decision will be easy, or that there won't be trade-offs. But if you know your values and are clear on what they mean to you, those decisions will get a lot easier. Sarah says it best: “It allows people to feel so much more empowered. Get clear on your values, define them, and you'll be amazed at the different kinds of results you get both in the workplace and at home.” Don't forget to check out Sarah's , and I'll see you next week! full interview Also published here.

What Do You Think Your Values Are?

“ ” Venture services focuses on solving a specific sub-section of founder needs: the need for quality software and the need for capital to pay for it. This post is the second part in a series on and . examined the ‘what’ and the ‘why’ of venture, primarily from the perspective of the providers. venture venturing Part One This second part deep dives into ‘how’ venturing actually works in the context of , the corner of the venture ecosystem I occupy as a partner of London-based . venture services Pivot Venture Services My first experience of venture was via , a pioneer of the corporate venture model. facilitate a number of innovation programmes globally and, in collaboration with NewsCorp, helped birth , for whom I was Product Director in 2015. PreHype PreHype Newsmart My first real break in venture came as MD for (aka ‘Castle Digital Partners’), a venture services firm based in Charlotte, NC. I set up Castle’s London Studio and inked their first European venture deals. Castle Venture, in all its forms, is the professionalisation of angel investing. Angel investment has a bad reputation. One that is characterised by chaos, amateurism and exceptionally high levels of risk. I once attended an investment seminar where a seasoned angel presented his view that one shouldn’t expect to make any money from angel investing and that it was more about having fun. This seemed counter-intuitive. But when your company is you, a big idea and a slidedeck, it’s hard to convince others to back you (especially if you’ve never started a company before.) Given the obstacles, it’s a miracle first-time founders ever make it out the gate. Super-early startups are a ‘dead-zone’ analytically. Valuations, forecasts, market sizes and business plans are like ; simultaneously worth zero and one million. Shrodinger’s Cat An early-stage startup valuation. Institutional investors don’t engage until much later on. They are , as opposed to startup, investors. They enjoy the comparative luxury of a working product, market traction, even revenue from the companies in which they invest. scaleup What is Venture Services? Venture services firms serve two sets of interests, those of and those of . They facilitate transactions that align the interests of both towards the rapid creation of equity value. early-stage founders early-stage (‘angel’) investors Jeff Bezos is well known for advising founders to if they want to build a durable business. He bet on customers’ desire for low prices and fast delivery not changing and both seem to have paid off well. ‘bet on the things that don’t change’ For venture services firms, the bets are as follows: We are betting that startup founders will continue to need capital and great quality software. We are betting that wealthy individuals will increasingly want to invest in startups whilst reducing the risk of losing their investment. (In the UK in particular, they will continue to take advantage of the UK Government’s generous tax relief schemes, such as , for as long as they are available. More on this later…) SEIS and EIS Bet #1: Founders Proptech. Lawtech. Connected Enterprise. Smart Cities. IoT. The tech-sector is diversifying rapidly. And so is the archetypal startup founder. Outside of Silicon Valley, it’s increasingly rare for first time founders to be hackers or coders. Instead, most are who have identified an opportunity to disrupt or transform the status quo with software. They are, in , “just some business person.”* domain experts the words of Paul Graham [ : Mr Graham uses this term in a somewhat derogatory manner. We most definitely do not.] *Important note And for “just some business person” with a vision but no technical experience, the challenge of building an initial product and finding sufficient capital to pay for it can be insurmountable. It’s a much greater obstacle to startup progress than is commonly acknowledged. Historically, non-technical founders looking to start a company have had two options to choose from. they find a technical co-founder / CTO and rely on them to build a product in exchange for a healthy slice of equity. Either they reach further outwards to one of the many providers of software services (aka ‘dev shops’), typically located on the eastern edge of the European Union (think Poland, Czechia, Belarus, Ukraine, Russia), or in India and pay them in cash. Or Each route has its pros and cons, both of which have been well documented and . here here Venture services firms offer a third option to founders, one that sits between an invested employee and a detached supplier. And one which, ideally, offers the benefits of both. These firms don’t create companies or lay claim to their ideation. They partner with mission-driven founders to solve a specific subsection of their needs: the need for working software and the need for capital to pay for it. For startup founders, finding a partner willing to remove the two biggest obstacles to early-stage progress can be compelling. It’s common to hear founders talk of their desire for . Done correctly, venture services should feel like an accelerator on steroids. value-added capital By trading in equity, venture services firms attain goal congruence with founders that suppliers are unable to achieve. This creates a partnership culture of trust and shared outcomes. What’s most appealing to investors, however, is that venture services significantly . This includes team risk, technical risk and (above all) execution risk at pre-product stage. reduces risk The percentage of startups who burn through their initial capital but fail to deliver an actual product remains surprisingly high. This usually occurs when CEOs unexpectedly part company with their CTO and find themselves with half a product that has little or no value. Experienced angels will have witnessed this on multiple occasions. Having confidence that a product will be delivered, on time and on budget, is rare and appealing. Bet #2: Investors Investment decisions don’t exist in a vacuum. They occur within a complex macro-economic environment of financial incentives and opportunities, notably tax relief and grant-funding. All governments seek economic growth. As startups are high-growth vehicles by definition, governments are increasingly trying to incentivise early-stage investment. Widespread awareness of the generosity of these schemes, combined with the explosion of interest in startups, has created a perfect storm of capital searching for a home. By way of example, let’s examine the UK government’s , better known as SEIS. For early-stage startups, SEIS is the financial equivalent of methamphetamine. Seed Enterprise Investment Scheme In the words of , CrowdCube “To say that the Seed Enterprise Investment Scheme, or SEIS, is incredibly generous is somewhat of an understatement. Of course we all hope our investments do well, but if you pay tax at 45% and make an investment of £10,000 that fails completely, you only lose £2,750 due to the tax relief.” Wealthy individuals with high income tax bills have stampeded to find startups that will allow them to take advantage of the tax relief offered by SEIS. But few have direct access to qualifying companies or the time to find them. Whilst the relief is an end in itself, investors also want to protect their investment and minimise downside potential. They’re not just looking for startups that qualify for SEIS tax-relief. They’re looking for that qualify for SEIS tax-relief. the best startups Via access to select early-stage startups, venture firms are well positioned to broker these trades. How Venture Services works In Part One, I observed that the economic details of venture deals are rarely disclosed. It’s hard to quantify how much equity (and cash) really changes hands between founders and venture firms. The spectrum is very broad. As a first stab at demystification, below is a breakdown of how a successful venture engagement works at . It’s a happy path, one in which all the boxes are checked. Pivot The very short summary is that venture services means building software in exchange for equity, syndicating part of the equity to cover costs, and retaining the remainder as margin. This aligns the firm’s incentives and upside with those of the founder(s) with whom they partner. The slightly longer version is as follows: The firm meets a startup founder and makes an investment decision based on the market opportunity and the extent to which quality software will create momentum and initial traction. Working closely with the founder(s), the firm scopes and estimates the cost of the software development required to help the startup reach its agreed goals. The firm’s Investment Manager determines if additional cash is necessary and calculates the equity premium the firm will add as margin. The firm agrees a pre-money valuation with the founder(s) and makes a proposal to develop their software in exchange for a % of the equity in the company. If both parties agree to terms of the venture proposal, the founder(s) transfer the agreed equity to the firm, typically via an option or an Advanced Subscription Agreement (ASA). The venture firm designs, develops and deploys the agreed software for the startup and maintains/iterates upon it for an agreed period of time. The software engineering team are paid directly by the venture firm from their fund capital or by syndicating a % of the equity to investors in exchange for cash. Why Venture Services is Hard Compared to other forms of venture capital, venture services remains niche. This is primarily due to the complexity of the business model. There are more moving parts, and more factors to align, in a successful venture deal than a conventional investment. Most investors prefer to avoid the additional overheads and responsibilities of being a venture partner. Additional reasons I have observed are as follows: #1: Cashflow In Part One, we explored why venturing can’t be funded by profits from a services business; at least not long-term. Services businesses generate revenue in the form of cash. Venturing is investing, which is deferred revenue. What makes the two irreconcilable is usually the cost base. As software companies are asset-light, the lion’s share of their cost base is payroll. If you have a lot of salaries to pay at the end of each month then you need the cash to pay them. And, if all of your profits are funding ventures, then you are dangerously exposed in a downturn. I have experienced services companies with large workforces all but go bankrupt trying to move into venture. The consequences were drastic and unfortunate, mainly for the employees. So, how can venture firms build software without the overhead of large team of developers and designers? We solve for this at Pivot by mobilising teams on a deal-by-deal basis and working within a tightly managed, fixed cost environment. We generally employ a mixed model, with CPO and CTO working ‘onshore’ (in our case, in London) and UX, design and engineering working ‘near shore’, typically in Eastern Europe. We always seek to build long term partnerships by working with the same individuals across a number of ventures. This creates bonds of familiarity and trust that generate real team effectiveness over time. The costs for the team are rolled into the overall capital raised by the venture firm and then exchanged for equity in the company. #2: Market-size Finding startups that are good investments is hard. Finding startups that are pre-product, pre-revenue a good investment is a treasure hunt of mythical proportions. and It’s hard to accurately size the market for venture services. It be very small based on the ratio of deals we have done versus the opportunities we have passed on. could The downside of working at the highest risk end of the investment spectrum is that it’s, well, risky. Most investors prefer to see proof of market traction. It’s a giant leap of faith to invest in a market for something that doesn’t yet exist. Pivot endeavours to de-risk this by backing founders who sell back into their primary network, at least in the first instance. Ex-lawyers are well positioned to sell lawtech solutions to other lawyers. Ex-marketers selling marketing platforms to their former (marketing) colleagues is a reasonable bet. Unlike hackers, “just some business person” often has the edge when it comes to sales. And sales is the bigger challenge long-term. #3: Culture Venture services firms combine expertise (and application) in two distinct fields: early-stage investment and software product development. Folks from these respective fields come from different places, have different mindsets and speak different languages. Service providers are instinctively inclined to work with whoever wants to work with them. Investors are instinctively sceptical. Service providers are default ‘yes.’ Investors are default ‘no’. I speak from experience when I say that the two are also very differently by founders and the startup community in general. treated Investors are top of the food chain and get treated accordingly. They enjoy high response rates to their emails and meeting requests. They don’t sell or hustle as a rule. Life is different for providers of software services. Much of their time is taken up writing estimates and proposals that go unacknowledged. Their communications are ignored. They have to hustle constantly. Part of the achievement of a venture services firm is finding a common ground for these very different individuals to interact within. This alone is a barrier to entry. TL:DR That’s pretty much all I have to share, so far. As you can probably tell, I’m bullish about the potential of venture (and venture services) as an emerging category within venture capital as a whole. Any other players (or aspiring players) within the space, I’d love to hear from you. : Here’s the TL:DR Venture is still a niche space, primarily due to the complexity of its business model. All equity deals (ie. where startups pay no cash) are the most compelling, differentiated, and potentially lucrative form of venture. Smart founders value the opportunity to onboard an experienced product team at speed and are prepared to trade their equity accordingly. It’s impossible to succeed in venture without access to a pool of capital, typically a fund or syndicate of investors. Keeping overheads as low as possible is key to success. Whilst the value proposition is compelling and the problem large and frequent, the size of the addressable market remains hard to quantify (and may be very small.)

On Venture Services

It is unlikely that the word “contributing” means anything to people who neither work in IT industry nor actively participate in the development of open source projects. However, in the light of recent events, there is every reason to believe that such terms as “contributing” and “contributor” will soon become very popular. You might ask “why?” Let’s start with the common IT terminology. Contributor is a person participating in the implementation and improvement of innovative ideas and promising projects. Contributing is a type of activity aimed at implementing interesting projects to gain experience in this field. Mostly, this means involvement in the process of developing new functions, improving open source projects, and fixing bugs. One of the largest hosting services for contributing to open source projects is GitHub. In the IT field, such activity is usually conducted by volunteers. However, in the Bitbon system, which was developed on the basis of Blockchain applying the smart contract technology, contributing acquired a fundamentally new meaning and became a separate type of business and financial activity. Unlike contributing in the field of investments and IT, contributing in the Bitbon system means the ability to earn money without studying in detail the niche for investment or spending huge amounts of time. A vivid example of unsuccessful investment as well as insufficient analysis of the market situation in an invested niche was the case when News Corp bought out MySpace for $580 million in 2006, at the peak of its popularity. Later, it turned out that MySpace could not withstand competition from Facebook, and the social network was sold again, but this time for only $34 million. In the Bitbon system, the contributor selects an interesting project from those already verified by qualified specialists, which not only increases the return on the contributor’s investments but also saves his/her precious time. It is exactly the reason why contributing can serve both as additional and main source of income — after all, contributors themselves can determine the degree of their involvement in a certain project. In 2009, Blockchain wasn’t internationally recognized, and Bitcoin just began to gain popularity. At that time, probably anyone could have invested in both projects. It is not surprising that, since then, the development and digitalization of economy led to the emergence of new technologically advanced occupations, one of which was contributor.

