yakkomajuri

Fair question! We're a much younger company and thus more malleable. Val Town was started basically pre-AI and had early traction that might have locked it into its path a little bit.

We're still figuring things out and this announcement is yet another data point about the simplified infra for devs space maybe not being very promising but most of our customers at the moment are actually non-technical teams who have struggled to use anything else. You could argue in some ways Val Town was tangentially in the same space as Lovable & Co and those companies are still going strong.

I think there's real signal in this announcement but it's also hard to derive general lessons for a whole market from a single company given how many variables come into play.

This is sad. We're in a similar space as Val Town (which technically made us competitors) but I've always appreciated the care they had for the craft and the values they drove forward as a company like their transparency (even their investor updates were public).

Steve and Tom also individually have some of the best blogs out on the web.

I wish them all the best with whatever comes next. I'm pretty sure they'll all do great things.

My main side project at the moment is an AI workspace that's fully customizable.

I've done 80-90% of my work from inside it the past two weeks and made it open source, but haven't made much noise about it yet.

https://lynx.railcode.dev

Was surprised to find Andorra grouped in here. If you're into nature it's a whole different story!

There's really amazing hiking out there.

I've also disabled memory in Claude for finding it more annoying than helpful.

However, I have it on for ChatGPT (use it mostly for on the go queries via mobile) and their memory experience is significantly better. I haven't enabled Claude's memory in a long time so things may have changed though.

If I'm reading this right, this is (really solid) future planning but won't help with the current outages?

I'd have thought so from the headline + initial content but then they say that the horizontal scale techniques are basically carrying over from the existing architecture (where it seems the problem is?) and this is mostly to help with what they call "vertical scale" (intra-repo scalability). They also admit right after that the most active repos today will receive short bursts of at most dozens of writes.

So this seems like really useful planning for a near future that likely helps prevent more issues but isn't really meant to help current issues. Am I reading this right?

Of course improvements like this are always useful either way. Not dunking here at all.

On occasion I get a glimpse at someone else's experience on the web and I'm so thankful that I haven't seen many ads at all in years.

I just disabled my adblocker on the BBC website out of curiosity and oh am I happy I don't live like this.

1. Claude $200 + Codex $100. I never hit limits since realizing that trying to have too many agents running at once just wasn't effective. I'm probably overpaying.

2. No

3. Lynx (lynx.railcode.dev) & CLIs (less and less since Lynx)

4. No

5. Tried a few but never stuck around

This is insane!

I also love the presentation on this, which is probably a byproduct of LLM usage (in a good way). As someone who doesn't actively read academic papers I'd definitely read more of them if they were presented like this.

Small clarification: some people in the thread seem to think I'm saying Bitcoin will no longer exist or will tank in value. I'm actually not making a point about the price of crypto at all, just its usefulness as currency.

The price seems to be fully independent of that and as someone who diversifies their investments I own a little bit of crypto and would like to see it go up :)

I think we can probably declare crypto as failed now.

Note that I was one of the folks early on that bought into the idea (but not as much of the coins sadly) and I still think the Bitcoin white paper is a beautiful intellectual exploration (as is the concept of blockchains).

But the centralized entry and exit points in practice mean all the theoretical benefits aren't really in play.

Yeah and they will happily give you all the context you need about their experience

I mean, fair play? I appreciate marketing that has effort involved, especially these days.

You do have to get your word out somehow

Hey, thank you! I think you've hit at the heart of what has been a struggle of mine which is what to call this thing.

Originally Railcode started out as an AI-native Retool let's say. But while Retool added an MCP and now supports custom code they're still a reasonably rigid platform. I don't really believe a platform that's rigid works out today, and I'm also all-in on the user's own harness [1]. As a result we ended up building something that's a lot closer to Vercel as far as the flexibility and ownership you have, but with the governance layer of Retool.

I'd say we're technically a layer above plain compute.

In practice:

- Projects are just code. We can host static frontends and JavaScript serverless functions for the backend.

