9 years ago
3
Google Is 2 Billion Lines of Code—And It’s All in One Place
How big is Google? We can answer that question in terms of revenue or stock price or customers or, well, metaphysical influence. But that’s not all. Google is, among other things, a vast empire of computer software. We can answer in terms of code. Google’s Rachel Potvin came pretty close to an answer Monday at an engineering conference in Silicon Valley. She estimates that the software needed to run all of Google’s Internet services...
Continue Reading
http://www.wired.com
Join the Discussion
One important thing to note is that while Google has everything in one repository they aren't a single project. They aren't "one thing" like the article claims. From what I heard from Googlers, Google actually uses lots of independent microservices, the single repository is just so the developers always have the latest versions of dependencies available.
I think my biggest take-away is this comment
I like the idea of a centralized library of functions already written that can be repurposed for other tasks easily. We have a gitlab server at my job where we've been squirreling away code projects. I can see creating one 'main' project containing subfolders for php, python, powershell, etc with discrete libraries that can be included in new projects with the minimum of fuss.
If you find yourself writing new functions in your new project, put them into the main library where everyone can use them for their projects.
Anyway, nice idea.
The biggest take away I got from talking with several people at Google is that they always build against the latest changes of other projects the depend on, so they catch incompatible changes early (they have special build machines that builds everything on demand). Having a library of functions could be solved with common packages in a package repository, for example, but this approach takes continuous integration to the next level.