Also I don’t know why but on suggested update by cursor is displayed v0.48, but after the update goes to 0.47.8, really hope that issue will be fixed in 48
Still having the same issue, even on 1.0 it doesn’t work anywhere near as well as on 0.46. Created my own post to try to get some support, but nothing helpful yet
@ajmeese7 I had the same issue, tracked it down to state.vscdb freezing my whole PC when being updated. My process monitor was showing massive writes to that sqlite3 database, even when it was only couple of MB with size it was still crippling a 16core 7950x with 128GB RAM… What I did was moved it to tmp and made an autostart script that copies the database file to /tmp which is stored in RAM memory and periodically does a backup where the original file is stored. I don’t encounter any freezes at all now.
Constantly freezing on Linux… seems to be linked to the amount of time it’s been running, or quite possibly the length of a conversation in progress. Longer conversations lag up the editor more. (yes I know it culls past a few past messages, but still seems linked)
It’s consuming a lot of hard drive, When working with fast ai like Claude 3.7 sonnet,it uses the hard drive to full potential 95% active time, I don’t know if it because it is creating backups or what. But that is my case, ram and CPU is okey.
Having same issue, even as i type the prompt if ofter freezes for a few seconds, and when switching in Windows between apps its a sure thing IDE will take 10-20 seconds to show up & unfreeze. Any help how to identify the issue ?
I’ve noticed that Cursor AI sometimes gets stuck processing a request for more than 20 seconds. There are no internet connectivity issues, and my system is working fine. This seems to happen intermittently and affects productivity. Has anyone else experienced this, or is there a known fix or explanation?
No long chats involved. I checked Activity Monitor — it’s not consuming much RAM. As shown in the attached screenshot, it’s stuck at this point. The only workaround is to stop and re-run the request.
OK thats a different case and topic. Could you please file a full separate bug report so this can be investigated in detail? If possible please attach Request ID there.