Heya friend!

Carl here. You signed up to receive updates about Ymir, the WordPress serverless DevOps platform that I’m building.

INTRO

We did it fam! I finally have the second Ymir case study published. 🥳

It's been a mad two weeks for me with Ymir. I've basically decided to go all in with supporting project types outside WordPress. I was aiming for Drupal next, but things are moving in Laravel land so I'm thinking of doing Laravel first. (More on that later.)

​Radicle support hasn't launched officially yet because I'm refactoring a lot of the code. I didn't architect Ymir to be a serverless PHP platform. And moving away from that architectural decision is delicate process.

Business is ok. I lost a customer. I'm trying to be more active marketing, but it's hard. I'll talk about it more inside.

PRODUCT

You can always view the history of Ymir's product development at https://ymirapp.com/changelog.

Lots of work is happening on the product right now. I'm focusing everything on moving away from just supporting WordPress. It's a huge lift like I mentioned. I'd never really thought of going outside of the WordPress ecosystem when I started working on Ymir.

One of the two areas that needed the most work was the CLI. I completed a huge refactor earlier in the week. It's now super easy to add project types. I've also tripled the number of tests in the CLI. That said, the CLI probably needs 4-5 times more tests than it has right now.

This refactor was also an interesting experience for another reason. I use AI a lot for work, but I don't think it's good at refactoring. I read an amazing article (I recommend everyone to read it) on AI coding workflow that discussed using AI for refactoring. (This is super rare. Most AI codegen examples are for new projects.)

I tried it out for this refactor. The result was still underwhelming. It got maybe 15% of what really needed to be refactored. It couldn't to generate good tests as well and hallucinated a lot.

I learned a lot from the experience. I plan to give it another go when I dedicate time to expanding the CLI test suite.

MARKETING

So I've published the case study for Open House New York. 40Q and myself are very happy with how it turned out. OHNY feels great about Ymir too. I'm trying to get a quote from them as well. I just decided to publish the case study without it for now.

After some heavy push from a friend, I'm trying to be more active on the marketing side. I wanted to start some discussions on Reddit, but it's honestly not easy. It's essentially self-promotion even if I'm trying to start a discussion.

I'm trying to inject myself in the Laravel subreddit after the launch of Laravel Cloud and the impact on Vapor. (More on that in the business section.) But I don't have the karma yet because I'm a lurker 😅

BUSINESS

You can always view Ymir's up-to-date business metrics at ymirapp.com/open. They're updated every 10 minutes.

Ok, there's a lot to discuss here as well. The easy, but bummer stuff is I lost a customer. They were actually using the product. So not sure what happened there, tried to see by email, but no answer sadly.

The larger discussion is revolves around Laravel. Last week, Laravel Cloud launched with the new Laravel site. The new site was a bit polarizing, but, more importantly, you couldn't find any mention of Laravel Vapor on it.

This led to a Reddit post asking if it was deprecated. There were some tweets on it as well. Officially, they say they're supporting the product still. In practice, it doesn't feel that way. (There was one product update officially last year.)

So now, I'm thinking maybe I need to support Laravel. Not only because there's a customer base that exists and isn't being served well anymore. But there's also the issue that Ymir also uses Vapor and I need to be ready if something does happen.

It makes business sense that Laravel needs to focus on Laravel Cloud. They took $57M in funding. They need a product that makes them a lot more money to match what the investors expect.

But does that mean that mean that the Laravel community doesn't like serverless anymore? I've been trying to figure that out. I wanted to post on the Laravel subreddit about it, but I can't unfortunately. For now, I'm discussing things in the Serverless Laravel course discord.

I don't have the ability to move on this fast. It'll be a few months before I can really host Ymir with Ymir. So I'm trying to see if it's worth that I go for it. I think it is, but I'm not the best at business (as my journey shows!) so maybe it's not worth it.

To be continued...

Carl