Contributing 2.0

Role models for women in can come in many forms. technology Plenty of impressive women with big ticket tech companies (and many exits) to their name have come from a non-technical background. Here’s just a few of them, many of whom have been granted British Honours for their extensive contributions to the world of technology & innovation —so, no small feat! As a History graduate I find this list extremely inspiring. Don’t let your past define your future — here’s to breaking down those invisible barriers :) Sarah Wood OBE — CEO & Co-Founder, Unruly Degree: BA English Literature, Uni of Cambridge and PhD Early American Literature, UCL Sarah founded Unruly in 2006 which has gone on to be the worlds largest video marketplace, selling to NewsCorp in 2015 for £114m. Sarah remains the Chair and has been awarded an OBE for her contribution to innovation and technology. Jess Butcher MBE — CoFounder, Blippar Degree: MA Ancient & Modern History, Uni of Oxford Jess founded Blippar in 2011 which went on to raise $45 Million in funding with >60M users. She is NED at Azoomee and has been named as one of the ‘Inspiring 50’- most inspiring women in European tech in 2016, BBC’s international list of ‘100 Women’ in 2014, Fortune Magazine’s global ‘Top 10 Most Powerful Women Entrepreneurs’ in 2012 and winner of the Natwest EveryWoman Iris (tech) award in 2013 amongst other accolades. Alex Depledge MBE — CEO of Resi, previously Hassle.com Degree: BA History, Uni of Nottingham and MA International Relations, Uni of Chichago Alex cofounded Hassle.com in 2013 with her business partner Jules after working with Accenture. Hassle sold to Helping for £24m in 2015. TechCity Insider named Alex ‘Entrepreneur of the Year’ 2014. Alex is also on the Board of London Leap & EDITD. She is now running Resi set out to bring transparency to the architecture & homeowner relationship for the better. Priya Lakhani OBE — Founder & CEO of Century Tech Degree: BA Law & Economics, Queen Mary and MA Media Broadcasting Law, UCL Priya founded Century Tech in 2013 — the platform for artificial intelligence, machine learning & data science in education. Priya has been a member of the Secretary of State for Business, Innovation and Skills’ Entrepreneurs’ Forum and also sits on the Board of several educational/skills organisations. Priya was awarded Business Entrepreneur of the Year in 2009, The Mayor of London Fund’s Special Recognition Award 2016 and Officer of the Order of the British Empire in 2014. Debbie Wosskow OBE — Co-Founder, AllBright, ex-CEO LoveHomeSwap Degree: MA Philosophy & Theology, Uni of Oxford Debbie started her first business aged 25 and has two exits behind her including LoveHomeSwap which sold to for $53m and she’s now investing in female founded companies through the AllBright fund. Debbie was the chairman of Sharing Economy UK (SEUK), she sits on the Mayor of London’s Business Advisory Board, and is also a Trustee of the Hampstead Theatre. Wyndham Alice Bentinck MBE — Co-Founder of Entrepreneur First Degree: BSc Management Studies, Uni of Nottingham Alice first started Code First: Girls in 2012 and has since built Entrepreneur First as Co-founder, an institution which over the last 6 years has created over 100 new tech companies across is global cohorts, including Magic Pony and AdBrain. Martha Lane Fox — Chair of Doteveryone.co.uk Degree: BA History, Uni of Oxford Early tech success pioneer cofounding Lastminute.com in 1997, Martha has since gone onto be Founder & Chair of doteveryone.org.uk — making the Internet work for everyone. She is also a Director at Twitter, Chancellor of the Open University and Co-founder of Lucky Voice. She is the UK Gvt Digital Champion and sits on Boards of M&S and mydeco.com. Tamara Lohan MBE — Co-Founder & CTO of Mr & Mrs Smith Degree: BA Spanish & Italian, Uni of Oxford Tamara co-founded the online travel destination for boutique hotels Mrs & Mrs Smith with her husband in 2002 which has over 1.5m customers worldwide. Tamara is also NED at notonthehighst.com and Board Member at TravelTechLab. Louise Broni-Mensah, Founder of Shoobs Degree: BSc Mathematical Economics, Uni of Birmingham Louise founded Shoobs, the marketplace for events and nightlife, in 2013 through the Y Combinator accelerator. She is now based in Silicon Valley but started life in Investment Banking in the UK. Louise was recognised as one of Britain’s most influential Black people under 40 by Powerful Media. Kathryn Parsons MBE— CEO & Founder of Decoded Degree: BA Classics & Classical Languages, Uni of Oxford Founded Decoded in 2011, the tech education which provides technical literacy to corporates and individuals worldwide. Kathryn has also been appointed Chairwoman of the UK Institute of Coding, London Mayor’s Business Advisory Board, UK Government Cyber Security Skills Advisory Board, Tech London Advocates Advisory Board, MBE as part of the New Year Honours List 2016. company Are you interested in getting into tech but don’t have the background? The Bakery is building a pipeline of entrepreneurs to build new tech companies from scratch. We provide that makes the first part a little easer. . a lot of the support Get in touch for more info

Impressive Women in UK Tech Without a Tech Degree

Photo by on Kelly Sikkema Unsplash The news that’s been grinding up the audience development world and spitting it out over the past couple of months has hit with all the surprise and inevitability of a car crash you’ve been watching spinning towards you, without actually considering if you should move or not. Facebook The interview was particularly brutal. Brown came out with some eyebrow-raising comments: Campbell Brown and Adam Mosseri did at Code Media Brown said that Facebook wants to help publishers who want to be on Facebook, but it’s doesn’t necessarily need them. “My job is not to go recruit people from news organizations to put their stuff on Facebook,” she said. “If someone feels that being on Facebook is not good for your business, you shouldn’t be on Facebook.” Her job title seems rather poor now. There was little or no sense of “partnership” in what she was saying. Come to Facebook. Or not. Doesn’t matter . to us In love with a digital sociopath Face it. Facebook never loved us. It had its fun, decided that the relationship was too much work, and has moved on. Oh, sure. It’s still got our number in its phone, for those moments when engagement demands a quick booty call, but we’re history. Any sense that Facebook was our friend and ally in developing a sustainable business model for should be done with. journalism Like any nasty break-up, could it be better for us as an industry to cut ties, rather than try to be friends? This relationship began over a decade ago, and back then we were the big beasts and Facebook was the scrappy little startup starting to challenge the News Corp-owned MySpace. Surely, with News Corp money and real adults in charge, MySpace was set to rule the world? No so fast. Facebook was cleaner, faster and more useable than MySpace, and it was acquiring users at a rapid rate. And this small company, led by a guy barely beyond university age, looked like a great partner for the news business, one that could make the tricky business of navigating the social web a lot easier. I remember the meeting where the slippery slope into Facebook’s clutches began for me clearly. It was a decade ago in one of those dull grey meeting rooms, with a view over suburban London, that either allow bright ideas to flourish against the drab background they provide, or which slowly deaden the spirit. And sometimes both. We started talking about Facebook, which had recently both made itself available to the general public, and introduced the News Feed (yes, there was Facebook before the news feed). Some of my colleagues were very excited by this, but I was less so. They thought that Facebook was a great option for us to use instead of the blogs and communities we’d been building, bringing sharing updates and photos and videos together, away from the messy freeform flow of blogs, Flickr and YouTube. I was less enthusiastic from the start, fearing that what they said was fundamentally correct, but really dangerous. To bring all that freeform creativity of the web under one company, to give it a central gatekeeping role in how we consume media, would be a terrible risk. I used the phrase “training wheels social media” in that meeting, to at least one colleague’s disgust — because that’s what it was. It was a dumbed-down version of the tools we were using to create communities of interest for ourselves, and for our titles. But unlike real training wheels, their purpose was not to allow people to learn and grow, and then move on without them. It was to trap people on the Facebook platform. These weren’t training wheels for learning, these were training wheels of infantilisation, and Facebook used them to suck up people, and trap them in a candy-coloured nursery version of the social web. Once they were happily clicking like and share, Facebook could slowly feed and grow fat on all the yummy, yummy data they were collecting. And many journalists, like my colleagues from that meeting, helped them do it. Over a decade later, and Facebook is the sort of omnipotent, international conglomerate (let’s remember that Instagram, WhatsApp and Oculus Rift are Facebook too) that cyberpunk books and films have been warning us about for decades. They have probably a greater control over the information flow than any organisation in history, and even as the West sours on Facebook, emerging economies where internet penetration is still growing are taking to it with gusto. Imagine if there was just one newspaper worldwide, where the editorial team decided what news the world saw. Would anyone truly be comfortable with that? Yet, that’s exactly the power we’ve handed the Facebook algorithm. That piece of code is influencing people’s perception of world events more than any editor ever has since the dawn of journalism. In less than a decade, the power relationship has shifted, and we have become the supplicants at Facebook’s altar. And Brown and Mosseri came down yesterday to tell us we were sinners, and must repent. For the path to paradise is not through organic reach. Facebook is a fickle and terrible god. Until recently, it has showered us with blessings: traffic to our stories, organically granted, unlike those poor marketing schmucks who had to for reach. It gave us money to make Facebook Lives, and didn’t seem to pay too much attention to how many people were watching, which was just as well, given how few were. It created Instant Articles just for us. pay But we failed to be sufficiently grateful. We abandoned Instant Articles, and let our news crowd out more data-rich social interactions in the feed. We even dared to express doubts about the usefulness of Live. And the we had the temerity to suggest that perhaps Facebook itself was to blame for the spread of intentional misinformation, the so-called “fake news”. Blasphemy! We became more trouble than we were worth, and so the great god Facebook turned the face of its algorithm from us, and we were cast out into the darkness. What so many of us missed is that Facebook was never a , it was a . It existed to grow itself as a business, rather than growing as a business by supporting the creative endeavours of others. And we were right on the cusp of the switch from the tool era to the platform era. Many of the “tools” that were popular then — Flickr, YouTube — were platforms built in a tool-like way. They were both built to allow the content you uploaded there to be shared freely elsewhere, like your own blog. The VC money was flowing away, though, in the direction of platforms. Platforms that were habit-forming. Platforms with lock-in. tool platform And Facebook was the greatest of these. Even as everyone celebrated Twitter as the Arab spring remade world politics, Facebook was growing ever stronger, to the point where it eclipsed the blue bird of terseness. It wasn’t interested in helping you connect, it wanted to be THE place where you connected. And if people were to do that, and give Facebook all the sweet, sweet data that resulted, well, more fool them. dumb enough fucks A building on algorithmic sand cannot stand A VC-backed platform grows by relentless self interest, not by helping others. Once you sign up to platform thinking, you live or die by the whim of that platform. Live by the algorithm, die by the algorithm. The lesson SEO-driven business leaned a decade ago in the fires of Panda were forgotten, and once again people built their businesses in the platform sand. The figures seem small — news would drop from just 5% of newsfeed content to 4%. Your average data illiterate journalist would look at that 1% drop and think it not too bad. But everyone else notes that it’s a 20% drop, and if a majority of your traffic is coming from Facebook, that’s ruinous. Smart businesses, of course, have always had a diversified acquisition strategy. They’re using search and social, and newsletters and chat apps and all sorts of tools to build relationships with their readers. Others haven’t. They have weeks, if not less, to adapt or die. And if your audience team has just learnt how to do Facebook by rote, by reading the “best practice” articles out there and copying an pasting the volume and timing advice, then you’re screwed. For those who have taken the time to understand both the underlying technology and the underlying psychology, they have reason to be cheerful. There’s no reason that 20% drop in traffic needs be equally shared. If you understand how to create content that makes people respond to it — create meaningful engagement in Facebook’s terms — it’s possible you can maintain your traffic, or even grow it. And yes, maybe it’s Groups, and maybe it’s more personal interactions from journalists on Facebook, and maybe it’s a new strategic approach to your Pages. More likely, it’s a mix of them all. But if you have the sort of smart people who can do that, they’re probably already hard at work diversifying your traffic base, with great newsletters, and subtle use of Pinterest and and… growing views on Apple News The greatest trick Facebook played Facebook is not the devil, because the greatest trick the devil ever played was making us think he didn’t exist. Facebook tricked us into thinking nothing else existed, and we fell for it. The Facebook era of journalism is over. We helped Facebook grow. We helped Facebook establish itself as a major destination for video and photos and general reading. But that’s done now, and Facebook is not one for gratitude. And when the presence of misinformation on the platform is perceived as being Facebook’s fault, and brings the ever-growing mega-corp to attention of regulators and legislators, well, it just has to go. that Sure, we could . But what will happen in a couple of years, when Facebook is sucking all local advertising into its cavernous maw? Down will go organic reach, and local news will be back where it started — or worse. help Facebook make itself nice and localised The industry has to start weaning itself off its Facebook habit. Yes, maintain your page. Yes, experiment with Groups. But also start writing journalism that our readers to share — and then leave them to it. Stories can go viral without any intervention from journalists. want Use the time you’ve freed up to investigate other approaches, other platforms — and your own platforms. Maybe you could start paying attention to your loyal commenters again, or building out some focused e-mails, or start a membership schemes, or… Embrace real community. It will embrace you back 10 years ago, before the Facebook distraction, we called this job a “community editor”. And then it became a “social media editor” or an “audience engagement” editor. But maybe we need to come back to that word “community”. Journalism’s is not an abstract art. Journalists do not labour in obscurity, before starving in a garret, with their work only found and celebrated after their death. Journalism is performed for a community, in the moment. The internet is the perfect tool for helping us connect more deeply with that audience, and we showed a profound disrespect for them by outsourcing our community-building to Facebook. Facebook doesn’t care about its users other than as abstract data nodes, as anyone who has ever suffered abuse or harassment on the platform knows. Facebook wants you to conform to it, not the other way around, as anyone who lives under an assumed name or a lifestyle outside the mainstream has experienced. Let’s be better than that. Let’s start caring about our readers, our again. Let’s make them part of what we do, because they’re the ones we do it for. And the first step on that road is to give up Facebook. A loveless relationship is no relationship at all. community

Facebook never loved journalism. It’s time to break up.

