A few years back I made a post on how I work with PHP + SQLServer. In the intervening years, the code samples in that post have aged like milk and no longer work.

I’m working on an infrastructure application that provides a GRPC API to its clients. This will be a high throughput system, so we are building it in Rust using tonic and tokio to build the gRPC API. I wanted to implement request authentication for this system so that the service was resilient to Server-Side-Request-Forgery, and unauthenticated requests.

Recently, I’ve been working on a system that needs to run tasks scheduled by either web requests, or other tasks. These tasks are trusted but sometimes problematic code. While developers generally write good code, we all make mistakes and the occasional slow job that can run for multiple hours gets written. These slow tasks can create problems if the volume of them goes up as they can consume all available resources.

I recently had to do some maintenance work on an old application hosted on the CakePHP server. The CakePHP project has 20 or so sites and applications deployed in dokku. Dokku allows us to build a platform-as-a-service for our sites that operates on a single server. Dokku encourages using heroku-style ‘buildpacks’ to deploy applications.

I wanted to share a new CakePHP plugin that provides feature flagging. In a Saas environment, I’ve been a long time advocate for using feature flags to release changes instead of deploys. Feature flags let you better manage release risk by allowing you deliver code with new features disabled.

During the development of CakePHP, we also wanted to update migrations to be compatible. However, we had stumbled into a tedious dependency graph. Currently phinx relies on cakephp/database. While cakephp/migrations depends on phinx. In order to do a major release on cakephp/migrations (to provide CakePHP 5.

As I use webcomponents more, I found myself wanting a way to define the HTML for webcomponents with a non-trivial amount of light DOM contents in a more reusable programatic way. So far, I’ve found that modelling webcomponents as template fragments is a reasonably ergonomic solution. It enables usage such as:

When re-building docket with htmx, I wanted to retain the confirm dialog experience I had with react. Instead of taking on a dependency for this, I chose to build my own combining HTMX and Webcomponents. The resulting UX feels snappy and similar to a client rendered experience. My end result looks like this: