because you're ex YC i feel ok giving some tough love that i dont think this is a good plan because cursor and v0 will eat your lunch if you get any traction, however, happy koding and may you find the thing you were meant to do.

▲

jsunderland323 10 hours ago | parent [-]

I don’t mind the tough love one bit. Yeah, could happen. Startups just aren’t defensible creatures, especially in the post network effect world of AI. But having spent a bit of time in the world of startups (I worked at YC), I think it’s easy to forget that people build in an incumbent’s space all the time... and yes, usually those startups get killed. But by your logic, Cursor should not exist. We should all be using GitHub copilot. The browser IDE is really a new category so I think it will be somewhat awkward for any incumbent.

To be clear, I fully expect Vercel to launch something and for them to have the best distribution in the world. I don’t think they will ever support Vite or things that aren’t Next but que sera sera.

▲

ontouchstart 10 hours ago | parent [-]

It is going to be a very competitive market but you still have a chance if you serve the needs of professional React development. Unlike POC and MVP projects by people who are learning React, professional React projects that serve production are very complex. Developers need to deliver features without regression, debug with whatever tools they have to find out root cause and fix the bug without causing regression and more issues. Tight unit tests, integration tests and QA cycles.

I have been using Cursor + Claude (Composer) in production code base with some success. Integrations with live React DevTools in the browser could make UI debugging and iterative development much faster. But you also need to think outside of the box of IDE. It is never about IDE. It is about high quality workflow.

▲

jsunderland323 9 hours ago | parent [-]

>you still have a chance if you serve the needs of professional React development

That's my entire thesis. I think it's a long long road. But I think you've actually done a great job of pretty much seeing the whole vision.

▲

ontouchstart 9 hours ago | parent [-]

That is why Playwright might be a critical piece of the puzzle instead of good to have.

I have instructed agent to generate Playwright test from screenshots that will dump API calls to debug and generate mocks to trigger the edge cases bugs. I am sure if jstool is there, it can fix the bug and reload.

▲

jsunderland323 9 hours ago | parent [-]

To be clear, I'm with you 100%. You have my word that I will add in playwright when we get to a point where we can support it.

	▲	ontouchstart 9 hours ago \| parent [-]
		If you can build browser automation into jsxtool, perhaps it can replace playwright. If I remember correctly, the guys who developed puppeteer/playwright used to work for chrome devTools or something.