Credit: Pixabay ( ) https://pixabay.com/en/node-js-logo-nodejs-javascript-736399/ I started off in my career fairly close to the hardware. I had learned C and Objective-C to work on the iOS platform. I had a lot of control over the software and I had to learn to manage memory and spend hours or days fixing an issue that was caused by my own negligence and oversight (this was pre- ). My career grew with a bias towards iOS and I started learning Java (and more recently Kotlin) to branch to Android. ARC With Android development, memory management was less of an issue due to a forgiving garbage collector. However, I still had a lot of control over the application lifecycle and I was forced to learn the differences between the two major mobile platforms and from interface to functionality that’s a lot to take in. I stayed in my comfortable bubble whilst scoffing at developers — mocking the terrible performance of web apps. I refused to touch JavaScript for a long time and would always reluctantly attempt a more familiar taste such as TypeScript. It’s still JavaScript but at least I didn’t have to type it. I liked my variables and constants to have an explicit type. Years of memory management had me on edge. JavaScript If you own a car, the chances are that you don’t get along with cyclists sharing the road. There is psychology behind this and it’s commonly known as the . It took me a long time to realise that the reason I didn’t like JavaScript developers was because of this psychology. I had put years of effort and pain into learning memory management and application lifecycles. I thought to myself “ ”. free-rider problem JavaScript developers get to share my title without paying the price of memory management or understanding of software lifecycles?! Credit: News Corp Australia As I grew and became less content in my career, I started to have my own ideas. My idea required a web application with a few basic features like access-control and authentication. It needed a few endpoints and resources. I did some research and came across . I thought I might as well give it a shot since it looked fairly easy to setup and was well supported. It took me one evening to get this web application up and running. . This is the stack that I threw together. NodeJS One evening NodeJS (Server) mLab (Mongo DB) PugJS (Template Language) Several calls npm install --save xxx A few hours hacking around and I had a functional web application with a functional authentication system and access control. I didn’t care about Redis, Docker, K8s or any of the good stuff yet. My MVP was built. I could test my product. I had a sandbox database with no backups, I had an unscalable product hosted on Digital Ocean. However, I tested this product with real users and realised it wasn’t as magical as I dreamed. Oh well, nothing wasted, just learned. I started learning more about JavaScript and the unfamiliar package economy it had. Developing on iOS and Android, you are expected to use a few common dependencies like and . I expected a similar level of dependencies in JavaScript. I was wrong. I soon realised why it was called the world wide web. It was literally a huge web of dependencies. This was a scary experience for me. Alamofire Retrofit so Credit: Pixabay ( ) https://pixabay.com/en/spider-silk-bokeh-rain-arachnid-1287407/ However, after learning a few rules such as using a with polyfills, the most common dependencies and the general ecosystem — I started to get much more comfortable. I was able to create an MVP of any product in a couple of days. That’s an impressive turn-around and it’s all thanks to NodeJS. transformer Once an MVP worked out, it was time to invest time elsewhere. I started to branch out and care about scaling for my specific needs. . NodeJS is great for some uses but like any other language, it’s not to be used just because you’re familiar with it. NodeJS is not a silver bullet As for the language itself, I used to be disgusted with JavaScript. The core reason for this is that I used to use ActionScript 3.0 a while back. Both JS and AS3 were ECMAScript so why was AS3 so much nicer than JavaScript? Compatibility. It took me a while to realise how important browser compatibility was for JS. Once Babel started to be adopted, JavaScript started changing fast. It made some mega improvements. I actually enjoy writing JavaScript now whilst being assured that legacy code is being created under the hood. This post is primarily targeted at people who scoff at JavaScript. I still don’t like the idea of “native” JS applications. However, I understand just how much flexibility this gives to desktop/mobile apps and it’d be hypocritical of me to be sat here on VSCode whilst slating them. I do like the idea of hybrid applications that have some core components written in a low-level language which is an approach a lot of intensive desktop web-apps are doing. If you see a developer that hates JavaScript, it’s probably because they perceive you as a cyclist taking over a space they believe they’re entitled to. If you liked this article, please clap and follow me to read my future content.

I used to hate JavaScript. Now I like it.

By . James Alctucher Originally published on JamesAltucher.com I was the worst pizza delivery guy. Fraternity guys would chase after me as I was peeling out of their driveways after a delivery. Why? The sauce and cheese fell all to one side. I couldn’t help it. I also never got tips. Wende, my partner in our restaurant delivery business, always got tips. But she was beautiful, blonde, great smile, had personality, etc. And I secretly loved her. I couldn’t compete. We also started a debit card for college kids. From the first day we were open for business we had college kids signing up for our card (there were no credit cards for kids then). And anyone who had our debit card could order food from the 20 or so restaurants in town and we’d deliver, but with a 25% markup. I loved delivering food because it gave me twenty, or even forty minute breaks from my girlfriend. We were having troubles at the time. I’d sometimes stop the car between deliveries and just read. I was a screwed up 19 year old then. Now I’m only a mildly-screwed up 43 year old. I’ve had seven startups since then. And some profitable exits. And another 20 or so that I’ve funded. When I think “entrepreneur” I think or or . I don’t usually think of myself. In part because I feel shame that after all of these startups I don’t have a billion dollars. Many startups fail. But I’ve had a few successes as well. Successes in a startup makes Mark Cuban Larry Page Steve Jobs you feel immortal. I was going to make this post: “the 12 rules to being a good entrepreneur” and I outlined the 12 rules that have consistently worked for me. But rule #1 is taking up 1500 words already. Tim Sykes tells me I need to break these posts up more. So this one rule is going to take up the whole post. But, for me, this is the most important rule. This is a corollary of the phrase, “ideas are a dime a dozen”. The MOST IMPORTANT RULE: Have a customer before you start your business. There is another corollary: lazy is best. If you have to work for two years before one dollar of revenue comes into business then thats too much work. I’m lazy so I like money coming in with as little work as possible. Mark Zuckerberg, of course, is different. He put in years of work before dollar one of revenue came in. But we’re different people. In about twenty minutes I’m going to go to the local café here, The Foundry, and bring a pad, order a coffee and muffin, and write down ideas for businesses. Then I’ll probably throw the piece of paper out. **Because ideas are useless.**They are just practice to keep your idea muscle in shape. People say, “Execution is important”. That’s not really true either. It’s a commodity. The only thing that’s important is money. You get money by having a customer. You get a customer by satisfying a need that’s so important to them they would be willing to pay for it. If you have a customer that’s willing to pay you money, then execution becomes a lot easier. Why make it harder for yourself? FAKE RULE: Execution is useless. Life as an entrepreneur is hard. I like stability as much as I like taking risks. So for me, I need a customer. It’s a matter of how much risk you want to take. and jump into the abyss. If a customer happens to be waiting for you in the abyss then you won’t be lonely there. Loneliness is bad for a startup. In an earlier post I suggest reasons why people need to quit their jobs Example: How Stockpickr Started Tom Clarke, the CEO of thestreet.com called me up in mid 2006. He wanted to meet and brainstorm ideas with me. So I had about two weeks to prepare. I called up a development firm in India. MySpace had just been acquired by NewsCorp so I sketched out what I considered the “MySpace of Finance”. I threw in every idea from my own trading. In other words, I wanted to create a site that I would use as a professional trader and so I knew other traders would benefit from it. And, I had a theory about making a quality financial site that basically had no news in it. I’m going to be blunt: 99% of financial news is useless and misinformed and misleading, if not outright lying. My ideas for the site were purely based on my own ten years experience as a professional trader. In fact, I had just turned down working for a multi-billion dollar hedge fund based on a specific strategy I had. Instead, I implemented that strategy within stockpickr.com. Within a week, (because I told the company in India that I would be building the site with them if they sent back good screenshots, which was true), the company sent me back screenshots. I met with Tom and told him, “I’m almost done with the site. Here it is.” (I exaggerated) And I showed him the screenshots. I had set up meetings with Yahoo and AOL as well to discuss them so it wasn’t a stretch to say, “I’m also talking to Yahoo and AOL.” Nobody wants to be the first customer, so you have to create the aura of many customers. for free And so he said, “why are you talking with them? You’ve been with us forever. Lets do this together.” So we negotiated right there. I said, “great, how about you guys take 10% of the company and put all your extra ads on all of our pages and let me link from every article back to Stockpickr.com (the name of the company).” He said, “I thought we were partners. Lets do it 50–50.” So right away, Most people I spoke to thought this was a horrible idea. In fact, one of my employees quit because I did this: but giving 50% of your company away is often better than giving 10% of your company away. When you give 50% of the company away, your partner is obligated to follow through and be a real partner. He can never forget about you. Its also always a good thing when the most popular person in financial media, was also a 50% partner in my business since he was the founder of thestreet.com. If you can get the top person in your field to take equity, you’re golden from day one. Again, its all about making life easy. With family responsibilities, health, life in general, why make things even harder for yourself? I had given up 50% of the company. Jim Cramer, Its like that saying when you owe the bank a million, they own you. When you owe the bank a billion, you own them. Three years later I watched another company thestreet.com only took 10% of almost disappear because Thestreet.com was not obligated to follow through. So, at that point, without even having a site finished : I knew I had three things going for me: (or even started) A) I was going to get traffic. I could write three or four articles a day and link each one back to stockpickr all over the article. The street.com got 100mm pageviews a month so I knew I would get some percentage of that. B) I was going to make money. If I got even 3 million pageviews a month and the average CPM of thestreet.com (according to their SEC filings) was $17, I would make about $50 thousand a month with expenses nearing zero. What if I put 3 ads a page on the site? Then I would make even more. Slap a 40x multiple on a growing company and before I even started I had real value. C) Profitability and growth from day one meant I could put the company up for sale almost immediately. But that’s another story. With my first successful company, Reset, I had about 10 paying customers before I finally made the jump to running the company fulltime: HBO, Interscope, BMG, New Line Cinema, and Warner Brothers were all paying customers before I jumped ship from HBO to run Reset fulltime. (the Wu-Tang Clan was a paying customer) When I started my fund of hedge funds I didn’t put one dime into the expense of setting it up until I had the first $20 million commitment. $20 million with a 1.5% management fee meant an instant $300 thousand in revenues, plus extra money for legal fees, etc. Enough to pay a salary or two. It took a year of cultivating my network before I had that commitment, but it worked. This is just me, personally. Some people don’t mind starting a company without any customers. I’m too conservative for this. I’m happy to even give up equity to get that first customer. 50% of a profitable company is better than 100% of a company that will probably quickly go out of business. How do you get that first customer? – ? List the 20 CEOs or high level executives you would like to meet. If you have a rolodex, great. If you don’t, then you might have to write to 40 people. This is why its important to and use some of the other ideas mentioned throughout this blog. But its ok if your rolodex is cold. Many successful businesses I’ve been involved with started with cold emails. Who Exploit your Employer – . Develop 20 ideas for each person. Ideas Get your idea muscle in shape first. – . Write them all, giving at least 10 of the ideas, in detail, and how you would implement them. Sometimes you have to dig for their email address. Communicate Like I did here. – . All this does is get you the meeting. Once you’re in the door, the conversation can go anywhere. Of the 20 people you write, rule of the universe is you’ll get six meetings. Meet – . In the meetings ask the question, “ ” Remember, execution is a commodity. Because of globalization you can build anything for cheap. I built the first working version of Stockpickr.com for $3000 in Bangalore. Throw in another $150 from a design made in Siberia. You need zero skillset for that. Just a good idea muscle, an ability to sell, and modest ability to manage a project. Ask what one product can I build for you that you will definitely buy? – . Never. If someone says, “can you do this?” Yes. “But can you do this?” Yes. Can you do it for this? Yes. Never say No – . Nice to meet you. Here’s my sketch of what you want. Is this right? Should I start today? Follow up every day – If necessary, give up equity for that first customer. Or first two customers. Equity. – . If more than one CEO wants the same product or service, and the price is right, then now you have a business. Build the product, sell it, and you’re in shape. Done – . You’re not really “Done”. Every day you have to go back and see if your customer is achieving more success BECAUSE of your products. If he is, then ask him what else he needs and then build it. If he isn’t, then ask him what else he needs and builds it. Your easiest new sales will be with your old customers. Repeat This has worked for me on three different occasions. And each time resulted in great profits for me. By the way, this technique has also worked for people who have contacted me with their own ideas. You’ve been hypnotized. You’ve been told you need a corporate job. You need a college degree. You need stability. You need the white picket fence. You need the IRA and the health insurance. Snap your fingers in front of your face. The American Religion is a myth, just like the movie, Thor, is based on a myth. Stability is only in your mind. There’s $15 trillion dollars in our economy, recession or no recession. Its falling like snow. Reach out with your tongue and taste it. Listen. ➤ for my newsletter. You’ll get a daily email of my latest content! Sign up ➤ Or visit my to read my FREE blog content website ➤ Do you have a question? Ask me on …I answer questions daily! Quora James Altucher is the author of the bestselling book Choose Yourself, editor at The Altucher Report and host of the popular podcast, The James Altucher Show , which takes you beyond business and entrepreneurship by exploring what it means to be human and achieve well-being in a world that is increasingly complicated.

