Hello, am I crazy to say that based on official changelog and my personal use of Cursor, the IDE itself has been overlooked for a long time? Last actual new feature for the IDE specifically was released in May, which is now 3 months ago. I personally dont like the fancy automations, slack, ipad things etc. I just use the IDE and it is lacking behind competition. Does anybody feel the same way?
All too familiar pattern. Shiny things like iPad’s first, features and bug fixes last. Support for the Apple Watch next… I can still work fine but man that is such a common issue. Whole team goes off chasing the goose while the egg sits cold and never hatches. Also Electron is a total hog. Yes Electron is great for some things but it has so many open issues that are outside of the IDE providers scope that they rely on the Electron team to fix them before the end User gets the fix for the software they use which is built upon it… Again all too common. Waiting for the Electron team to fix… Result equals frustration from Users of the tools built upon it. Cursor being built upon Electron is a real negative in that regard.
Electron team? Usually companies dont do that. Is there a bug with https://www.electronjs.org/ that we should know about? Btw Electron isn’t a team as much as it is an open source solution and I use it for a few things myself. Curious what should this team be fixing? It would be great to learn more because I’m always looking to get better at this.
Enterprise features might seem like a miss but I bet they are high priority asks for their biggest accounts. I worked at tableau during a time where they went between community based requests and enterprise. It’s always been like this, even ten years ago it was like this. Best thing to do is if you have a feature idea, make it known, get people to add some fire to the idea, and offer the alternative if it’s not around. That’s called the workaround.
Hope this helps, it really applies a lot to my life and I’m sure I’m not the only one who has gone through tragedy!
Entirely but still a team of contributors.
Never thought about looking at issues here, thanks for the help. Been using electron for awhile now, not having any issues a good engineer can’t figure out. I’m not talking about myself.
edit; briefly overviewing the titles, i could be wrong, but it looks like a lot of user errors directed in wrong direction.
This is a common occurrence with issues raised on GitHub. It all becomes quite messy and I pity the developers having to sort through it all now given the AI generated errors on top of human reported errors that may be similar errors or the same as others but wording makes them appear they are new errors. What a headache.
My point about Electron and Cursor being built upon it is that its like building the most amazing home but on top of shifting tectonic plates. Above ground looks great on the exterior but below ground and internally not so great (with big cracks in the walls and floors and foundations of the building beginning to fracture). I’ve reported an issue recently where CPU spikes are significant when the context of a chat is not even that great. Other applications (CLIs) can handle this no problem at all but GUI’s built upon Electron just struggle. Same with other software I use built upon Electron so it is very familiar as the similarities are so obvious. High CPU load matched with increased memory load over a sustained (not even long lived) period of usage. Like I say CLI’s no problem at all. My CPU doesn’t even register a difference essentially. Chuck Electron on top as the GUI wrapper and fans go brrrrrrrrrrrrrrrrr. My CPU is not that old and my laptop is relatively decent so it shouldn’t be getting hit 40% of CPU with memory use all over the place.
- 32gb of 7467 MT/s LPDDR5.
- Intel Core Ultra 9 185H × 22 cores.
- Ubuntu 26.04 LTS.
The memory is fine for the most part that never exceeds 20/30% or so but my CPU is getting punished. I’m only running Cursor with no browsers open and no other applications (SSH in to another laptop and use a remote session for Docker etc). Cursor just seems to be incredibly CPU intensive right now and like I say the context is not even that large. So I hope the Cursor team rectify it or at least consider a rewrite away from Electron (which is a pipe dream), but the Electron side, well that is up to Electron to rectify although I’m sure the Cursor team must be in discussion with them…
Hi all, thanks for sharing your thoughts here.
You’re correct that most of our interface work is going into the Agents window rather than the IDE.
The IDE isn’t being wound down. There are still many important features there. Many of which we bring into the Agents Window, but some features remain in the IDE only like certain extensions.
We’ve also been focusing a lot on developing our own models e,g. Composer 2.5 and Cursor Grok 4.5. We’re working hard to make these the best they can be. Some of the prioritization can make it feel like the IDE is receiving less attention in terms of features, a fair critique. We’ve also been putting a lot of effort into improving cloud agents both in terms of product surface area and reliability/seamlessness of the experience.
tbh, they haven’t shifted that much and I still have an app running on 6 year old electron code, despite how crazy that sounds. I have a ton of experience with electron and I think it’s just more of a best practice given where we are at with tech and the desires for cross platform native web apps. I hear you, there’s a churn on the CPU that feels uninspiring. That’s why I went and made my own code editor the way I wanted to do it, and I think if you did the same you’d be surprised by how fast you’d get there. I’m not saying i got rid of cursor/claude, I just instead built something I needed that fits into that paradigm.