- When you deploy an app it's immediately behind company auth

- If you want to connect some data, you can e.g. connect an MCP and allow the Growth team to use that connection. Projects the Growth team builds can then use railcode.connector('posthog') and no API keys are ever shared

- If you connect a warehouse, you can create specific queries and assign them to users or roles to be able to build on top of, same model as connectors

- If you need to store some data, you can just do railcode.db.collections.create('orders') and it works. For an LLM it's railcode.llm.generate and admins can configure keys and manage spend centrally

- You can set a design system that agents follow when building apps in Railcode

Non-technical users don't need to manage infra or spin up a database (like they would on Vercel) and admins retain control and observability over everything that gets built.

Companies have started to build some version of this internally on top of Vercel, Railway, and Cloudflare, so I'd say we're somewhere in between an application layer thing and compute.

Genuinely curious: what would you call it if not a cloud?

[1] https://railcode.dev/blog/best-harness-is-your-own

Something I've been thinking a lot about recently is how much we still need UIs vs if we'll just interact with agents.

We have a platform that enables building both and I still find UIs to be quite relevant. Especially if it's so easy to build like in this case because you can always throw together a visualization for something quickly and then never look at it again.

This looks really cool and because I'm in a tangential space I know how fun and useful it's been to build and host stuff on the web quickly.

We're geared towards fully-fledged (serverless) apps and agents for internal use only though, in case it's of interest to anyone: https://railcode.dev/

Not at all! I agree with what you said

This makes sense yeah, I guess following a similar approach used on the buttons for the tabs as well. Point taken!

It definitely does. Particularly under the hero imo.

Doing this is really time-consuming so certainly there were some corners cut -- need to also focus on product after all!

Particularly after we changed a lot of the messaging (after this post was written) I slacked off a bit. Keen to hear what's most annoying to you though, always looking for feedback.

ah thanks for the feedback!

AI is just another tool and I have actually found it really fun to dive into design as a non-designer. It still takes a lot of time and effort but I can now create designs that I'm quite happy with that I'd have never been able to do before.

Wrote a bit about my process in a blog post [1]. Obviously my designs are far from perfect but I'm really happy with my company's website.

[1] https://railcode.dev/blog/vibe-coded-website

ofc we meet in the ClickHouse thread

> "Every single company handling ClickHouse® struggles with ingestion."

Very true. Reading about "too many parts" gave me flashbacks.

(previously owned ingestion into CH at PostHog, no longer)

The lack of any decoration is what struck me.

Otherwise really cool! Enjoyed the FAQ and appreciate the courage to put this out given the types of comments one receives online (see thread).

The resumes are impressive :) didn't realize Kenton had created Cap'n Proto

Ha interesting! Well I don't hate it but much prefer the current version ofc

But that's how taste works! It's personal

Well yeah I liked it as well at the time but also it feels a bit out of place? Like what does a rollercoaster have to do with this product?

Thank you! Yeah it's all part of the process

Oh yeah I mean check these out:

https://imgur.com/a/kLKTtAz

https://imgur.com/a/9fidnRv

The rollercoaster on the second one was the breakthrough that then eventually led to what I have today

Interestingly, I've found that AI has actually made me more interested in design because I'm now able to actually do some things, whereas before I couldn't get my creative visions built.

As an example, I recently built my company's website and I'm really happy with the result. I wanted it to have some hints of Rollercoaster Tycoon, and with a lot of experimentation, direction, tinkering, and examples I got to something that I really liked. Two years ago my company website would have been some generic Bootstrap thing until I hired a designer.

There's a lot to improve and certainly some aspects of it feel like AI still but it's generally gotten a lot of positive feedback and above all I really like it. Considering writing a blog post about my creative process here because I'm a dev with no design background at all.

Sharing a screenshot as to not promote: https://imgur.com/a/xqKi6X6 (in the website everything in the park moves btw)

Finnish libraries are fantastic. Many had free-to-use 3D printers as far back as 2012!

Libraries are a place of possibilities and fun, and it makes people want to be there. You can imagine the long-term positive impact this has.

I love this. My very own static site generator, Teeny [1], is going on five years now. It's 350 lines of hand-written JS and has powered all my blogs and personal websites for years. It currently powers https://yakko.dev.

I added plugins this year which made it really powerful and allows me to keep the core small.

[1] https://github.com/yakkomajuri/teeny