最新

I&#x27;ve seen a few anecdotes lately that say that they use Claude Code on legacy codebase and with relatively little supervision it can work on complex problems. Then the claim that Claude Code writes most of its own code, and that they no longer mentor their newcomers - instead, AI answers their questions and they can start making meaningful changes within the first few days. To me it sounds almost too good to be true, so I&#x27;d love to have some reality check.<p>I&#x27;ve spent most of my career in legacy codebases, reading, tracing behavior, making careful changes, and writing tests to protect them. I&#x27;ve taken a sabbatical though, which ends soon, and I&#x27;m quite worried and excited to what has happened during this time.<p>For those working on legacy codebases:<p>- Has the workflow really shifted to prompting AI, reviewing output, and maintaining .md instructions?<p>- Does your company allow Claude Code, Codex or similar tools? If not, what do you use?<p>- Do companies worry about costs and code privacy?<p>- Where do you think this is headed, a year from now?<p>Concrete examples, good or bad, would be especially helpful. Thanks.

My name is Adrian. I&#x27;m a Software Engineer and I spent 6 years developing a perimeter-based geofence-based social media app.<p>What it does:<p>- Allows you to load a custom perimeter anywhere on the geographic map (180° E and W longitude and 90° N and S latitude), to cover area any area of interest<p>- Chat rooms get loaded within the perimeter<p>- You can chat with people within the perimeter<p>I developed a mobile app that uses an advanced geofence-based networking system from 2013 to 2019. My goal was to connect uses within polygon geofences anywhere in the world. The app is capable of loading millions of polygon geofences anywhere in the world.<p><a href="https:&#x2F;&#x2F;enterpriseandroidfoundation.com&#x2F;assets&#x2F;images&#x2F;other&#x2F;manhattan-polygon-geofences.png" rel="nofollow">https:&#x2F;&#x2F;enterpriseandroidfoundation.com&#x2F;assets&#x2F;images&#x2F;other&#x2F;...</a><p>But people didn&#x27;t really have a need for this. So after failing, I spent the next 6 years trying to ideas to use FencedIn for. I tried a location-based video app and a place-based app that had multiple features. Nothing worked, but now I&#x27;m almost finished developing ChatLocal, an app that allows you to load a perimeter anywhere on the geographic map, which loads chat rooms.<p>The tech stack is 100% Java (low-level mostly). I have a backend, commons library and an Android app. Java was the natural choice back in 2013. However, I still wouldn&#x27;t choose anything else today. Java is the best for long-term large-scale projects. (I&#x27;m also using WildFly. PostgreSQL and a Linux server.)<p>This app is still not fully finished, but I think the impact on society might be tremendous.<p>The previous app to ChatLocal, LocalVideo, is fully up on the Google Play store and can be tested. It has 88% of the features of ChatLocal, including especially the perimeter-based loading system.<p>The feedback I&#x27;m mostly looking for is new ideas and concepts to add to this location-based social media app. And how strong of a value proposition does the app have for society.