Pragmatic supremacy: "The father of Php" Rasmus Lerdorf interview

"The father of Php" Rasmus Lerdorf character directly magnanimous, concise and incisive, the lines of a "pragmatic supremacy" spirit of Qi. On the eve of the "PHP Global Developers Conference", the "doer" received a "programmer" interview, sharing his own programming sentiment.

Work Focus

The projects I participate in are always directly related to the users. I have been a Yahoo engineer for many years and are responsible for connecting the infrastructure of hundreds of millions of end users, these basic settings are still in service. Today, I am also the backend infrastructure responsible for connecting millions of users in Etsy. Technology is really just a problem-solving tool, an abstract hammer, a saw, and nothing great, and the really exciting thing is to use technology to improve the quality of life for millions of people.

"Problem" as a guideline

Compared to "computer scientists", "engineer" the title more to my heart. On the difference between the two, I think the latter is more focused on solving the immediate problem. PHP was developed not because I liked programming or language design, but because I didn't like it. The 1993 programming Bashi made it impossible for me to solve web problems easily and quickly. In my case, only when encountering difficulties, will turn over the books, check the information to enrich themselves, skills are not to improve and improve, my every decision is to solve the problem-centered.

experience and lessons learned

I made a lot of mistakes, some of which I realized later, but also achieved better than expected results. The most important lesson is that solving web problems really should focus on the entire ecosystem from the outset. For more than 20 years, solutions to web problems have been endless, but the quality is uneven, and few can build a complete ecosystem and be used by ordinary people.

PHP 7 where to ponder?

Developing PHP 7 The last 10% is the most time-consuming and boring. However, the powerful new features and performance leaps and bounds, but also the positive mentality of the upper hand, but also inspired the entire team, to help us quickly endure the past. However, each update has to do not complete the test, solve the problem of the platform, the investigation of the strange edge of the situation, see the vulnerability report, there is no end.

Recruit New people

I know that the version control system of Emacs was replaced by git from BZR, but I think they are actually equal in terms of absorbing new contributors. Git has recently been more popular, facilitating Dragonfly water-style building blocks. However, for long-term loyal contributors, the version control system does not matter. It is important to have good documentation and a process that facilitates new contributions.

If I could redesign PHP

If the time can be reversed, there must be something I want to improve, such as distinguishing keyword case. Just beginning PHP is just an HTML template language. In the early 90, people debated whether the HTML tag was the uppercase, lowercase, or mixed-case. I do not want to contend for, the PHP template tags made of the case-insensitive, this practice is still in use.

does JavaScript devour other languages?

The evolution of PHP and JavaScript is almost synchronous. I started with Brendan Eich (JavaScript Designer) at the same time, and his focus was obvious to the client, and I was the server. If you write a client application, there is no choice but JavaScript-which language the browser supports, but it is very different if the center of gravity is the server.

The client code is written by many people, so the number of JavaScript, and now it in the server this Chamber also bear fruit. But JavaScript is just as much a solution as PHP, not the only one, and that's fine. As I have just said, language is only a tool for solving problems, not a cult religion. If there is a problem and you are more inclined to JavaScript, then it is the best choice.

Programming Principles

As long as it is effective, safe, fast enough, release it, and then solve the next problem. Three are indispensable, or you will go back to check the code, good time to solve the next problem.

Future Prospects

As for programming languages, I never really thought about what these tools would be going through in the future, and I'm more concerned about whether they can solve the problems of the moment. Take Etsy for example, as a handmade crafts website, it can be between wealthy buyers and poor artists matchmaking, let them out of poverty dilemma? Can our infrastructure help other companies deal with the same major problems? Is our solution strong to a blink of an eye the customer's problem is gone? That's all I care about.

In the past, I didn't like programming, and I still don't like it. I only like to solve the problem as the center, this will never change.

Reporter: Lu thrush xiang, Zhang Xinhui

This article is "programmer" original article, not allowed to reprint, more wonderful articles please subscribe to 2016 programmer: http://dingyue.programmer.com.cn/

