Coding in the past

Unveiled in 1946 at the University of Pennsylvania, ENIAC was the first general-purpose computer. See Figure 1-2. It was the size of a large room, and programmers punched holes in paper cards to code programs that could take hours to complete. Sometimes bugs would crawl inside these large computers, causing the circuits to malfunction and resulting in errors. Removing these bugs from the computer was called debugging, which is the name used even today.

Gradually, with advances in hardware, computers became smaller and more powerful. Whereas the ENIAC’s tens of thousands of resistors and capacitors took up almost 2,000 square feet, later microprocessors could fit all these electronics onto a chip the size of a postage stamp. Eventually, these microprocessors would be built using silicon, which is both cheap and plentiful.

Increased computing power from powerful microprocessors allowed programmers to write more complicated and resource-intensive programs. For example, computer games became faster, used more complex graphics, and displayed on-screen smoothly and realistically. Writing code, or software programming, depends on and is constrained by the underlying hardware on which the code runs. As computing power increases, code is written to provide more features at a faster speed to users.

Programming languages were also invented to take advantage of this new computing power. You may remember languages such as Basic, Fortran, Pascal, C++, and Java. Like spoken languages, programming languages were created to fill a need. If other programmers coded using the language, the programming language would survive and thrive; otherwise, it would die.

Popular programming languages can decline in popularity or die, but this can take a long time if the language is used for core processes. For example, Fortran is not nearly as popular as it was 30 years ago, but it continues to be used in the scientific community and in the financial sector, where it powers applications for some of the biggest banks in the world.

Coding today

In 2011, Marc Andreessen, creator of Netscape Navigator and now a venture capitalist, noted that “software is eating the world.” He predicted that software companies would rapidly disrupt existing companies. Traditionally, software was used on desktops and laptops. The software had to be installed, and the installation process at a minimum varied by computer type and might not even work or might be incompatible with your computer hardware and software. After the software was installed, you had to supply data to the program.

Four trends have dramatically increased the use of code in everyday life:

• Web-based software: This software operates in the browser without requiring installation. For example, if you want to check email, you previously had to install an email client by downloading the software or from a CD-ROM. Issues arose when the software was not available for your operating system or conflicted with your operating system version. Hotmail, a web-based email client, rose to popularity in part because it allowed users visiting www.hotmail.com to instantly check email without worrying about installation or software compatibility. Web applications increased consumer appetite to try more applications, and developers in turn were incentivized to write more applications.
• Internet broadband connectivity: Broadband connectivity has increased, providing a fast Internet connection to more people in the last few years than in the previous decade. Today, more than 2 billion people can access web-based software, up from approximately 50 million only a decade ago.
• Coding repositories: Anyone can publish code for others to view and use. Popular coding repositories, such as Github, are making coding a more collaborative, open, and public process than ever before. Programmers publish code to show others what they can build, to solicit feedback to increase functionality or find vulnerabilities, and to quickly spread software to other programmers.
• Mobile phones: Today’s smartphones bring programs with you wherever you go and help supply data to programs. Many software programs became more useful when accessed on the go than when limited to a desktop computer. For instance, the use of maps apps greatly increased thanks to mobile phones because users need directions the most when lost not just when at home on the computer planning a trip. In addition, mobile phones are equipped with sensors that measure and supply data such as orientation, acceleration, and current location through GPS. Now instead of having to input all the data to programs yourself, mobile devices can help. For instance, a fitness application such as RunKeeper automatically tracks your distance, speed, and time.

The combination of these trends has resulted in software companies that have upended incumbents in almost every industry, especially ones typically immune to technology. Some notable examples include the following:

• Airbnb: A peer-to-peer lodging company that owns no rooms, yet books more nights than the Hilton and Intercontinental, the largest hotel chain in the world. See Figure 1-3.
• Uber: A car transportation company that owns no vehicles but books more trips and has more drivers in 200 cities than any other car or taxi service.
• Groupon: A daily deals company that generated almost $1 billion after just two years in business, growing faster than any other company in history, let alone any other traditional direct marketing company.

Coding in the future

The one constant in technology and coding is change. Improvements in existing computer architecture will lead to the creation of newer, faster, and smaller hardware devices, and developers will then write code to operate and control those hardware devices.

Moore’s Law, a rule of thumb used in the computer hardware industry, predicts that the number of transistors per square inch on an integrated circuit will double every year. The prediction has proved to be true for the last 50 years, although some experts doubt whether it will continue to hold true for the next 50 years.

