Hacker News Reader

Comment by ramraj07 | Hacker News Reader

Can someone point to a good write-up on how to introduce and start using feature flags in a repo? Seems like I just can't find any good resources on this topic!

View on HN

I'm a Developer Advocate at LaunchDarkly, and we've written a bunch of different guides over the years. For a good intro that discusses first steps with flags and also how to approach testing flagged code, see our eBook "Effective Feature Management"[1] (Download page is right behind the form, you don't need to wait for an email or anything)

One of the best ways to start is add a simple kill-switch to low-impact code. (More about that in the book.) But, I have to ask: what are the main issues that are blocking you from starting? Is it that you need ideas for cases, or are worried about breaking something, or just want to ensure best practices from the start?

We're constantly writing new guides, both as blog posts[2] and long-lived guide articles on our documentation site[3]. Feel free to message me on Twitter[4] if you need more help!

[1] https://launchdarkly.com/effective-feature-management-ebook/

[2] https://blog.launchdarkly.com/

[3] https://docs.launchdarkly.com/guides

[4] https://twitter.com/yoz

If you're looking to try it out in a simple way, you could probably dynamically load different functionality or modules with environment variables. But I'm also looking for a good resource

This is what we are trying now, but that can't do everything. Some teams use launch darkly, but I'm also looking at what level of abstraction we should try to do, and things such as testing both branches, etc. It seems an important architectural refactor to go into feature flag territory yet I see very little discussion on best practices there.

Hi - we're developing a 100% open source feature flagging platform that can achieve this - https://bullet-train.io/ - would love feedback!

Reply on news.ycombinator.com