The MOST IMPORTANT RULE: Have a customer before you start your business.

On September 27th, in partnership with , we held an hour long conference call with four thought leaders on SEC complaint ICOs. The speakers were Mike Jones, CEO of , Emma Channing, General Counsel of Lee Schneider, Counsel at , and Tom Kineshanko, Co-Founder of Geektime Science Argon Group , Debevoise & Plimpton Protos Cryptocurrency Asset Management Each speaker shared their perspective on the process of, and the risks involved in, executing and investing in SEC complaint ICOs. After the speakers shared their prepared remarks, the call was opened for questions from the audience. A video replay of the call is below (only the last speaker had slides). The big take away from the call for me was how much uncertainty and risk there remains in trying to execute an SEC compliant ICO. That said, there were lots of thoughtful insights shared on the call. The chart below, which identifies the three types of tokens, and their attributes, can be a helpful reference given the discussions on the call. For purposes of this post, we’ll use the term Security Token instead of Crypto-Securities, as that was the term used on the call. Below are what I believe to be the seven most important thoughts shared on the call: . Emma was emphatic about the incredible amount of detail and guidance the SEC gave in the report, without it being an enforcement action. Emma stated that the SEC has a very sophisticated team looking at ICOs, and that the ICO community should take advantage of the fact that the SEC is happy to take emails or phone calls with ICO related questions. The SEC DAO Report was a remarkable gift to the ICO industry Emma says she looks to for guidance on Security Token White Papers, but there is no similar reference for a Utility Token. Is 20–30 pages sufficient, or should it be 70–80 pages long like offering memorandums? Should the Utility Token White Paper include risk factors, use cases, or use of proceeds? Should it have audited financials or management financials? Emma believes we don’t have answers to ANY of these questions. So there is risk that a regulator in five years could say that a certain piece of information that wasn’t in the White Paper was material, and sue an issuer on the basis of contract fraud misstatement. There’s been no guidance on how long a White Paper for an SEC Compliant Utility token should be, or what should be in it. S-1s Lee said there is a range of acceptable ways to codify the use of proceeds. The Use of Proceeds can be given in great detail or at a higher level. The use of proceeds can be linked to milestones, with amounts targeted to spend once those milestones are reached. While there is no “right” way to explain the use of proceeds, it’s critical to be thoughtful. In addition, issuers should have contingency plans in case they raise more money than expected. Lee Believes That the Use of Proceeds Section is the Most Important Part of a White Paper. Lee pointed out that all developed countries have securities laws and regulations that should be considered. ICO issuers then have to decide in which countries they’ll be regulatory compliant, as they can’t be compliant everywhere due to cost and time constraints. Generally issuers care about the countries where they can sell the most tokens, and take a risk in other countries. Emma said she’s often asked why the industry can’t just evolve away from regulation. She explains that there are laws that apply when you cross the road, let alone when you issue ICOs. So while some people say she’s scaring people, she believes she’s educating people. Solving For The SEC/U.S. Regulators Does Not Solve For The Rest of the World. Mike stated that ssuing ICOs has been more complicated than his past raises where his LPs were U.S. investors. Opening a crowd sale for people outside the U.S. and working within the legal constraints of different countries, is not only more complicated, it’s about twice the price of a traditional raise. In addition, in a traditional raise, the movement of shares is controlled. But in a token sale, where the tokens can move around the world, there are new things to think about like information control and how information is distributed. Designing the token economy adds another layer of complexity. For Mike, the Science ICO has been a four month process. . SEC Compliant ICOs Are More Complicated and Costly Than Traditional Fund Raising. i Blockchain Before Science**,** Mike worked at News Corp and was tasked with seeing if they could leverage MySpace to better battle Facebook. During that period, Mike spent a lot of time with people from MySpace and Facebook. There was a moment when Zuckerberg clearly articulated the value of social as it related to software. Once Mike got his mind around that, every piece of software he looked at, he looked at thru the social lens, and saw how the software could be more powerful. Mike said the moment he got t , everything he sees now, is through that Decentralization lens, and he sees opportunity everywhere. Mike Jones now looks at everything through a Decentralization lens, and sees opportunities everywhere. he concept of decentralized software Tom pointed out that traditional hedge funds may take a shorter term view on their investments than might otherwise be optimal, because of the risk of redemptions. Traditional funds can sometimes have 15 or 30 day redemptions (when investors ask for their money back). Redemptions can force the fund to sell assets to meet the redemption. Sometimes redemptions can cascade, as forced asset sales can happen at sub optimal prices, which can cause further redemptions, or cascading redemptions, resulting in the demise of a fund (see 2008). In a tokenized fund, investors who want out simply sell their tokens to other investors, leaving the pool of capital stable, and leaving the Token Fund manager able to take a longer term view. A Major Advantage of a Tokenized Hedge Fund Is A More Stable Pool of Capital Due to the Elimination of Redemptions. There were numerous other things mentioned on the call that were quite meaningful. Emma suggested that every ICO should be taking addresses and checking IDs for purposes. Emma also mentioned Argon’s with Overstock and RenGen. Lee talked about the importance of communications post ICO. Mike talked about the structure of his Science Blockchain incubator and how some exits will lead to tokens being burned. It was an hour well spent. KYC new ATS (Alternative Trading System) venture 0.00000001 If you got at least Bitcoin worth of value from this post please “Clap” below so others will see the post.

The 7 Things You Need To Know From The SEC Compliant ICO Conference Call

Court documents reveal how the late Apple CEO won over News Corp. on eBook pricing. The US government’s price-fixing lawsuit against Apple goes to trial next month in New York. Ahead of its court date, that purport to show Apple was the “ ” in a scheme to set artificially high ebook prices with some of the largest American publishers, which have already settled the case. the US released emails ringleader — This story helped me to negotiate hard at , I hope it will help you the same. Waydev The emails have mostly been viewed in the context of the lawsuit, but they also provide an extraordinary view of high-stakes negotiation between the leaders of two powerful firms, Apple and News Corp. They start far apart, but over the course of five days, Apple’s then-CEO Steve Jobs successfully pulls the son of News Corp. CEO Rupert Murdoch over to his side. Jobs was a famously hard-nosed negotiator who won these kinds of battles all the time. Before book publishers, there was the movie industry. And before that, music record labels. But most of those negotiations were hidden from view. What follows are the emails released last week along with some context; spelling and grammar have been preserved from the originals. News Corp.’s Opening Move It was a Friday morning, January 22, 2010. Apple was preparing to release its newest product, a long-rumored tablet computer, the following week. Part of the iPad’s appeal was supposed to be the vast array of media that could be consumed on it, but one of the largest American publishers, HarperCollins, was holding out from signing a deal to sell its ebooks in Apple’s iTunes store. Those were the stakes as Eddy Cue, Apple’s head of iTunes and the App Store, visited executives of HarperCollins and its parent company, News Corp. The substance of that meeting was conveyed in an email sent to Cue later that day by Brian Murray, the CEO of HarperCollins. It detailed the publisher’s opening bid in the negotiation, with five days remaining until the iPad’s unveiling. Eddy, Thanks for coming in again this morning. We’ve talked over the proposal and I want to make sure that you have a summary of the deal that HarperCollins would be willing to do in your timeframe. 1. Pricing: We need flexibility to price on a title by title basis outside the prescribed tiers in the contract. We will use our best efforts to meet the tiers we discussed. 2. MFN [“most favored nation” status]: In the event that HarperCollins and Apple disagree on a consumer price for a title, HarperCollins needs the ability to make that title available through other agents who support the higher price. 3. Commissions: We need a lower commission on new releases for the economics to work for us and our authors. We believe a 30% commission will lead to more authors asking for ebooks to be delayed a result that will not work for Apple or HarperCollins. 4. The new release window: We need to have flexibility on the agency window. We believe this window should be 6 months rather than 12 months in the event that one or more large retailers do not move to an agency model. Leslie will be sending Kevin a contract that reflects these points in the event you wish to move forward on these terms. ThanksBrian Those terms were never going to fly with Apple, which had successfully signed deals with HarperCollins’s rivals, like Penguin (a division of Pearson) and Simon & Schuster (part of Viacom). Those deals would allow Apple to set prices for new ebooks at $12.99 — three dollars higher than the typical rate at that time on Amazon — and take a 30% cut of each sale. But HarperCollins wanted the freedom to set its own prices and worried that $12.99 per ebook would hurt its sales on the new iPad as well as the Kindle. It also didn’t want to give up 30%. To back up its position, James Murdoch, a high-ranking News Corp. executive, forwarded Murray’s email to Apple’s then-CEO Steve Jobs, and included the following note. It was still Friday. Steve, Thanks for your call earlier today, and for the time last week. I spoke to Brian Murray and Jon Miller [then the head of digital media at News Corp.] — and Brian is sending a note to Eddy today. I thin I have a handle on this now. In short — we we would like to be able to get something done with Apple — but there are legitimate concerns. The economics are simple enough. [Amazon] Kindle pays us a wholesale price of $13 and sells it for 9.99. An author gets $4.20 on the sale of a hardcover and $3.30 on the sale of the e-book on the Kindle. [A portion of this email was redacted by the court.] Basically — the entire hypothetical benefit of a book without raw materials and distribution cost accrues to Apple, not to the publisher or to the creator of the work. The other big issue is one of holdbacks. If we can’t agree on the fair price for a book, your team’s proposal restricts us from making that book available elsewhere, even at a higher price. This is just a bridge too far for us. Also, we are worried about setting prices to high — lots of ebooks are $9.99. A new release window with a lower commission (say 10[%]) for the first six months would enable us to proce much more kenly for Apple customers. We’d like to da that. More on this below in Brian’s note to Eddy. We outline a deal we can do. Feel free to call or write anytime over the weekend to discuss if you like. I am in the UK (so eight hours ahead of CA). My home number is [redacted]. I check the email regularly. Steve, make no mistake that across the board (TV, Studios, Books, and Newspapers) we would much rather be working with apple than not. But we, and our partners who produce, write, edit, and otherwise make all this with us, have views on fair pricing, and care a lot about our future flexibility. I hope we can figure out a way, if not now and in time for this launch of yours, then maybe in the future. Best,JRM Jobs Digs In Jobs wasn’t willing to compromise. He sent this reply to Murdoch the same day, arguing that Amazon’s pricing wasn’t sustainable and would train people to think that ebooks were cheap. Jobs also reminded Murdoch of Apple’s vast reach — “over 120 million customers with credit cards on file.” You need us more than we need you, he seemed to be saying. James, A few thoughts to consider (I’d appreciate it if we can keep this between you and me): 1. The current business model of companies like Amazon distributing ebooks below cost or without making a reasonable profit isn’t sustainable for long. As ebooks become a larger business, distributors will need to make at least a small profit, and you will want this too so that they invest in the future of the business with infrastructure, marketing, etc. 2. All the major publishers tell us that Amazon’s $9.99 price for new releases is eroding the value perception of their products in customer’s minds, and they do not want this practice to continue for new releases. 3. Apple is proposing to give the cost benefits of a book without raw materials, distribution, remaindering, cost of capital, bad debt, etc., to the customer, not Apple. This is why a new release would be priced at $12.99, say, instead of $16.99 or even higher. Apple doesn’t want to make more than the slim profit margin it makes distributing music, movies, etc. 4. $9 per new release should represent a gross margin neutral business model for the publishers. We are not asking them to make any less money. As for the artists, giving them the same amount of royalty as they make today, leaving the publisher with the same profits, is as easy as sending them all a letter telling them that you are paying them a higher percentage for ebooks. They won’t be sad. 5. Analysts estimate that Amazon has sold more than one million Kindles in 18+ months (Amazon has never said). We will sell more of our new devices than all of the Kindles ever sold during the first few weeks they are on sale. If you stick with just Amazon, Sony, etc., you will likely be sitting on the sidelines of the mainstream ebook revolution. 6. Customers will demand an end-to-end solution, meaning an online bookstore that carries the books, handles the transactions with their credit cards, and delivers the books seamlessly to their device. So far, there are only two companies who have demonstrated online stores with significant transaction volume — Apple and Amazon. Apple’s iTunes Store and App Store have over 120 million customers with credit cards on file and have downloaded over 12 billion products. This is the type of online assets that will be required to scale the ebook business into something that matters to the publishers. So, yes, getting around $9 per new release is less than the $12.50 or so that Amazon is currently paying. But the current situation is not sustainable and not a strong foundation upon which to build an ebook business. [A portion of this email was redacted by the court.] Apple is the only other company currently capable of making a serious impact, and we have 4 of the 6 big publishers signed up already. Once we open things up for the second tier of publishers, we will have plenty of books to offer. We’d love to have HC among them. Thanks for listening. Steve Murdoch Starts to Bend Murdoch’s reply came the following afternoon, Saturday, January 23. Jobs had made clear that Apple wouldn’t budge. Murdoch was about to indicate that HarperCollins would. He proposed two possible compromises, then noted that News Corp. and Apple were negotiating on a number of fronts. “Is it worth considering in the round, over the next few months or weeks, whether or not some of these loose ends can be tidied up?” Murdoch wrote. The iPad announcement was four days away. Steve, I think the crux of this is our flexibility to offer product elsewhere at price-points you don’t like. If we could offer to you that a certain percentage of releases (>50%) would be available within your pricing structure (< or = 14.99), does that give you enough comfort? I think we are worried more about the absolute holdback of product elsewhere, and our ceding of pricing to Apple, than we are about the actual haggle over what the price will be. I haven’t shared this with HC directly — so this is only hypothetical. But if you were willing to accept that a supplier can exploit other avenues (at prices not disadvantageous to you), with a guarantee of substantial volume through Apple — maybe I could work with HC to get to some common ground. Please let me know. A different question: we have four areas of discussion (related to our product) between our teams right now: Books, US Video, Int’l Video, and newspapers. All at different stages of maturity, these discussions are all centered, for us, around the desire to make our product widely available, and to make yours and our products more attractive for our customers. It seems though that we in each one we largely encounter a “take it or leave it” set of terms, and predictably we’ve so far failed to really strike the kind of partnerships that could move things forward. Is it worth considering in the round, over the next few months or weeks, whether or not some of these loose ends can be tidied up? It’s clear that Apple is already becoming an attractive platform for so many of our customers — all over the world. As a creative company at our core, NWS [News Corp.] should be more engaged with Apple, and I think Apple could be more engaged with NWS, globally, than either of us are today. Best,JRM Jobs Goes in for the Kill With Murdoch indicating that HarperCollins was willing to compromise, Jobs pressed harder. “As I see it, HC has the following choices,” he wrote in a reply to Murdoch the following morning, Sunday, January 24. Jobs outlined three stark choices. Accept our terms, he was saying, or good luck with Amazon. “Maybe I’m missing something, but I don’t see any other alternatives,” Jobs wrote, almost daring Murdoch to spurn Apple. “Do you?” James, Our proposal does set the upper limit for ebook retail pricing based on the hardcover price of each book. The reason we are doing this is that, with our experience selling a lot of content online, we simply don’t think the ebook market can be successful with pricing higher than $12.99 or $14.99. Heck, Amazon is selling these books at $9.99, and who knows, maybe they are right and we will fail even at $12.99. But we’re willing to try at the prices we’ve proposed. We are not willing to try at higher prices because we are pretty sure we’ll all fail. As I see it, HC has the following choices: 1. Throw in with Apple and see if we can all make a go of this to create a real mainstream ebooks market at $12.99 and $14.99. 2. Keep going with Amazon at $9.99. You will make a bit more money in the short term, but in the medium term Amazon will tell you they will be paying you 70% of $9.99. They have shareholders too. 3. Hold back your books from Amazon. Without a way for customers to buy your ebooks, they will steal them. This will be the start of piracy and once started there will be no stopping it. Trust me, I’ve seen this happen with my own eyes. Maybe I’m missing something, but I don’t see any other alternatives. Do you? Regards,Steve On Tuesday, a day before the iPad announcement, HarperCollins agreed to Apple’s terms. The publisher’s ebooks were included in the iBookstore unveiled on January 27 along with new tablet, more than 100 million of which have now been sold. Original article: http://www.theatlantic.com/business/archive/2013/05/the-steve-jobs-emails-that-show-how-to-win-a-hard-nosed-negotiation/276136/

