From 2005 until 2022 I spent a lot of energy trying to convince you to write tests for your PHP code, using a variety of techniques and approaches. Then I decided to stop teaching people to test at conferences.

Instead, I want to share my thoughts on where we are in 2024 in terms of overall commitment to testing. What other tools are out there? Does it matter what tools you use? Does it matter if you actually have automated tests? Mutation testing? Static code analysis? Code reviews? Snarky chat sessions?

Even though the basics haven't changed, how we can learn from the successes and failures of others can greatly help you come up with a testing plan that works for you. The goal is confident deployments to production at any time!

Comments

Please login to leave a comment

Chris Abbey at 09:55 on 26 Oct 2024

Good to hear we’re doing the right things, and get a new tool out of the talk. Nice easy way to start the last day.

Enjoyed the talk - both the history, the now, and the look forward.

Chris has long been a staple of the PHP community and inspired other developers—myself included—to make automated testing a huge focus of their careers. He grumped so that we could run, and this talk really covers how far we've come (and where we might focus next).

Lane Staples at 13:10 on 27 Oct 2024

I always try to make it to Chris' talks when possible as he consistently provides solid, approachable, actionable advice for making projects more reliable and easier to maintain and/or encouragement that the efforts already being invested in are part of what he sees, from a wealth of experience-won wisdom, as the happy path. This talk was no exception.

Awesome talk, was super excited to hear the shout-out for PHPStan and Psalm. We have used these since the beginning of our API, on very strict settings. We always say that every thing PHPStan finds is one less bug we have to fix later. I would almost suggest a short demo of this to help drive this point home, but perhaps that would be out-of-scope for this talk. Mutation testing was very interesting to hear about, but a little hard to conceptualize. Perhaps an example of that would be helpful too?

Betcha this talk would slay as a 3-hr workshop, if you're not already doing those. Thanks Chris!