I've always wanted my own place to control on the internet. Couldn't find one, so I built one. This is going to be where I write and build little code experiments, and I'm excited about the possibilities.
Every couple of years, some new service or platform would catch my eye and I'd think to myself, 'nice, yeah, let's move everything over there and I'll finally be able to stop looking'. So I'd load up my digital hankerchief on a stick, collect all my disparate content, make the trek over to the new place and then find myself inevitably underwhelmed. I really liked a place called 'read.cv', before it was bought and killed by Perplexity in 2025. It had a nice mix of customisability, with a niche but engaged design-focused community. After it shuttered, I spent about a year wandering around until finally deciding to return 'home' and just build out my own place using all the weird half-finished stuff I had lying around.
Under the Hood
I'm writing this right now in my own bespoke Markdown editor that's built for maximum flexibility, anywhere I find myself with the urge to write. Nothing intrusive gets in my way, no silly AI-powered content scanner gets suggested to me, and most importantly I can control what features make it into the editor. Image upload is as simple as drag-and-drop from my desktop. It has a slack-like custom emoji service, so if I write :aah:, it properly shows up like 
All of this is being served by a Vercel project in the background. I've got two databases:
- Neon - backend data storage for both internal and public storage (auth, drafts, posts, emojis)
- Vercel Blob - public files storage (mostly images)
At work in late 2025, I started using Quick extensively, and I'm gonna be vulnerable and admit it probably taught me more about how backend engineering works in a few afternoons than years of tinkering ever did, at least in my world. I always assumed backend work was risky and needed really dedicated attention to manage, especially the security side. I think the elegant simplicity of Quick showed me that if you got the fundamentals right, you could effectively set it and forget it. Backend work, in my mind, became as much a part of the user experience as the frontend.
Obviously, another major catalyst to undertaking this work now was how good and easy Agentic coding has gotten. I've had a few years of really dedicated practice to catch up to the devs and engineers I've idolised for years. I believe I know what good looks like (not always immediately, but I'm willing to fail a lot to get there!) and my skillset really suits a fullstack approach. Plan, design, test, deploy, reconfigure, repeat. I wrote about that process more here.
For this project, I've got all my service credentials for the deploy and build pipeline controllable via any agent handling the changes. Local project files are committed to Github, which triggers a new deployment in Vercel automatically. Agents also have access to my Vercel CLI so they can create... 
new data tables? Whatever they need over there. None of my business, tbh.
Benefits of Control
While I was writing this post, I realised I kept uploading screenshots, then wanting to change them. This left a bunch of unused uploaded files in blob storage. That kinda thing was going to bother me long-term. So, while I kept writing, on the side I built a new Admin page to control uploads.
This took about 5 minutes. Most of that time was my agent smoke-testing the implementation by running through the full user journey to validate used images don't get purged during an unused image cleanup.
Really loving the speed at which I can experiment in this playground - I can dream up something locally and have it live in a new deployment in less than 15 minutes. The custom emoji build 
was exactly that, I wanted it and knew what it should look and feel like, and the longest effort was just me bringing in all the little .gifs and figuring out their unique handles.
What's Next
I want to consider the idea of open-sourcing this build. Self-hosting is definitely not for everyone, but I truly don't see how services that are charging monthly or yearly for subscriptions to host an online presence will survive long-term. This entire project has cost me... about $35US? Somewhere around there. So happy to invest that and come out the other side with a product I actually enjoy.
Trying to decide currently if it's uncouth to pull my email list that I've collected over on Substack for my creative-focused posts and fold it into this new home, with an email delivery service. I've seen other blogs rely instead on RSS feeds, which I also love for it's simplicity. I wish there was something in-between; approachable for anyone, but not as presumptuous as a direct email. Not sure what I'm looking for yet, honestly.
I can't wait to move over my main https://mackiec.xyz site, that's my next personal project. It's currently using Format which is a silly amount of money per year to host some images, but I've already paid for the year 
so the sunk cost fallacy is keeping me over there for now.