Leveling Up Queries, Indexes, and Platforms: What’s New in Ditto SDK 4.12

KubeCon + CloudNativeCon North America 2025 wrapped up in Atlanta last week, and it sure did feel like a big one for OpenTelemetry. Between Observability Day, the project updates, and the activity around the OpenTelemetry Observatory booth, you could feel how quickly the ecosystem is maturing.

Using Projections to Improve Query Efficiency

One habit I’ve developed over the years is validating performance ideas before I start changing code. If I think there’s a way to make the app run more efficiently, I want to measure the potential benefit up front so I know whether the idea is worth pursuing.

For the main movie listing screen in my Mflix demo app, I realized I was pulling back entire movie documents—even though that screen only uses a handful of fields. The movies collection schema contains a lot of data, including cast lists, full reviews, and multiple nested objects. None of that is needed for a simple listing view.

Using Simple Indexes to add Search functionality

One of my testing patterns is to insert or update data frequently to make sure those parts of the app are working as expected. The challenge is that in a large dataset, finding the movie I just added or updated can be slow and tedious, especially when I want to validate it quickly. At the same time, giving users the ability to search for a movie by title would be a nice usability improvement.

I expected the LIKE %search% pattern to be a performance killer—especially since I initially skipped adding debounce logic, meaning the query runs on every keystroke. To my surprise, results were nearly instantaneous. In testing with physical devices,  results came back in as little as 120 ms, and even on a very old Android phone with slow storage, the longest query took around 400 ms.

For a production app, I’d still recommend adding debounce logic to reduce excessive disk I/O and prevent unnecessary battery drain. For a demo, though, the real-time updates are pretty fun to watch. In Flutter, I added debouncing because on older Android devices with slower storage, performance suffered without it. The index not only made the search possible, but also fast enough to feel instant, a big win for both development testing and potential end-user search experiences.

Moving Forward with DQL

In this article, I’ve covered a lot of exciting new capabilities that DQL brings to the table. Over the past year, our engineering team has put in a tremendous effort to make DQL a fully featured, production-ready query interface. We’ve now reached the point where anything you could previously build with the legacy Query Builder can be built with DQL—only better. DQL offers more flexibility, richer functionality, and a more expressive syntax for shaping and retrieving your data.

With that milestone achieved, we’re beginning the process of deprecating the Query Builder API. To make the transition as smooth as possible, we’ve published a migration guide to help you update your apps and take full advantage of what DQL offers today.

I see this as an exciting turning point. DQL is now the clear path forward for building fast, powerful, and maintainable queries in Ditto—and it’s only going to get better from here. I’m looking forward to showing you what’s coming next in future releases as our engineering team continues to expand DQL’s capabilities.