How Steve Jobs negotiated with James Murdoch

Building Your Own Programming Language From Scratch: Part X - Exceptions Handling

Building Your Own Programming Language From Scratch: Part VII - Classes

Construyendo su propio lenguaje de programación desde cero: Parte VII - Clases

Construire votre propre langage de programmation à partir de zéro : partie VII - Classes

स्क्रैच से अपनी खुद की प्रोग्रामिंग लैंग्वेज बनाना: भाग VII - कक्षाएं

ゼロから独自のプログラミング言語を構築する: パート VII - クラス

Construindo sua própria linguagem de programação do zero: Parte VII - Aulas

Xây dựng ngôn ngữ lập trình của riêng bạn từ đầu: Phần VII - Các lớp học

从头开始构建您自己的编程语言：第七部分 - 类

In this part of creating your programming language, we will implement loops. Please see the previous parts: Building Your Own Programming Language From Scratch Building Your Own Programming Language From Scratch: Part II - Dijkstra's Two-Stack Algorithm Build Your Own Programming Language Part III: Improving Lexical Analysis with Regex Lookaheads Building Your Own Programming Language From Scratch Part IV: Implementing Functions Building Your Own Programming Language From Scratch: Part V - Arrays The full source code is available over on . GitHub Our language will provide the following types of loops: loop - to iterate through a range: For loop - to iterate as long as a specified condition is true: While loop - to iterate through elements of arrays and structures: For-each (Iterable) 1 Lexical analysis In the first section, we will cover lexical analysis. It’s a process to divide the source code into language lexemes, such as keyword, identifier/variable, operator, etc. To make our lexical analysis work with loop constructions, we add the following lexemes: , , to the TokenType enum: Keyword loop in by package org.example.toylanguage.token;

...
public enum TokenType {
    ...
        Keyword("(if|elif|else|then|end|print|input|struct|fun|return|loop|in|by)(?=\\s|$)"),
    ...
} Then, in order to separate the lower and upper bounds of the For loop range, we add the two-dot lexeme: GroupDivider [.]{2} ...
public enum TokenType {
    ...
    GroupDivider("(\\[|\\]|\\,|\\{|}|[.]{2})"),
    ...
} And finally, if we don’t want the lexeme to catch the two-dot lexeme declared after the loop’s lower bound, we add a negative look ahead construction before the dot expression: Numeric GroupDivider (?![.]{2}) ...
public enum TokenType {
    ...
    Numeric("([-]?(?=[.]?[0-9])[0-9]*(?![.]{2})[.]?[0-9]*)"),
    ...
} 2 Syntax analysis Let’s move to the next section - the syntax analysis, which will convert the lexemes received from the previous step into the final statements following our language rules. As was said at the beginning, our language will support three types of loops: For, While and Iterable. Therefore, first we will declare the class, which will contain common execution process for all the loop types: AbstractLoopStatement package org.example.toylanguage.statement.loop;

public abstract class AbstractLoopStatement extends CompositeStatement {
    protected abstract void init();

    protected abstract boolean hasNext();

    protected abstract void preIncrement();

    protected abstract void postIncrement();
} Inside this class, we declare the following abstract methods, which can have different implementation for each loop type: - it will be invoked to perform initialization operations, like initializing a counter variable when a loop starts the execution. init() - It will be called before each iteration to define that we still have remaining items to iterate. hasNext() - It’s a prefix operation that will be called before each iteration. preIncrement() - It’s a postfix operation that will be called at the end of each iteration. postIncrement() Before creating the loop implementations, we complete the method by utilizing declared abstract methods. To start the execution, we invoke the method to initialize the loop state, but before that, we should open the new MemoryScope for the variables declared within the loop, as they shouldn’t be accessed after the loop ends. In the end of the execution, we release the earlier opened MemoryScope: Statement#execute() init() MemoryContext.newScope();

try {
	init();
} finally {
	MemoryContext.endScope();
} Next, to iterate a loop, we utilize the abstract method. Inside each iteration, we invoke the loop’s inner statements, defined within the inherited CompositeStatement class. When the iteration starts, we invoke the method to perform the prefix increment operations. At the end of each iteration, we call the method to perform the postfix increment operations accordingly: hasNext() preIncrement() postIncrement() while (hasNext()) {
	preIncrement();

	try {

		// execute inner statements
		for (Statement statement : getStatements2Execute()) {
			statement.execute();
		}
	} finally {
		postIncrement();
	}
} Next, to make sure that variables declared inside the loop statements are isolated from each iteration, we open the inner MemoryScope during the statements' execution. And finally, if our loop is defined inside the function and one of the loop’s inner statements contains a return statement, we should stop the iteration immediately because it’s a sign to exit a function. Therefore, we validate that the function's return statement hasn’t been called after each statement execution. The complete method: AbstractLoopStatement#execute() public abstract class AbstractLoopStatement extends CompositeStatement {
    ...

    @Override
    public void execute() {
        MemoryContext.newScope();
        try {

            init();

            while (hasNext()) {
                preIncrement();

                MemoryContext.newScope();

                try {

                    // execute inner statements
                    for (Statement statement : getStatements2Execute()) {
                        statement.execute();

                        if (ReturnContext.getScope().isInvoked())
                            return;

                    }
                } finally {
                    MemoryContext.endScope(); // release each iteration memory

                    
                    postIncrement();
                }

            }
        } finally {
            MemoryContext.endScope(); // release loop memory
        }
    }
} 2.1 For loop Let’s begin with the first For loop implementation. By our language design, it will contain the variable expression, the lower bound, the upper bound, and the step expression: package org.example.toylanguage.statement.loop;

@RequiredArgsConstructor
public class ForLoopStatement extends AbstractLoopStatement {
    private final VariableExpression variable;
    private final Expression lowerBound;
    private final Expression uppedBound;
    private final Expression step;
} Next, we implement the abstract methods. Within the first method, we assign the lower bound to the variable by evaluating an instance of AssignmentOperator, which accepts the expression as a first operand and the expression as a second operand: init() variable lowerBound ...
public class ForLoopStatement extends AbstractLoopStatement {
	...

	@Override
	protected void init() {
		new AssignmentOperator(variable, lowerBound)
				.evaluate();
	}
} Next, we override the . It should return a boolean value if we remaining have items. To make a compare operation, we create an instance of the LessThenOperator by passing the expression as a first operand, and the as a second operand. As a result, we expect to get the LogicalValue with boolean value inside it: hasNext() variable upperBound ...
public class ForLoopStatement extends AbstractLoopStatement {
    ...

	@Override
	protected boolean hasNext() {
		LessThanOperator hasNext = new LessThanOperator(variable, uppedBound);
		Value<?> value = hasNext.evaluate();
		return value instanceof LogicalValue && ((LogicalValue) value).getValue();
	}
} Lastly, to perform the increment operations for each iteration, we need to implement the and methods. The For loop should use the step increment only at the end of each iteration. Therefore, we perform the increment only within the method. To implement the increment operation, we create an instance of the AdditionOperator, which will sum up the value with the value, and assign the result to the expression again using an instance of the AssignmentOperator. If the step expression hasn’t been provided, we accumulate the NumericValue instance with the default value equal to : preIncrement() postIncrement() postIncrement() variable step variable 1.0 ...
public class ForLoopStatement extends AbstractLoopStatement {
	...

	@Override
	protected void preIncrement() {
	}

	@Override
	protected void postIncrement() {
		AdditionOperator stepOperator = new AdditionOperator(variable, Objects.requireNonNullElseGet(step, () -> new NumericValue(1.0)));
		new AssignmentOperator(variable, stepOperator)
				.evaluate();
	}
} 2.2 While loop The next loop implementation is the While loop. It’s more straightforward than the For loop, as we don’t need to perform addition and assignment operations at all. This operator will contain only the expression, which will only serve to calculate the result. The other overridden methods will be empty: hasNext hasNext() package org.example.toylanguage.statement.loop;

@RequiredArgsConstructor
public class WhileLoopStatement extends AbstractLoopStatement {
    private final Expression hasNext;

    @Override
    protected void init() {
    }

    @Override
    protected boolean hasNext() {
        Value<?> value = hasNext.evaluate();
        return value instanceof LogicalValue && ((LogicalValue) value).getValue();
    }

    @Override
    protected void preIncrement() {
    }

    @Override
    protected void postIncrement() {
    }
} 2.3 Iterable loop The last loop implementation is the Iterable loop. It will allow us to perform the for-each iterations with arrays and structures. First, we define the new Value implementation, which will override the interface and, accordingly, will provide us with the implementation: java.util.Iterable java.util.Iterator package org.example.toylanguage.expression.value;

public abstract class IterableValue<T> extends Value<T> implements Iterable<Value<?>> {
    public IterableValue(T value) {
        super(value);
    }
} Next, we extend the ArrayValue by implementing IterableValue and passing array values to the Iterator instance: package org.example.toylanguage.expression.value;

public class ArrayValue extends IterableValue<List<Value<?>>> {
    ...

    @Override
    public Iterator<Value<?>> iterator() {
        return getValue().iterator();
    }
} And we do the same extension for StructureValue as well. For now, we will iterate only the structure’s values. Later, we can introduce the dictionary data type to operate with key-value types: package org.example.toylanguage.expression.value;

public class StructureValue extends IterableValue<Map<String, Value<?>>> {
    ...

    @Override
    public Iterator<Value<?>> iterator() {
        return getValue().values().iterator();
    }
} Now we have the two IterableValue implementations which can provide us with the Iterator instance, we can complete the Iterable loop with the corresponding implementation. It will contain the expression and the expression. Also, we declare an auxiliary non-final field that will be used to iterate the loop: variable iterable Iterator<Value<?> package org.example.toylanguage.expression.value;

@RequiredArgsConstructor
public class IterableLoopStatement extends AbstractLoopStatement {
    private final VariableExpression variable;
    private final Expression iterableExpression;

    private Iterator<Value<?>> iterator;
} Next, we override and resolve the abstract methods. Inside the method, we initialize the field using the IterableValue’s method: init() iterator iterator() ...
public class IterableLoopStatement extends AbstractLoopStatement {
    ...

