njaremko

Hey everyone, Nathan (other co-founder) also here.

We're bootstrapped, and on a mission to give people cheap fast CI.

We've already saved companies millions in CI costs (and cut CI times in a lot of cases).

I'll also be around to answer questions :)

This slipped a day because I want to make sure I'm handling all the edge cases with locales properly. Sorry about that.

Everything's from scratch, we've been working on this for a long time :)

Yep, if you run into issues let me know :)

- Today, you'd likely have the LLM write a script against the CLI, that you'd run whenever you want to refresh (assuming the query is fixed, given your example)

- Nobie will continue getting more customizable and programmable. You can see a hint of what's to come with our IDE inspired triple pane support with draggable UI elements. We'll get to a point in the near future where an LLM (or a developer/company) will make you the plugin and UI you want directly in Nobie, and you click a button to sync from your data source (or bind a to a key chord)

- We'll be putting out more content showing off Nobie. Too many screenshots felt distracting for the launch page.

- There's no GUI for it at the moment, but you can programmatically load data into Nobie. If you have access to a data store, and you use LLM tools, you can just tell it to load the data in, and it'll work. The target market is all spreadsheet users, so we've definitely put thought into this exact problem.

You're right, we'll update the script to help facilitate uninstall. You just need to delete `~/.nobie` and:

- Delete `/usr/local/bin/nobie`

- If you're on mac, you also need to delete the .app from `/Applications`

I'm one of those developers :)

I linked in this thread, and Matt replied above in this thread, but we charge companies making >$100M in revenue for using Nobie to provide a service to users, or for using it to help train models.

We've gone through a bunch of iterations. One of those iterations was canvas, but we care a lot about the user experience, and found getting the frame times we wanted on huge workbooks wasn't practical, so we graduated to webgpu/webgl, and that was better, but we kept hitting bugs in the web rendering engines that we couldn't tolerate, so now we've ended up going down to the metal and writing an engine on top of metal/vulkan/opengl (we still have a webgpu/webgl backend, but we're not using it at the moment).

You didn't ask a specific question, but it basically comes down to getting really good at calculating layout fast, getting good at figuring out how to incrementally maintain your rendered view, then getting really good at figuring out how to optimally use the hardware you're targeting.

I'm not sure I follow, but the math in Nobie ties with the math in Excel.

When you use codex / claude in our built-in terminal, the agents get very high signal context about the workbook and what you're doing in it.

As far as I'm aware, Nobie is the the best spreadsheet for agentic use-cases explicitly because they don't have to guess anymore.

Yes, you can print an xlsx to PDF as excel would if you printed the same file (respecting whatever print layout has been done in excel). It'll be vectorized, so you have infinite resolution on zoom for text.

We do not yet support editing print layout, so you can't adjust the pagination / change the page size in / with Nobie yet, but that will come soon :)

Yeah, we're not providing subsidized AI to people. It costs us nothing to give people a great, computer-use friendly, native spreadsheet app

We're saying we don't take your data. If you choose to use codex or claude, and send your data to them yourself, that's your choice.

You don't need to use AI. We're fine if you just want a really good, free, spreadsheet tool :)