Google is 2 billion lines of code
“What’s the capital of Grenada?”
We’ve all heard that line millions of times throughout the course of our lives. Humans have evolved to such an extent that the phrase “I don’t know” has been phased out by “Google it.” It is also for the first time perhaps in history that a brand has become a verb from a noun.
Google, then, is an immense part of our lives. It can be considered an empire, built on the foundations of its vast network of ever changing software. But, Google isn’t just a leading name in search engines; it has various other applications like Gmail, Google drive, YouTube and so much more, all under its belt. To run such a vast empire takes an extremely complicated set of codes; 2 billion lines of it to be exact. Microsoft Windows, one of the world’s most popular operating systems averages out at around 50 million lines of code. Combine the entire open source code for Windows 40 times and you have 1 Google. However, the devil is in the details. The kicker here is; these 2 billion lines of code lie in one single repository accessible by the 25,000 engineers of Google worldwide. Google treats its titanic amounts of code as one single project. To put that in layman’s terms; it is like letting one man be the president of the entire world. Sounds like a terrible plan doesn’t it? And yet, in the 18 years of Google’s existence, they have not only made it work but have also moved leaps and bounds ahead in terms of technological development.
Sharing 2 billion lines of code with 25,000 people will undoubtedly create some communication gaps and errors. Managing such a megalith is no walk in the park. People will eventually completely mess up the huge lines of code and there will be, as the saying goes, too many cooks; the broth will be ruined. Google thought of this issue and addressed it efficiently. They introduced Piper. Piper is Google’s bot used to control and regulate the code. Piper’s responsibilities include, taking most of the burden off human workers, juggle the constant commits(changes) and keep the code running smooth and most important of all, make the lines of code simple and minimalist so the engineers don’t fall into pits of confusion and error. In doing so, Piper scans through the 85TB worth of data and keeps it bug free, without the need for constant human interference.
Google has one secret up its sleeve though. Not ALL of its code lies in one single repository. There is a separate set of codes that manages Google’s PageRank search algorithm. PageRank is a software used by Google to rank websites in their search results. It is used by many people, mostly for picking out the apt websites to obtain authority links from. Access to the coding of PageRank is limited to a certain select bunch of engineers. This is so that this set of code does not fall into the wrong hans. Getting a high rating on PageRank is desirable for every website on the internet and so these lines of code are guarded as tightly as a vault at Fort Knox.
Google, then, is the final word in software infrastructure. It stands testament to the continuously developing nature and complexity of software in today’s day and age. Without Google, people might have to walk to a library to learn about the capital of Grenada. And who has time for that?
BTW the capital of Grenada is St George’s.