We Started a Blog. What Happens Next May Shock You.
Welcome to the Skylight blog!
Skylight helps you keep your customers happy by pointing out performance problems in your Rails applications. Unlike most performance tools on the market, which drown you in as much data as possible, Skylight gives you higher quality, curated information and proactively alerts you to problems in production. If you haven't tried it yet, we'd love to have you sign up for a free, 30-day trial.
Building a tool that can detect trends in performance across billions of requests is tough. Collecting and processing enormous amounts of data is a hard technical challenge, but technology alone is not enough. We felt strongly that helping you understand the performance of your app required a fast, beautiful and intuitive UI. The tight feedback loop gives you the freedom to explore and takes performance tuning from drudge work to outright fun.
In this blog, we'll give you a play-by-play of how we build Skylight, from technical details to design decisions.
Early in Skylight's life, we started a daily mailing list, where we send a short email describing what we've worked on that day. Those emails became one of our customers' favorite "features," because they're full of interesting technical details and devoid of marketing/sales BS. Now we'd like to share our process with everyone.
Skylight is built with badass hipster tech: Ruby on Rails, Ember.js, Rust, Storm, Dropwizard, Kafka, Cassandra, and more. We want to share the nitty gritty details about how we use them, but also how to get all of these things working together. For an overview of our architecture, see the talk I gave with Yehuda Katz at RailsConf:
While we’re excited about sharing our technical experience, we’re even more excited to talk about how we’ve built a sustainable, bootstrapped company centered on open source. We all had previous experience working for VC-backed companies, and we knew that balancing the need for explosive growth with the care and feeding of open source projects could be tricky. We wanted the ability to control our own destiny.
Three years ago, we started Tilde after the last company we worked at sold to Facebook. (Don’t worry, we’re not internet millionaires—our stock never vested.) We’ve learned a lot about how to run a company without investors—sometimes from other companies like Basecamp, but mostly through trial and error.
We’d love to see more companies adopt this model, but we don’t want to be out-and-out cheerleaders, either. As with our technical posts, we’ll discuss the tradeoffs without the BS.