	@Override
	protected void init() {
		Value<?> value = iterableExpression.evaluate();
		if (!(value instanceof IterableValue))
			throw new ExecutionException(String.format("Unable to loop non IterableValue `%s`", value));
		this.iterator = ((IterableValue<?>) value).iterator();
	}
} The method will utilize the implementation. Lastly, we should perform the prefix increment operation to know the value that is currently being iterated. Within , we create an instance of the AssignmentOperator, which will accept the as a first operand, and the value received from the iterator as a second operand: hasNext() Iterator#hasNext() preIncrement() variable ...
public class IterableLoopStatement extends AbstractLoopStatement {
    ...

	@Override
	protected boolean hasNext() {
		return iterator.hasNext();
	}

	@Override
	protected void preIncrement() {
		Value<?> next = iterator.next();
		new AssignmentOperator(variable, next)
				.evaluate();
	}

	@Override
	protected void postIncrement() {
	}

} 2.4 StatementParser To make the loop statements work, we need to map the loop lexemes to the defined statements using the StatementParser class. All the loops defined by the language grammar start with the keyword. Therefore, inside the parseExpression() method, we add a new case for the corresponding Keyword lexeme: loop loop package org.example.toylanguage;

public class StatementParser {
	...

	private Statement parseExpression() {
		Token token = tokens.next(TokenType.Keyword, TokenType.Variable, TokenType.Operator);
		switch (token.getType()) {
			case Variable:
			case Operator:
                ...
			case Keyword:
				switch (token.getValue()) {
					case "print":
                        ...
					case "input":
                        ...
					case "if":
                        ...
					case "struct":
                        ...
					case "fun":
                        ...
					case "return":
                        ...
					case "loop":
                        ...
				}
			default:
                ...
		}
	}
} After the Keyword, we read the next expression using the Dijkstra Two-Stack algorithm defined within the ExpressionReader class. If we read the For or the Iterable loop, we expect this expression to be only a variable. If we read the While loop, this expression will act as a loop condition, which can be an instance of any Expression implementation, including operator or function invocation. Therefore, we split the execution. In case the expression is the VariableExpression instance and the next lexeme is the Keyword, we continue to read the For and Iterable loop. In the negative case, we build the WhileLoopStatement using the condition expression: loop in case "loop": {
    Expression loopExpression = new ExpressionReader().readExpression();

    AbstractLoopStatement loopStatement;
    if (loopExpression instanceof VariableExpression && tokens.peek(TokenType.Keyword, "in")) {
        // loop <variable> in <bounds>
    } else {
        // loop <condition>
        loopStatement = new WhileLoopStatement(loopExpression);
    } Inside the For and Iterable loop clause, we skip the next Keyword lexeme, and then read the following bounds expression. This expression can be an instance of any Expression implementation which can be evaluated to the lower bound value or IterableValue only when the code will be executed. To define which loop type it is, we can rely on the next lexeme. In the case of the For loop, after the lower bound expression, we expect to read the two-dot GroupDivider as a spliterator for lower and upper bounds. In the negative condition, we build the IterableLoopStatement: in if (loopExpression instanceof VariableExpression && tokens.peek(TokenType.Keyword, "in")) {

    // loop <variable> in <bounds>
    VariableExpression variable = (VariableExpression) loopExpression;
    tokens.next(TokenType.Keyword, "in");
    Expression bounds = new ExpressionReader().readExpression();

    if (tokens.peek(TokenType.GroupDivider, "..")) {
        // loop <variable> in <lower_bound>..<upper_bound>
    } else {
        // loop <variable> in <iterable>
        loopStatement = new IterableLoopStatement(variable, bounds);
    }
} To build the last For loop statement, we skip the two-dot GroupDivider and read the upper bound expression. In our language grammar, for the For loop, we can define the step expression with the keyword. Therefore, if the next expression is the lexeme, we read the following step expression. At the end, we build an instance of the ForLoopStatement: by by if (tokens.peek(TokenType.GroupDivider, "..")) {
    // loop <variable> in <lower_bound>..<upper_bound>
    tokens.next(TokenType.GroupDivider, "..");
    Expression upperBound = new ExpressionReader().readExpression();

    Expression step = null;
    if (tokens.peek(TokenType.Keyword, "by")) {
        // loop <variable> in <lower_bound>..<upper_bound> by <step>
        tokens.next(TokenType.Keyword, "by");
        step = new ExpressionReader().readExpression();
    }

    loopStatement = new ForLoopStatement(variable, bounds, upperBound, step);
} We have read every loop type declaration. Finally, we can read the body of the loop. To gather all internal statements, we invoke the recursive method, which will return an instance of the Statement interface. We should stop the process when we reach the finalizing lexeme: StatementParser#parseExpression() end while (!tokens.peek(TokenType.Keyword, "end")) {
    Statement statement = parseExpression();
    loopStatement.addStatement(statement);
} At the end, we skip this keyword and return the loopStatement. The complete clause: end loop case "loop": {
    Expression loopExpression = new ExpressionReader().readExpression();
    AbstractLoopStatement loopStatement;

    if (loopExpression instanceof VariableExpression && tokens.peek(TokenType.Keyword, "in")) {
        // loop <variable> in <bounds>
        VariableExpression variable = (VariableExpression) loopExpression;
        tokens.next(TokenType.Keyword, "in");
        Expression bounds = new ExpressionReader().readExpression();

        if (tokens.peek(TokenType.GroupDivider, "..")) {
            // loop <variable> in <lower_bound>..<upper_bound>
            tokens.next(TokenType.GroupDivider, "..");
            Expression upperBound = new ExpressionReader().readExpression();

            Expression step = null;
            if (tokens.peek(TokenType.Keyword, "by")) {
                // loop <variable> in <lower_bound>..<upper_bound> by <step>
                tokens.next(TokenType.Keyword, "by");
                step = new ExpressionReader().readExpression();
            }

            loopStatement = new ForLoopStatement(variable, bounds, upperBound, step);

        } else {
            // loop <variable> in <iterable>
            loopStatement = new IterableLoopStatement(variable, bounds);
        }

    } else {
        // loop <condition>
        loopStatement = new WhileLoopStatement(loopExpression);
    }

    while (!tokens.peek(TokenType.Keyword, "end")) {
        Statement statement = parseExpression();
        loopStatement.addStatement(statement);
    }

    tokens.next(TokenType.Keyword, "end");

    return loopStatement;
} 3 Tests We have finished embedding loops in both lexical and syntax analyzers. Now we should be able to perform For, While and Iterable loops. Let’s create and test a more real task - the bubble sort algorithm: fun bubble_sort [ arr, n ]
    loop i in 0..n - 1
        loop j in 0..n - i - 1
            if arr{j+1} < arr{j} then
                temp = arr{j}
                arr{j} = arr{j + 1}
                arr{j + 1} = temp
            end
       end
    end
end

fun is_sorted [ arr, n ]
    loop i in 0..n - 1
        if arr{i} > arr{i+1} then
            return false
        end
    end
    return true
end

arr = {}
arr_len = 20
loop i in 0..arr_len by 1
    arr << i * 117 % 17 - 1
end

print arr # [-1, 14, 12, 10, 8, 6, 4, 2, 0, 15, 13, 11, 9, 7, 5, 3, 1, -1, 14, 12]
print is_sorted [ arr, arr_len ] # false

bubble_sort [ arr, arr_len ]
print arr # [-1, -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 12, 13, 14, 14, 15]
print is_sorted [ arr, arr_len ] # true

Building Your Own Programming Language From Scratch: Part VI - Loops

Building Your Own Programming Language From Scratch Part IV: Implementing Functions

I've recently become obsessed with , a word puzzle game created by Brooklyn-based software engineer for his word game-loving partner. As a homage to Josh, and just-for-fun, I created a version of the game that can be played via text message. Read on to learn how to build an SMS word game using the , , the , , and , and play Twordle yourself: ! Wordle Josh Wardle Dictionary API Twilio Functions Twilio Serverless Toolkit Twilio Assets cookies in Twilio Runtime text a 5-letter word or "?" to +12155156567 Prerequisites A Twilio account - sign up for a free one here and receive an extra $10 if you upgrade through this link A Twilio phone number with SMS capabilities - configure one here Node.js installed - download it here Get Started with the Twilio Serverless Toolkit The Serverless Toolkit is CLI tooling that helps you develop locally and deploy to . The best way to work with the Serverless Toolkit is through the . If you don't have the Twilio CLI installed yet, run the following commands on the command line to install it and the Serverless Toolkit: Twilio Runtime Twilio CLI npm install twilio-cli -g
twilio login
twilio plugins:install @twilio-labs/plugin-serverless Create your new project and install our lone requirement , an HTTP client library to make HTTP requests in Node.js, by running: got twilio serverless:init twordle
cd twordle
npm install got Add a Static Text File to Twilio Assets Twilio Assets is a static file hosting service that allows developers to quickly upload and serve the files needed to support their applications. We want our Twilio Asset to be private–this means it will not be accessible by URL or exposed to the web; rather, it will be packaged with our Twilio Function at build time. . For more information on Private, Public, and Protected Assets, check out this page Copy this and add it to our folder as . We will read the file from our Twilio Function and generate a random word from it that will be used for each Wordle game. The word will be different for each person, and each person can play multiple times a day. GitHub file containing five-letter words from the English dictionary Assets words.private.text Write the Word Game Logic with JavaScript into the directory and make a new file called game.js containing the following code to import the module, read the file from Twilio Assets, create a function to return a random word from the Asset, and initialize two constants (the user always has five chances to guess the word, and all the words are five letters): cd \functions got words.txt randomWord const got = require('got');
let words = Runtime.getAssets()['/words.txt'].open().toString().split("\n");
const randomWord = () => {
    return words[words.length * Math.random() | 0];
}
const maxGuesses = 5;
const wordLength = 5; Next we have the meaty function that takes in a parameter (an object representing each player), and a guess (the word they text in as a guess.) We make a score card which will contain the boxes we return based on how close the user's guess is to the generated random word. In a block, we make an HTTP request with to the dictionary API using : if the page exists, the guess is a word, and we increment the attribute of the player object. For each letter in the guess, we check if it is in the goal word: if a letter is in the same spot, that spot in the score card will contain a green square (🟩). If a letter in the guess is not in the same index as the generated word for the player but the letter is in the generated word, the score card will contain a yellow square (🟨). Else, the score card will contain a black square (⬛). If our HTTP request is unsuccessful, our score card will be a string telling the user to try again. handleGuess player try got guess guessesAttempted const handleGuess = async (player, guess) => {
  let newScoreCard = [];
  try {
    const response = await got(`https://api.dictionaryapi.dev/api/v2/entries/en/${guess}`).json();
    if (response.statusCode !== 404) {
      player.guessesAttempted+=1;
      for (let i = 0; i < guess.length; i++) {
        if (guess.charAt(i) == player.randWord.charAt(i)) {
          if (player.dupLetters[i] != null) {
            player.numCorrectLetters+=1;
          }
          player.dupLetters[i] = null;
          newScoreCard.push('🟩');
        } else if (guess.charAt(i) != player.randWord.charAt(i) && player.randWord.includes(guess.charAt(i))) {
          newScoreCard.push('🟨');
        } else {
          if (!player.incorrectLettersArr.includes(guess.charAt(i))); {
            player.incorrectLettersArr.push(guess.charAt(i));
          }
          newScoreCard.push('⬛');
        }
      }
      console.log(`newScoreCard ${newScoreCard}`);
      return newScoreCard;
    } 
    else { //404 word not in dict
      newScoreCard = "word not found in dictionary! try again!";
      console.log('Word not found!');
      return newScoreCard;
    }
  }
  catch (err) {
    newScoreCard = "word not found in dictionary! try again!";
    console.log('Word not found!');
    return newScoreCard;
  }  
} After our function to handle each guess, let's make a function to check if the game is over. For parameters it accepts the object and . If the number of attempted guesses for the player is greater than or equal to five (the most number of guesses a player can have), the number of correct letters guessed is equal to the word length (five), or the score card contains five green squares, the game is over and returns true. Else, the game continues and returns false. player scoreCard endFunc const endFunc = (player, scoreCard) => {
  if (player.guessesAttempted >= maxGuesses) { 
    console.log(`guessesAttempted >= maxGuesses`);
    return true;
  }
  else if (player.numCorrectLetters == wordLength) {
    console.log("in numCorrect");
    return true;
  }
  else if(scoreCard == `🟩,🟩,🟩,🟩,🟩`) {
    console.log(`scorecard = 🟩,🟩,🟩,🟩,🟩`);
    return true;
  }
  else {
    console.log(`game still going`);
    return false;
  }
} Call Game Logic in Twilio Functions' Handler Method The handler method is like the entry point to your app, similar to a function in Java or in Python. In this tutorial, it will run each time someone texts our Twilio number. . main() __init__ For more information on Function invocation and execution, read this page First in the method, we initialize a Twilio Messaging Response object to respond to the player's guess text message, a variable which is whatever the player texted in, a string as empty text that we will append to depending on the guess, create a Twilio Response object to handle memory management with cookies,and a object whose attributes we will initialize based on the guess. guess responseText player exports.handler = async function(context, event, callback) {
  let twiml = new Twilio.twiml.MessagingResponse();
  let responseText = '';
  let guess = event.Body.toLowerCase().trim();
  let response = new Twilio.Response();
  let player; If the player texts in a question mark, we return a message about Josh Wardle who made the game as well as instructions on how to play the game. if (guess == "?") {
    twiml.message(`Wordle was made by Josh Wardle, a Brooklyn-based software engineer, for his partner who loves word games. You guess a 5-letter word and the responding tiles reflect how close your guess was to the goal word. 🟩 means a letter was in the right spot, 🟨 means the letter was correct but in the wrong spot, and ⬛️ means the letter is not in the goal word.`)
    return callback(null, twiml); //no need for cookies
  } Then we check if the player has texted in before. If the player does not exist, we generate a new word for them and initialize a new player object with the random word, the guesses attempted (none so far), number of correct letters (none so far), an array of duplicate letters, and an array of incorrect letters guessed (currently empty.) If the player does exist, we pull data off the cookie to get the player state and make that the object. with cookies player if (!event.request.cookies.player) { //any guesses attempted? -> new player
    let randWord = randomWord(); //new random word
    player = { //init new player
      randWord: randWord, 
      guessesAttempted: 0,
      numCorrectLetters: 0,
      dupLetters: [...randWord],
      incorrectLettersArr: []
    }
  } else { //else pull data off cookie to get player state
    player = JSON.parse(event.request.cookies.player);
  } We check the length of the guess and if it's five letters, we run the method and pass it and from above. We then check if the game is over and if it was a win, we send a congratulatory response; otherwise if it's a loss, we send a more apologetic message. Under both conditions, we remove the from cookie memory to start the player over using . handleGuess player guess player response.removeCookie("player"); If the game is not over, the response message is the scorecard with squares and we save the game state with the object with . It's in that we also set a four-hour time limit so the user has four hours to guess before the game state is lost–the default time limit for cookies in a Twilio Function is one hour. Lastly, if the guess is not five letters-long, we tell the player to send a five-letter word. player response.setCookie setCookie if (guess.length == wordLength) { //5 letters
    let scoreCard = await handleGuess(player, guess); //guessesAttempted increments
    console.log(`scoreCard ${scoreCard}`);
    if(endFunc(player, scoreCard)) { //over, win
      if(guess == player.randWord) {
        responseText += `Nice🔥! You guessed the right word in ${player.guessesAttempted}/${maxGuesses} guesses. You can play again by sending a 5-letter word to guess a new random word 👀 \nThanks for playing our SMS Twordle game. Do head to https://www.powerlanguage.co.uk/wordle for web-based word fun! Original Wordle creator Josh Wardle: as fellow builders we salute you and thank you for inspiring us to create our SMS experiment`
        response.removeCookie("player");
      }
      else if (guess != player.randWord) { //over, lose
        responseText += `Game over 🙈\nThe correct word was ${player.randWord}. Send a 5-letter guess to play again! \nThanks for playing our SMS Twordle game. Do head to https://www.powerlanguage.co.uk/wordle for web-based word fun! Original Wordle creator Josh Wardle: as fellow builders we salute you and thank you for inspiring us to create our SMS experiment`;
        response.removeCookie("player");
      }
    }
    else { //keep guessing, not over
      responseText += `${scoreCard.toString()} \n${player.guessesAttempted}/${maxGuesses} guesses`;
      response.setCookie("player", JSON.stringify(player), [
        'Max-Age=14400' //4 hour time-limit
      ]);
    }
  }
  else { //not 5 letters
    responseText += `"${guess}" is not valid. Please send a word in the dictionary that is 5 letters to get started!`;
    // twiml.message(`"${guess}" is not valid. Please send a word in the dictionary that is 5 letters to get started!`);
    console.log(`randWord ${player.randWord} in invalid `);
  } At the bottom of the handler method we append header, add information to our response about playing if the player has only guessed one time, send our in , and add twiml to return to our Twilio Response to both send our response text message to the player as well as update the object in cookie memory. responseText twiml.message player response.appendHeader('Content-Type', 'text/xml');
  // see if player.guessesAttempted == 1
  if (player.guessesAttempted == 1) {
    responseText += `\nText "?" for help on how to play`
  }
  twiml.message(responseText);
  response.setBody(twiml.toString());
  return callback(null, response); Wow that was a lot! The complete handler method is below. exports.handler = async function(context, event, callback) {
  let twiml = new Twilio.twiml.MessagingResponse();
  let responseText = '';
  let guess = event.Body.toLowerCase().trim();
  let response = new Twilio.Response();
  let player;
  if (guess == "?") {
    twiml.message(`Wordle was made by Josh Wardle, a Brooklyn-based software engineer, for his partner who loves word games. You guess a 5-letter word and the responding tiles reflect how close your guess was to the goal word. 🟩 means a letter was in the right spot, 🟨 means the letter was correct but in the wrong spot, and ⬛️ means the letter is not in the goal word.`)
    return callback(null, twiml); //no need for cookies
  }
  