The following technology developments are increasing in popularity and should remain relevant at least for the next five years:

• Internet of Things (IOT): Computing power is transforming dumb hardware devices into smart, connected, self-regulated devices. For example, the Nest thermostat uses a motion detector to record when people are present, and then heats and cools homes when people are expected to be at home instead of all day. Similarly, Lockitron makes a device that allows you to lock and unlock your front door with your smartphone. Other connected devices, such as the FitBit fitness tracker and the Apple Watch, need coders to add functionality and connect people in new ways.
• Machine learning: For years, databases just stored data. Now, code is finally being written to analyze the data and make intelligent predictions. For example, mapping applications use real-time and historical data to predict traffic and the time your route will take to complete. 23andme, a genetics company, compares your human genome against its database to predict which diseases you are more likely to have. General Electric has outfitted industrial machines such as hospital equipment and jet engines with sensors, and uses historical data to repair machines before they break, decreasing downtime and increasing revenue. Coders will continue to write analytics programs to crunch large datasets and generate predictions with increasing accuracy.
• Interconnected applications: An application programming interface (API) allows one program to talk to and request data from another external program, which provides a response. Although APIs are powerful, their functionality can be limited and they rarely talk to one another. For example, Dropbox, the storage provider, has an API to allow third-party applications to back up data, and Facebook has an API that lets third-party applications retrieve a user’s photos. However, using just those two APIs, you cannot automatically back up every Facebook photo to Dropbox. Companies such as IFTTT (If This Then That) allows users to create recipes that combine APIs.
• Virtual software containers: Traditionally, software programs could be described as an interconnected web of your code and code written by others. To incorporate someone else’s code into your own program, you had to check that both programs were compatible and that any third-party code used by the external program, called a dependency, was also compatible with your code. The process of resolving conflicts was frequently time-consuming and frustrating. One solution is to move away from the current interconnected system of software programming to an independent self-contained system. Docker is one company that hosts an open-source project to help programmers package software and its dependencies into a self-contained program called a virtual container. These virtual containers have standardized inputs and outputs, run on many operating systems, and can connect to each other with little need to check for compatibility. Just like standardized shipping cargo containers make it easier and faster to load and unload ships, so too do virtual containers make it easier and faster to package programs to work easily with other programs.

Full-time jobs

Companies of various sizes hire people who have just learned how to code for full-time positions. The size of the company can have pros and cons when it comes to hiring people who have just learned how to code:

• Large companies: Companies with more than 1,000 employees, such as Fortune 500 companies and large tech companies including Yahoo!, Google, and Facebook have high standards for hiring employees. Given the number of applications they receive for each open position, recruiters at these companies usually use a strict screening process and grant interviews only to people who have a computer science, math, or engineering-related major. However, for those people who do pass the hiring screen and are eventually hired, there are many resources, both formal programs and people who can help coach and train you to increase your skills.

  Almost every large company has an online application. Send in your application online, and then find an advocate, someone at the company who believes in your candidacy, to help your application pass to the interview stage.

• Medium-sized companies: Of the three types of companies, getting hired at a medium-sized company can be hardest. With their large recruiting departments, candidates have to interview with as many people as in large companies. In addition, medium-sized companies typically do not spend as much money on training as large companies.

  One successful strategy to getting hired permanently in a medium-sized company is to freelance first, which helps you build up your reputation and allows the company to assess your skills in a low risk way.

• Startups: With less than 20 employees, startups often desperately need coding talent and are small enough to make hiring decisions quickly. They don’t have a formal recruiting staff, so you should develop a personal connection with the person doing the hiring. Startups don’t have extensive training programs, and you are expected to contribute immediately. However, the small company size should help you form personal relationships with your engineering coworkers, who can help answer questions and informally train you.

  In the beginning, successful startups often have so much work and are so short staffed that having anyone do the work is better than having no one. For this reason, startups decide on candidates quickly, rather than wait for the best person for each role.

• Government: City, state, and federal governments and their agencies have thousands of internship and full-time job openings for coders. Depending on the agency, the application process can be time consuming, and require proof of U.S. citizenship, extensive background checks, and completion of qualifying exams. Applicants can use www.usajobs.gov to search across all federal opportunities, and individual state government websites for opportunities in a specific state or city government.

Freelancing and contract jobs

Companies of all sizes hire freelancers to do discrete projects that are not overly complex and have a definite end date. For example, freelancers often build a website with a few defined pages, create mobile apps, or scrape and store data from websites into databases. Getting some of the initial work performed by a freelancer helps a company see how much time and money a project will cost and whether there is a need to hire a full-time employee.

Contract work also provides freelancers with some valuable benefits. Even for full-time coders, doing some contract work is a great way to build up skills in a new programming language or framework. Also, because coding work can be done from anywhere, freelancers have a good deal of flexibility and don’t necessarily have to sit behind a desk in an office. For example, some freelancers travel often for pleasure, and can be found working in cities such as Boston one month and Bali the next month. Finally, some coders freelance full-time, and build their business by doing work for existing clients and pitching new work to client referrals.

One issue with freelancing is that you are always looking for the next job. A few websites, such as Freelancer (www.freelancer.com) and Upwork (www.upwork.com), formerly odesk.com, help provide freelancers with steady work by creating communities that connect employers and freelancers. See Figure 1-5.

These sites create online reputations for both freelancers and companies, which helps each side feel more confident that the work will be completed and the agreed upon amount will be paid.