It’s been a long, long year. So much has changed and so much has stayed the same. If you’re like me, you likely stayed up late to watch 2025 die a happy death.
But, in the world of technology and software — there was lots to talk about too. From my perspective, I have some opinions about what the last year brought us. Let’s talk a look:
Front-end Web Development
We began 2025 with four technologies used the most as front-ends:
- React
- Angular
- Vue
- Svelte
This hasn’t changed a lot. Each of the libraries have incrementally changed. If you’re using one of these platforms, you’re in a good stead of common web projects.
Meta-Frameworks
What has changed this year is that it feels like development is moving more towards meta-frameworks like Next, Nuxt, and Analog. But it has to be said, it matters a lot what you’re building, but meta-frameworks are a great solution for building line-of-business apps.
Plain JavaScript
The other solution for web development that is gaining in popularity is Plain JavaScript. The browser JavaScript version are incredibly well heeled at this point, and it is possible to create complex apps without platforms or meta-frameworks. Though, I think it still involves a lot of work.
Vite
One winner this year continues to be Vite. More and more, Vite is the center of most platforms and meta-frameworks. With it’s great development experience, and pluggable system for using your own compilation step (i.e. replacing Rollup if you want), Vite continues to thrive.
What About Web Components?
Lastly, I wanted to mention Web Components. Every year, it seems, Web Components are set to take over front-end development. But again, a year has gone by and it continues to be a small community around it. I still think it’s got a lot to offer, but there isn’t a single, great solution to building them.
Mobile Development
I’ll admit, I don’t really build anything for mobile devices any longer, but I continue to keep my ear to the ground as it’s so important to most architectures.
As many know, you can no longer deploy iOS apps with Xamarin now that it’s at end-of-life. For many Xamarin developers, MAUI is the obvious choice. While Microsoft has been making big strides in making MAUI work for mobile developers, I think it has upset a lot of them.
I’ve heard many of these people just moving to other platforms like React Native and Flutter. While both of these require a completely re-write, them seem more mature for experienced mobile developers.
I don’t work with many developers that work with the tons of other platforms for mobile (e.g. Swift), so I don’t know how those have changed this year.
.NET Got a New Version
In it’s usual cadence, Microsoft released a new version of .NET 10 (and C#). This is a long-term supported version, so many projects are upgrading to the latest version. There are too many features to outline here (I did do a talk at OreDev that you can watch here).
There are a couple features I’ll mention here that are important but maybe missed by some:
- The new Garbage Collector (DATAS) that handles smaller workloads much better, is now the default. It seems to fit containers much better since those loads are often running on limited memory.
- New support for Contains clause in Entity Framework 10. (Coverts collection of values to an
INclause). - Validation support in Minimal APIs.
- C# adds Extension members and the new
fieldkeyword.
Aspire Matured
Aspire got a lot more mature this year. It finally broke out of it’s .NET shell and now works in other ecosystems. Some of the changes that Aspire brought this year that I love include:
- Aspire CLI support.
- Single-File App support for AppHost.
- Python app support.
- Improved Node/JavaScript app support.
- MCP Support for CoPilot.
- More deployment targets.
Artificial Intelligence (mostly LLMs)
This is probably the biggest changes this year. AI has promised a lot for developers and the general public. But in some ways it’s started to fulfill those promises for developers.
In 2024, the idea of prompt engineering replacing developers was all the talk. Lots of companies took these promises to heart and started to cut developer headcount. But I think in 2025, they started to see the real benefits of using AI in software development.
From the developer’s perspective, lots of us have been using Agent-Mode to help us build, refactor, and modify our codebases. Has it removed the need for developers? No. But, like every page in the history of software, it has made us more productive.
Just like higher level languages (e.g. C++), garbage collected languages (e.g. Java, C#), and even containers have made it much easier to build complex systems. We no longer think of these as reducing the number of developers, but enabling us to build better systems. We take these advances for granted. We just use these tools without thinking. I suspect that AI agents will be part of our tool chain for some time.
I shared my thoughts on this two years ago in one of my Rant Videos. Ultimately, I haven’t changed my mind much. Code generated by AI models still needs to be validated. Once we generate code, we own it. Just like you’d do a code review before developers check in code; we should be treating generated code the same.
Final Thoughts
There is a lot of fear among developers right now. I don’t think this is going away in 2026. This reminds of of 2008 and 2000 when there were big developer purges. I’m hearing of a lot of people that have been without work for over a year. This is a scary time.
My only advice is to keep in touch with your community. It’s time to go back to user groups. Keep up with your friends from prior jobs. I think that’s key to most of us to survive other year of chaotic hiring.
What do you think?