  if (!event.request.cookies.player) { //any guesses attempted? -> new player
    let randWord = randomWord(); //new random word
    player = { //init new player
      randWord: randWord, 
      guessesAttempted: 0,
      numCorrectLetters: 0,
      dupLetters: [...randWord],
      incorrectLettersArr: []
    }
  } else { //else pull data off cookie to get player state
    player = JSON.parse(event.request.cookies.player);
  }
  
  if (guess.length == wordLength) { //5 letters
    let scoreCard = await handleGuess(player, guess); //guessesAttempted increments
    console.log(`scoreCard ${scoreCard}`);
    if(endFunc(player, scoreCard)) { //over, win
      if(guess == player.randWord) {
        responseText += `Nice🔥! You guessed the right word in ${player.guessesAttempted}/${maxGuesses} guesses. You can play again by sending a 5-letter word to guess a new random word 👀 \nThanks for playing our SMS Twordle game. Do head to https://www.powerlanguage.co.uk/wordle for web-based word fun! Original Wordle creator Josh Wardle: as fellow builders we salute you and thank you for inspiring us to create our SMS experiment`
        response.removeCookie("player");
      }
      else if (guess != player.randWord) { //over, lose
        responseText += `Game over 🙈\nThe correct word was ${player.randWord}. Send a 5-letter guess to play again! \nThanks for playing our SMS Twordle game. Do head to https://www.powerlanguage.co.uk/wordle for web-based word fun! Original Wordle creator Josh Wardle: as fellow builders we salute you and thank you for inspiring us to create our SMS experiment`;
        response.removeCookie("player");
      }
    }
    else { //keep guessing, not over
      responseText += `${scoreCard.toString()} \n${player.guessesAttempted}/${maxGuesses} guesses`;
      console.log(`randWord in not over ${player.randWord}`);
      response.setCookie("player", JSON.stringify(player), [
        'Max-Age=14400' //4 hour time-limit
      ]);
    }
  }
  else { //not 5 letters
    responseText += `"${guess}" is not valid. Please send a word in the dictionary that is 5 letters to get started!`;
    // twiml.message(`"${guess}" is not valid. Please send a word in the dictionary that is 5 letters to get started!`);
    console.log(`randWord ${player.randWord} in invalid `);
  }
  response.appendHeader('Content-Type', 'text/xml');
  // see if player.guessesAttempted == 1
  if (player.guessesAttempted == 1) {
    responseText += `\nText "?" for help on how to play`
  }
    // Add something to responseText that says: "Text 'HELP' for help" or whatever
  twiml.message(responseText);
  response.setBody(twiml.toString());
  return callback(null, response);
}; You can . view the complete code on GitHub here Configure the Function with a Twilio Phone Number To open up our app to the web with a public-facing URL, go back to the <em>twordle</em> root directory and run . Grab the link ending in . In the phone numbers section of your Twilio Console, select a purchased Twilio phone number and scroll down to the <em>Messaging</em> section. Under , change to and then under select , for select , and then for select . twilio serverless:deploy /game A MESSAGE COMES IN Webhook Function Service Twordle Environment dev-environment Function Path /game Click the button below and tada🎉! You can now text your Twilio number a 5-letter word to get started playing Twordle! Save What's Next for Twilio Serverless, Assets, and Word Games? Twilio's Serverless Toolkit makes it possible to deploy web apps quickly, and Twilio Runtime seamlessly handles servers for you. Let me know online what you're building with Serverless and what your current Wordle streak is! Mine is Wordle 208 5/6 ⬛⬛⬛⬛⬛ 🟧🟧⬛⬛⬛ 🟧🟧⬛⬛⬛ 🟧🟧⬛🟧⬛ 🟧🟧🟧🟧🟧 Twitter: @lizziepika GitHub: elizabethsiegle Email: lsiegle@twilio.com Livestreams: lizziepikachu

Build a Wordle-like SMS game with Twilio Serverless in JavaScript

In this article, we will go over a project on image recognition using Go. We will also create a Telegram bot, through which we can send images for recognition. The first thing we need is an already trained model. Yes, in this article we will not train our model. For this exercise, let's take a ready-made module from the docker image of ctava/tfcgo. To launch our project, we will need 4 terminals at the same time. In the first case, we will launch an image recognition server. In the second case, we will launch the bot. In the third case, we will launch a public tunnel for sending our bot "out". In the fourth - we will execute the command to register our bot. To start the recognition server, create a Dockerfile: FROM ctava/tfcgo

RUN mkdir -p /model && \
  curl -o /model/inception5h.zip -s && \
  unzip /model/inception5h.zip -d /model

WORKDIR / /src/imgrecognize
COPY src/ .
RUN build
ENTRYPOINT [ ]
EXPOSE "http://download.tensorflow.org/models/inception5h.zip" go go "/go/src/imgrecognize/imgrecognize" 8080 This way we will run our server in the image. Inside this image, we will have our server: src/imgrecognize. In addition, we will unpack the model in the directory: /model. For the server, the first thing we need is to set the value of the constant os.Setenv( , ) "TF_CPP_MIN_LOG_LEVEL" "2" This is necessary so as not to get an error: I tensorflow/core/platform/cpu_feature_guard.cc: ] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4 SSE4 AVX AVX2 FMA
unable to a tensor from image: Expected image (JPEG, PNG, or GIF), got empty file 140 .1 .2 make Here, we will not optimize our server, but simply run it through "ListenAndServe". On port 8080. Before starting the server, we will load our model (loadModel) and get our graph (modelGraph) and labels (labels). From the graph, which is stored in a file in the protobuf format "/model/tensorflow_inception_graph. pb". { model, err := ioutil.ReadFile(graphFile) err != { , , err
	}
	graph := tensorflow.NewGraph() err := graph.Import(model, ); err != { , , err
	} labelsFile, err := os.Open(labelsFile) err != { , , err
	} labelsFile.Close()
	scanner := bufio.NewScanner(labelsFile) labels [] scanner.Scan() {
		labels = (labels, scanner.Text())
	} graph, labels, scanner.Err()
} func loadModel () (*tensorflow.Graph, [] , error) string // Load inception model if nil return nil nil if "" nil return nil nil // Load labels if nil return nil nil defer var string for append return Actually, in "modelGraph" we keep the "structure" of our model and the key tools for working with it. And "labels" contains a "dictionary" for working with our model. Inside our HTTP handler, we are required to normalize the resulting image-normalizeImage. In order to pass it on to the recognition input in the future. To normalize, we convert our image from a Go value to a Tensor: tensor, err := tensorflow.NewTensor(buf.String()) After that, we get three variables graph, input, output, err := getNormalizedGraph() "graph" - we need to decode, resize, and normalize an image. The "input", together with the tensor, will be the "input point" for "communication" between our application and tensorflow. The "output" will be used as the output signal. Through "graph", we will also open a session to start normalization directly. session, err := tensorflow.NewSession(graph, ) nil Normalization Code: { buf bytes.Buffer
	_, err := io.Copy(&buf, imgBody) err != { , err
	}

	tensor, err := tensorflow.NewTensor(buf.String()) err != { , err
	}

	graph, input, output, err := getNormalizedGraph() err != { , err
	}

	session, err := tensorflow.NewSession(graph, ) err != { , err
	}

	normalized, err := session.Run( [tensorflow.Output]*tensorflow.Tensor{
			input: tensor,
		},
		[]tensorflow.Output{
			output,
		}, ) err != { , err
	} normalized[ ], } func normalizeImage (imgBody io.ReadCloser) (*tensorflow.Tensor, error) var if nil return nil if nil return nil if nil return nil nil if nil return nil map nil if nil return nil return 0 nil After normalizing the image, we create a session for inference over modelGraph. session, err := tensorflow.NewSession(modelGraph, ) nil With the help of this session (session), we will start the recognition itself. The input is our normalized image modelGraph.Operation( ).Output( ): normalizedImg, "input" 0 The result of the calculation (recognition) will be saved in the "outputRecognize"variable. From the received data we get the last 3 results: res := getTopFiveLabels(labels, outputRecognize[ ].Value().([][] )[ ]) { resultLabels []Label i, p := probabilities { i >= (labels) { }
		resultLabels = (resultLabels, Label{Label: labels[i], Probability: p})
	}
	sort.Sort(Labels(resultLabels)) resultLabels[:ResultCount]
} 0 float32 0 [] func getTopFiveLabels (labels [] , probabilities [] ) string float32 Label var for range if len break append return And for the HTTP response, we will give only one most likely result: msg := fmt.Sprintf( , res[ ].Label, res[ ].Probability* )
_, err = w.Write([] (msg)) "This is: %s (%.2f%%)" 0 0 100 byte All the code of our server for recognition: main ( tensorflow ) (
	ResultCount = ) (
	graphFile  = labelsFile = ) Label {
	Label Probability } Labels []Label { (l)
} {
	l[i], l[j] = l[j], l[i]
} { l[i].Probability > l[j].Probability
} (
	modelGraph *tensorflow.Graph
	labels     [] ) { err := os.Setenv( , ) err != {
		log.Fatalln(err)
	}

