Devlog #1: got the bones working

The first real version of Rollforge is up. It is still very much a site I am building in public, which is a nicer way of saying there are probably edges I will notice five minutes after writing this. But the important bones are working.

I wanted to get to a point where I could add a new generator without it becoming a little web project every time. That meant doing enough setup to stay out of my own way, then stopping before the setup became the project.

What is actually live

The current version has a static Astro site aimed at Cloudflare Pages, a growing set of generators, and a generators hub where you can browse by category or filter down to something specific. That part matters because a site full of random tools gets annoying fast if you cannot find the one you need.

There is also this blog, which is plain local Markdown for now. That is intentional. I do not need a full CMS to write a few build notes, and keeping posts in the repo makes them easy to publish with the rest of the site. If that becomes annoying later, I can revisit it. Right now it is exactly enough.

I also added about and privacy pages. They are not exciting, but they should exist before a site starts pretending it is ready for ads or wider use. A small project can still be clear about what it does and does not collect.

Why Astro and a shared generator engine

Astro was a good fit because most of Rollforge is pages people read and buttons people click. The site can build into static files, stay lightweight, and deploy cleanly to Cloudflare Pages. I did not need to run a server just to assemble a random tavern name.

The more important choice is the shared generator engine. Each generator has its own content, rules, and personality, but they do not each need a whole separate interface. A fantasy name generator and a loot generator ask different questions, yet the basic job is the same: give someone a useful result quickly and let them roll again.

That separation makes adding things less painful. I can improve a word list, tune a result format, or add a new category without rebuilding the surrounding page every time. It is not glamorous code. It is the kind that lets me spend more time making tools and less time copying buttons around.

SEO and mobile choices that matter

I am trying to keep the pages light and direct. Each generator has a real URL, not a result trapped behind a single app screen, which makes the site easier to link to and easier for search engines to understand. A person looking for a fantasy tavern name should be able to land on that exact tool.

Mobile matters just as much. A lot of these tools are for the moment someone is in a game and needs an NPC name immediately. That is not when you want to pinch-zoom through a giant dashboard. The pages should load quickly, explain what they do, and put the generate button where you can find it.

I am not trying to perform SEO magic here. I just want clear pages, useful text, and a site that does not make people work for a random result.

What is next

The next thing is probably another generator. It might be a more specific name tool, a tiny game, or something dumb that only five people care about. I am deliberately not locking myself into a huge roadmap.

For now, the main win is that Rollforge has moved past the folder-of-ideas stage. There is a site, there are tools, and there is a place to write down what changes. You can browse the blog when you want the behind-the-scenes version, or just go generate something and get back to your game.