programming language used to design Facebook?

programming language used to design Facebook?.
arkrishna

Facebook and its Dirty little secrets:

 facebook headquarters arkrishna

Ever wondered if the Facebook is written using  PhP/python/C/C++ or ruby.  Today we  are going to talk about  all dirty secrets of Facebook and the programmers behind the Facebook.

Which programming language is Facebook based on?

There is no specific language that facebook uses.mainly they modify the languages according to their requirement .

mainly Hiphop for php HipHop for PHP.

cassandra databse .

facebook chatbox is coded in Erlang.

and need c++ also .

One of the key values at Facebook is to move fast. Simple to learn, simple to write, simple to read, and simple to debug. 

     

Facebook secrets arkrishna

HipHop for PHP isn't technically a compiler itself. Rather it is a   source code transformer. HipHop programmatic-ally transforms your PHP   source code into highly optimized C++ and then uses g++ to compile it.   HipHop executes the source code in a semantically equivalent manner   and sacrifices some rarely used features — such as eval() — in   exchange for improved performance. HipHop includes a code transformer,   a reimplementation of PHP's runtime system, and a rewrite of many   common PHP Extensions to take advantage of these performance   optimizations.

     

Scaling PHP as a Scripting Language  

   

PHP's roots are those of a scripting language, like Perl, Python, and   Ruby, all of which have major benefits in terms of programmer   productivity and the ability to iterate quickly on products. This is   compared to more traditional compiled languages like C++ and   interpreted languages like Java. On the other hand, scripting   languages are known to generally be less efficient when it comes to   CPU and memory usage. Because of this, it's been challenging to scale   Facebook to over 400 billion PHP-based page views every month.     

Facebook secrets arkrishna

One common way to address these inefficiencies is to rewrite the more   complex parts of your PHP application directly in C++ as PHP   Extensions. This largely transforms PHP into a glue language between   your front end HTML and application logic in C++. From a technical   perspective this works well, but drastically reduces the number of   engineers who are able to work on your entire application. Learning   C++ is only the first step to writing PHP Extensions, the second is   understanding the Zend APIs. Given that our engineering team is   relatively small — there are over one million users to every engineer   — we can't afford to make parts of our codebase less accessible than   others.     

Scaling Facebook is particularly challenging because almost every page   view is a logged-in user with a customized experience. When you view   your home page we need to look up all of your friends, query their   most relevant updates (from a custom service we've built called   Multifeed), filter the results based on your privacy settings, then   fill out the stories with comments, photos, likes, and all the rich   data that people love about Facebook. All of this in just under a   second. HipHop allows us to write the logic that does the final page   assembly in PHP and iterate it quickly while relying on custom   back-end services in C++, Erlang, Java, or Python to service the News   Feed, search, Chat, and other core parts of the site.     

Since 2007 we've thought about a few different ways to solve these   problems and have even tried implementing a few of them. The common   suggestion is to just rewrite Facebook in another language, but given   the complexity and speed of development of the site this would take   some time to accomplish. We've rewritten aspects of the Zend Engine —   PHP's internals — and contributed those patches back into the PHP   project, but ultimately haven't seen the sort of performance increases   that are needed. HipHop's benefits are nearly transparent to our   development speed.

     

Hacking Up HipHop 

    

One night at a Hackathon a few years ago (see Prime Time Hack), I   started my first piece of code transforming PHP into C++. The   languages are fairly similar syntactically and C++ drastically   outperforms PHP when it comes to both CPU and memory usage. Even PHP   itself is written in C. We knew that it was impossible to successfully   rewrite an entire codebase of this size by hand, but wondered what   would happen if we built a system to do it programmatically.     

facebook secrets arkrishna

Finding new ways to improve PHP performance isn't a new concept. At   run time the Zend Engine turns your PHP source into opcodes which are   then run through the Zend Virtual Machine. Open source projects such   as APC and eAccelerator cache this output and are used by the majority   of PHP powered websites. There's also Zend Server, a commercial   product which makes PHP faster via opcode optimization and caching.   Instead, we were thinking about transforming PHP source directly into   C++ which can then be turned into native machine code. Even compiling   PHP isn't a new idea, open source projects like Roadsend and phc   compile PHP to C, Quercus compiles PHP to Java, and Phalanger compiles   PHP to .Net.     

Needless to say, it took longer than that single Hackathon. Eight   months later, I had enough code to demonstrate it is indeed possible   to run faster with compiled code. We quickly added Iain Proctor and   Minghui Yang to the team to speed up the pace of the project. We spent   the next ten months finishing up all the coding and the following six   months testing on production servers. We are proud to say that at this   point, we are serving over 90% of our Web traffic using HipHop, all   only six months after deployment.

     

How HipHop Works     

The main challenge of the project was bridging the gap between PHP and   C++. PHP is a scripting language with dynamic, weak typing. C++ is a   compiled language with static typing. While PHP allows you to write   magical dynamic features, most PHP is relatively straightforward. It's   more likely that you see if (...) {...} else {..} than it is to see   function foo($x) { include $x; }. This is where we gain in   performance. Whenever possible our generated code uses static binding   for functions and variables. We also use type inference to pick the   most specific type possible for our variables and thus save memory.     

Facebook secrets arkrishna

The transformation process includes three main steps:    

Static analysis where we collect information on who declares what and dependencies,

Type inference where we choose the most specific type between C++ scalars, String, Array, classes, Object, and Variant, and

Code generation which for the most part is a direct correspondence  from PHP statements and expressions to C++ statements and expressions.

We have also developed HPHPi, which is an experimental interpreter   designed for development. When using HPHPi you don't need to compile   your PHP source code before running it. It's helped us catch bugs in   HipHop itself and provides engineers a way to use HipHop without   changing how they write PHP.     

Overall HipHop allows us to keep the best aspects of PHP while taking   advantage of the performance benefits of C++. In total, we have   written over 300,000 lines of code and more than 5,000 unit tests.

arkrishna