	modelGraph, labels, err = loadModel() err != {
		log.Fatalf( , err)
	}

	log.Println( )
	http.HandleFunc( , mainHandler)
	err = http.ListenAndServe( , ) err != {
		log.Fatalln(err)
	}
} {
	normalizedImg, err := normalizeImage(r.Body) err != {
		log.Fatalf( , err)
	} session, err := tensorflow.NewSession(modelGraph, ) err != {
		log.Fatalf( , err)
	}

	outputRecognize, err := session.Run( [tensorflow.Output]*tensorflow.Tensor{
			modelGraph.Operation( ).Output( ): normalizedImg,
		},
		[]tensorflow.Output{
			modelGraph.Operation( ).Output( ),
		}, ,
	) err != {
		log.Fatalf( , err)
	}

	res := getTopFiveLabels(labels, outputRecognize[ ].Value().([][] )[ ])
	log.Println( ) _, l := res {
		fmt.Printf( , l.Label, l.Probability* )
	}
	log.Println( )

	msg := fmt.Sprintf( , res[ ].Label, res[ ].Probability* )
	_, err = w.Write([] (msg)) err != {
		log.Fatalf( , err)
	}
} { model, err := ioutil.ReadFile(graphFile) err != { , , err
	}
	graph := tensorflow.NewGraph() err := graph.Import(model, ); err != { , , err
	} labelsFile, err := os.Open(labelsFile) err != { , , err
	} labelsFile.Close()
	scanner := bufio.NewScanner(labelsFile) labels [] scanner.Scan() {
		labels = (labels, scanner.Text())
	} graph, labels, scanner.Err()
} { resultLabels []Label i, p := probabilities { i >= (labels) { }
		resultLabels = (resultLabels, Label{Label: labels[i], Probability: p})
	}
	sort.Sort(Labels(resultLabels)) resultLabels[:ResultCount]
} { buf bytes.Buffer
	_, err := io.Copy(&buf, imgBody) err != { , err
	}

	tensor, err := tensorflow.NewTensor(buf.String()) err != { , err
	}

	graph, input, output, err := getNormalizedGraph() err != { , err
	}

	session, err := tensorflow.NewSession(graph, ) err != { , err
	}

	normalized, err := session.Run( [tensorflow.Output]*tensorflow.Tensor{
			input: tensor,
		},
		[]tensorflow.Output{
			output,
		}, ) err != { , err
	} normalized[ ], } {
	s := op.NewScope()
	input = op.Placeholder(s, tensorflow.String)
	decode := op.DecodeJpeg(s, input, op.DecodeJpegChannels( )) output = op.Sub(s,
		op.ResizeBilinear(s,
			op.ExpandDims(s,
				op.Cast(s, decode, tensorflow.Float),
				op.Const(s.SubScope( ), ( ))),
			op.Const(s.SubScope( ), [] { , })),
		op.Const(s.SubScope( ), ( )))
	graph, err = s.Finalize() graph, input, output, err
} package import "bufio" "bytes" "fmt" "io" "io/ioutil" "log" "net/http" "os" "sort" "github.com/tensorflow/tensorflow/tensorflow/go" "github.com/tensorflow/tensorflow/tensorflow/go/op" const 3 var "/model/tensorflow_inception_graph.pb" "/model/imagenet_comp_graph_label_strings.txt" type struct string float32 type func (l Labels) Len () int return len func (l Labels) Swap (i, j ) int func (l Labels) Less (i, j ) int bool return var string func main () // I tensorflow/core/platform/cpu_feature_guard.cc:140] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA // unable to make a tensor from image: Expected image (JPEG, PNG, or GIF), got empty file "TF_CPP_MIN_LOG_LEVEL" "2" if nil if nil "unable to load model: %v" "Run RECOGNITION server ...." "/" ":8080" nil if nil func mainHandler (w http.ResponseWriter, r *http.Request) if nil "unable to make a normalizedImg from image: %v" // Create a session for inference over modelGraph nil if nil "could not init session: %v" map "input" 0 "output" 0 nil if nil "could not run inference: %v" 0 float32 0 "--- recognition result:" for range "label: %s, probability: %.2f%%\n" 100 "---" "This is: %s (%.2f%%)" 0 0 100 byte if nil "could not write server response: %v" func loadModel () (*tensorflow.Graph, [] , error) string // Load inception model if nil return nil nil if "" nil return nil nil // Load labels if nil return nil nil defer var string for append return [] func getTopFiveLabels (labels [] , probabilities [] ) string float32 Label var for range if len break append return func normalizeImage (imgBody io.ReadCloser) (*tensorflow.Tensor, error) var if nil return nil if nil return nil if nil return nil nil if nil return nil map nil if nil return nil return 0 nil // Creates a graph to decode, rezise and normalize an image func getNormalizedGraph () (graph *tensorflow.Graph, input, output tensorflow.Output, err error) 3 // 3 RGB "make_batch" int32 0 "size" int32 224 224 "mean" float32 117 return Now, we need to build this image (build it). Of course, we can build an image and run it in the console using the appropriate commands. But it is more convenient to build these commands in a Makefile. So, let's create this handy file: recognition_build:
	docker build -t imgrecognition .

recognition_run:
	docker run -it -p : imgrecognition 8080 8080 After that, open the terminal and run the command: recognition_build && recognition_run make make Now, in the first terminal, we have a local HTTP server that can accept images. In response, it sends a text message containing information about what was recognized in the image. This is so to say the "core" of our project. Creating a Telegram Bot Next, we need to create a Telegram bot. We need to "build" the bot; to do this, we need to write a second HTTP server. The first server recognizes our images and uses port 8080. The second one will be the Bot's server and will use port 3000. First, we need to create a bot through your account in the app via . With this registration, you will receive the bot's name and its token. Don't tell anyone about this token. BotFather Let's put this token in the "BotToken" constant. You should get something like this: BotToken = const "1695571234:AAEbodyrfOjto2xNE5yjpQpW2Gyq0Ob5X24D5" Our bot's handler will decode the JSON response body. json.NewDecoder(r.Body).Decode(webhookBody) We are interested in the photo in the sent message By the unique image ID- let's collect a link to the image itself . And download it webhookBody.Message.Photo. photoSize.FileID fmt.Sprintf(GetFileUrl, BotToken, photoSize.FileID) downloadResponse, err = http.Get(downloadFileUrl). We will send the image bytes to the handler of our first server: msg := recognitionClient.Recognize(downloadResponse) In response, we get a certain message - a text string. After that, we simply send this string to the User, as is, in the Telegram Bot. The entire bot code: main ( ) (
	BotToken = GetFileUrl       = DownloadFileUrl  = SendMsgToUserUrl = ) webhookReqBody {
	Message Msg
} Msg {
	MessageId Text From {
		ID FirstName Username } Photo *[]PhotoSize Chat {
		ID FirstName Username } Date Voice {
		Duration MimeType FileId FileSize } } PhotoSize {
	FileID Width Height FileSize } ImgFileInfo {
	Ok Result {
		FileId FileUniqueId FileSize FilePath } } {
	log.Println( )
	err := http.ListenAndServe( , http.HandlerFunc(Handler)) err != {
		log.Fatalln(err)
	}
} { webhookBody := &webhookReqBody{}
	err := json.NewDecoder(r.Body).Decode(webhookBody) err != {
		log.Println( , err) } downloadResponse *http.Response webhookBody.Message.Photo == {
		log.Println( , webhookBody) } _, photoSize := *webhookBody.Message.Photo { imgFileInfoUrl := fmt.Sprintf(GetFileUrl, BotToken, photoSize.FileID)
		rr, err := http.Get(imgFileInfoUrl) err != {
			log.Println( , err) } rr.Body.Close() fileInfoJson, err := ioutil.ReadAll(rr.Body) err != {
			log.Println( , err) } imgInfo := &ImgFileInfo{}
		err = json.Unmarshal(fileInfoJson, imgInfo) err != {
			log.Println( +imgFileInfoUrl, err)
		} downloadFileUrl := fmt.Sprintf(DownloadFileUrl, BotToken, imgInfo.Result.FilePath)
		downloadResponse, err = http.Get(downloadFileUrl) err != {
			log.Println( +downloadFileUrl, err) } downloadResponse.Body.Close()
	} recognitionClient := recognition.New()
	msg := recognitionClient.Recognize(downloadResponse) err := sendResponseToUser(webhookBody.Message.Chat.ID, msg); err != {
		log.Println( , err) }
} sendMessageReqBody {
	ChatID Text } { msgBody := &sendMessageReqBody{
		ChatID: chatID,
		Text:   msg,
	} msgBytes, err := json.Marshal(msgBody) err != { err
	} res, err := http.Post(fmt.Sprintf(SendMsgToUserUrl, BotToken), , bytes.NewBuffer(msgBytes)) err != { err
	} res.StatusCode != http.StatusOK {
		buf := (bytes.Buffer)
		_, err := buf.ReadFrom(res.Body) err != { err
		} errors.New( + res.Status)
	} } package import "bytes" "encoding/json" "errors" "fmt" "io/ioutil" "log" "net/http" "github.com/romanitalian/recognition/src/bot/recognition" // Register Bot: curl -F "url=https://9068b6869da7.ngrok.io "  https://api.telegram.org/bot1695571234:AAEbodyrfOjto2xNE5yjpQpW2Gyq0Ob5X24D5/setWebhook const "1695571234:AAEbodyrfOjto2xNE5yjpQpW2Gyq0Ob5X24D5" "https://api.telegram.org/bot%s/getFile?file_id=%s" "https://api.telegram.org/file/bot%s/%s" "https://api.telegram.org/bot%s/sendMessage" type struct type struct int `json:"message_id"` string `json:"text"` struct int64 `json:"id"` string `json:"first_name"` string `json:"username"` `json:"from"` `json:"photo"` struct int64 `json:"id"` string `json:"first_name"` string `json:"username"` `json:"chat"` int `json:"date"` struct int64 `json:"duration"` string `json:"mime_type"` string `json:"file_id"` int64 `json:"file_size"` `json:"voice"` type struct string `json:"file_id"` int `json:"width"` int64 `json:"height"` int64 `json:"file_size"` type struct bool `json:"ok"` struct string `json:"file_id"` string `json:"file_unique_id"` int `json:"file_size"` string `json:"file_path"` `json:"result"` func main () "Run BOT server ...." ":3000" if nil // This handler is called everytime telegram sends us a webhook event func Handler (w http.ResponseWriter, r *http.Request) // First, decode the JSON response body if nil "could not decode request body" return // ------------------------- Download last img var if nil "no photo in webhook body. webhookBody: " return for range // GET JSON ABOUT OUR IMG (ORDER TO GET FILE_PATH) if nil "unable retrieve img by FileID" return defer // READ JSON if nil "unable read img by FileID" return // UNMARSHAL JSON if nil "unable unmarshal file description from api.telegram by url: " // GET FILE_PATH if nil "unable download file by file_path: " return defer // --------------------------- Send img to server recognition. if nil "error in sending reply: " return // The below code deals with the process of sending a response message // to the user // Create a struct to conform to the JSON body // of the send message request // https://core.telegram.org/bots/api#sendmessage type struct int64 `json:"chat_id"` string `json:"text"` // sendResponseToUser notify user - what found on image. func sendResponseToUser (chatID , msg ) int64 string error // Create the request body struct // Create the JSON body from the struct if nil return // Send a post request with your token "application/json" if nil return if new if nil return return "unexpected status: " return nil The client code that sends the image request from the Bot to the Recognition Server: recognition ( ) imgRecognitionAddress = Client {
	httpClient *http.Client
} { &Client{
		httpClient: &http.Client{},
	}
} { msg method := req, err := http.NewRequest(method, imgRecognitionAddress, downloadResponse.Body) err != {
		log.Println( , err) msg
	}
	req.Header.Add( , ) recognitionResponse, err := c.httpClient.Do(req) err != {
		log.Println(err) msg
	} {
		er := recognitionResponse.Body.Close() er != {
			log.Println(er)
		}
	}()

	recognitionResponseBody, err := ioutil.ReadAll(recognitionResponse.Body) err != {
		log.Println( , err) msg
	}
	msg = (recognitionResponseBody) msg
} package import "io/ioutil" "log" "net/http" const "http://localhost:8080/" type struct * func New () Client return func (c *Client) Recognize (downloadResponse *http.Response) string var string "POST" if nil "error from server recognition" return "Content-Type" "image/png" // do request to server recognition. if nil return defer func () if nil if nil "error on read response from server recognition" return string return By the way, to make our bot work correctly-register our handler. To do this, run: ngrok http 3000 Immediately after executing this command, you will see a list of public addresses. The last one will be an address with HTTPS - we need it. For example, it can be: https://9068b6869da7.ngrok.io. And directly register our bot-say Telegram where to send webhooks: curl -F https: "url=https://9068b6869da7.ngrok.io" //api.telegram.org/bot1695571234:AAEbodyrfOjto2xNE5yjpQpW2Gyq0Ob5X24D5/setWebhook Now you can send a file with a photo to your bot and get information about what is depicted on it. Thanks for your attention.

Stories, News & Mentions about NEWSCORP

Newscorp’s Stories on HackerNoon

Newscorp's News & Mentions