On December 18, 1974, while being transported in a car by Sheriff Earl Lee and the Georgia Bureau of Investigation Agent

The Engineering Productivity team at Pinterest came across a small change that had a large impact in reducing build times across pipelines. We found that setting the refspec option during git fetch reduced our cloning step by 99%.,Urvashi Reddy | Software Engineer, Engineering Productivity Team Adam Berry | Tech Lead, Engineering Productivity Team Rui Li | Software Engineer, Engineering Productivity Team,Tortured, Starved, Abused: The Disabled People Imprisoned in Philadelphia ‘Linda Ann Weston’ the accused who left no stone unturned to make their life miserable.medium.com,By simply adding the refspec option and specifying which refs we care about (master, in our case), we can limit the refs to the branch we care about and save a lot of time. Here’s what that looks like in our pipeline code:,All our larger efforts are designed to progress on the goal of making software development and delivery at Pinterest a fast and painless experience. Recently, we were reminded of how small details can also have a large impact on that goal. We discovered an overlooked Git option that significantly reduced the build times in our continuous integration pipelines. In order to understand how this small change had such a large impact, we need to share a little information on our monorepos and our pipelines.,Cloning monorepos that have a lot of code and history is time consuming, and we need to do it frequently throughout the day in our continuous integration pipelines. For Pinboard alone, we do more than 60K git pulls on business days. Most of our Jenkins pipeline configuration scripts (written in Groovy) start with a “Checkout” stage where we clone the repository that the later stages will build and test. Here’s what a typical “Checkout” stage looks like:,View the field as a mathematical framework that allows us to learn representations and insights from data. It is a very promising area because it offers a lot of opportunities for automation, based on historical data, which could not be done before by using an algorithm.,We have six main repositories at Pinterest: Pinboard, Optimus, Cosmos, Magnus, iOS, and Android. Each one is a monorepo and houses a large collection of language-specific services. Pinboard is as old as the company and is the largest monorepo. Pinboard has more than 350K commits and is 20GB in size when cloned fully.,This article explained why it’s worth having a look at Deep Learning and move away from traditional Machine Learning methods, especially when you’re dealing with a lot of data.,Sadly, this is a learned behavior pressed on all of us at a young age. We are taught that our mistakes are both the center of our focus and a bad thing. This mentality is at the core of our formal education. Academics have made us think that mistakes are nothing more than incorrect answers that hurt our grades, and the most successful of our classmates were the students that had the least amount of wrong answers. If you did poorly on a test, your job was to take a bad grade and hope it didn’t torpedo your odds of passing. This experience taught us to dread mistakes instead of learning how to deal with them appropriately.,The Engineering Productivity team at Pinterest is responsible for supporting the engineers who build and deploy software at the company. Our team maintains a number of infrastructure services and often works on large scale efforts — migrating all software at Pinterest to Bazel, creating a Continuous Delivery platform called Hermez, and maintaining the monorepos that get committed to a few hundred times a day, to name a few.,Even though we’re telling Git to do a shallow clone, to not fetch any tags, and to fetch the last 50 commits, we’re still not running this operation as fast as we could be. That’s because we aren’t setting the refspec option. Notice that by not setting it in the pipeline configuration script, we’re telling Git to fetch all refspecs: refs/heads/*:refs/remotes/origin/*. In the case of Pinboard, that operation would be fetching more than 2,500 branches.,It’s easy for us to explain that a smarter person is more likely to win from a dumb person in a game of chess. Well, guess what. If a computer learns every path to victory, he’s likely to be the better person. Deep learning tries to remember every path to victory, allowing it to compete and even beat world-class players in games like chess and go.,“The Casanova Killer had no compunction about killing you, makes no difference whether he strangled you, whether he shot you, whether he stabbed you or what. He was a martial arts expert. He was tough. He was mean.”,Tortured, Starved, Abused: The Disabled People Imprisoned in Philadelphia ‘Linda Ann Weston’ the accused who left no stone unturned to make their life miserable.medium.com