New ask Hacker News story: Ask HN: Does OpenClaw need a re-architecture to be usable?
Ask HN: Does OpenClaw need a re-architecture to be usable?
4 by xinbenlv | 0 comments on Hacker News.
I’ve been using OpenClaw intensively for about two weeks. The first few days were exciting. It felt like we’re finally getting closer to autonomous agents that can actually operate a computer end-to-end. But after the initial excitement faded, I started noticing some consistent issues: - It frequently stops responding mid-task - Execution fails without clear recovery - Task success rate feels inconsistent and unpredictable - Long-running tasks degrade over time It made me wonder whether the current architecture is fundamentally limiting reliability. Right now, it feels closer to a “single program trying to do everything” model. But if we look at the history of computing, systems only became truly robust when we moved toward operating system–like abstractions: - event-driven execution - proper failure recovery - watchdog / heartbeat monitoring - task supervision trees - state persistence and resumability In other words, less like a script, more like an OS. My current hypothesis is that tools like OpenClaw might need a deeper re-architecture — not just better prompting or incremental patches — but a system-level rethink focused on reliability and scalability from day one. Curious what others think: Is this mainly an engineering maturity issue that will be fixed incrementally? Or is there a more fundamental architectural gap in current agent frameworks? Has anyone tried building agents with more OS-like supervision models? Would love to hear perspectives from people building in this space.
4 by xinbenlv | 0 comments on Hacker News.
I’ve been using OpenClaw intensively for about two weeks. The first few days were exciting. It felt like we’re finally getting closer to autonomous agents that can actually operate a computer end-to-end. But after the initial excitement faded, I started noticing some consistent issues: - It frequently stops responding mid-task - Execution fails without clear recovery - Task success rate feels inconsistent and unpredictable - Long-running tasks degrade over time It made me wonder whether the current architecture is fundamentally limiting reliability. Right now, it feels closer to a “single program trying to do everything” model. But if we look at the history of computing, systems only became truly robust when we moved toward operating system–like abstractions: - event-driven execution - proper failure recovery - watchdog / heartbeat monitoring - task supervision trees - state persistence and resumability In other words, less like a script, more like an OS. My current hypothesis is that tools like OpenClaw might need a deeper re-architecture — not just better prompting or incremental patches — but a system-level rethink focused on reliability and scalability from day one. Curious what others think: Is this mainly an engineering maturity issue that will be fixed incrementally? Or is there a more fundamental architectural gap in current agent frameworks? Has anyone tried building agents with more OS-like supervision models? Would love to hear perspectives from people building in this space.
Comments
Post a Comment