A Letter From
The Workshop

3 A.M.

It is three in the morning as I write this, which is either the worst or the only honest time to write something like it. The build server is humming in the corner of the office doing its fourteenth full deploy of the day, the coffee went cold two refactors ago, and somewhere across the internet a streamer named N1Hawk is probably live right now, unaware that the website behind his community has quietly been taken apart and put back together again like an engine block on a garage floor.

You deserve better than a changelog for what just happened here. Changelogs are for people who already know the story. Most of you don’t, because most of you were never told one. For months this site simply was — sometimes slow, sometimes strange, occasionally wearing a Halloween costume in July because somebody left an old stylesheet loaded. You clicked things. Some of them worked.

So before anything else: this letter is the story I owe you. It’s long. It wanders, because that’s how the work actually went. And it ends in a place that might surprise you, because the honest ending of six months of building is not a roadmap.

It’s a shrug, and a grin, and a promise to keep going anyway.

What Happened While I Was Quiet

Let me tell you the part nobody says out loud about software: sometimes the studio building your favorite thing has other favorite things. Vertali Cloudworks — my shop — had client commitments pile up the way they do, and this site, through no fault of its own or yours, slipped down the stack for a while. Weeks became months. The dust didn’t gather so much as settle in.

Some of you have asked whether Hamish Robertson — Cloudworks’ CEO, and the reason this studio exists at all — has quietly left the picture. He hasn’t. He and his family are battling some demons right now, the kind that don’t care about deploy schedules, and the rest of us are holding the fort while he handles the things that matter more than any website ever will. He’s still around, he still reads everything we ship, and his chair stays exactly where it is for as long as his family needs.

When I finally came back and opened the codebase like a pilot opening a hangar door after winter, I found things I’d genuinely forgotten writing. A leaderboard with its own quest system. Two different navigation systems fighting over the same sidebar. A page that rendered fine on my machine and wore a pumpkin theme on everyone else’s. Feature pages whose entire data source was a hard-coded list of fictional people with invented loyalty stats. That last one still embarrasses me more than I can properly express.

The honest diagnosis was this: the site hadn’t been maintained so much as accreted. Layer on layer on layer, each added in good faith during some earlier sprint, none ever removed, because removal doesn’t demo well. Every old page was a little promise someone made once. But promises rot when nobody keeps them, and keeping nine half-promises is worse than keeping three real ones.

So we made a decision that felt terrifying and turned out to be the only sane move available:

What follows is what that looked like.

What Actually Shipped

Here is the part where I get to brag a little, because the crew earned it and honestly? Nobody else is going to list these things, and they should exist somewhere outside of git history.

• A real store. Product detail pages, a multi-item cart, prices that live in a database instead of trusting the browser, checkout handed to Stripe’s hosted page with shipping collection and promo codes, and a webhook that records every sale into our own order ledger. When the keys drop in, gear goes live without another line of code changing.
• Three separate dashboards. Crew Quarters for every viewer — points, streaks, channel relationship, connection health, a readiness score computed from real signals instead of invented ones. A Broadcaster Hub with a working question moderation queue. A Command Center with sales, catalog status and role management. Each walled off by role checks enforced server-side, each wearing the same amber-and-charcoal design language.The old portal had two navigations rendering at once. That’s not a metaphor. It was literally two sidebars.
• A song request bay that actually breathes. The queue used to re-render itself every fifteen seconds with markup from a stylesheet that wasn’t even loaded anymore. Now the live updates and the page speak the same language, and the polling endpoint that had been silently 404-ing since forever finally answers.
• A subscriber wall that stopped lying. When Twitch data went down, the old page quietly swapped in eight imaginary subscribers with fake tenure stats. It shows an honest syncing state now. Honesty is a feature. It ships.
• A contact form worth submitting. Amber glass fields, segmented urgency pills, a rebuilt notification email that no longer carries branding from a company phase that ended two years ago.
• Sponsors, sponsors’ images, and the truth about both. Every sponsor asset now lives on our servers instead of hotlinking Wikimedia and praying. Gatorade’s logo had been broken for who knows how long.
• A maintenance experience people might actually enjoy. Live radar, countdowns, chat embed — because if you’re going to lock people out, at least be charming about it.
• An archive of missions past, a rig page for the hardware curious, emote walls with click-to-copy, event exports for your calendar, and a login screen that finally looks like it belongs to this decade.

And underneath all of it, less visible but more important: the whole site now speaks one visual dialect. Warm amber on near-black, glass panels, thin amber scrollbars, hover states that behave. The blue era is over. If you find a page still wearing it, that’s a bug, and I want to hear about it personally.

The Sunset List

Now the harder paragraph. Features are easy to announce; funerals are not. But if this letter is going to be honest, then it has to read like one, and honesty means naming names. The following pages and widgets are being retired from n1hawk.com. Some already have. Some will over the coming weeks. None of them failed exactly — they just stopped earning their complexity.

Since not everyone speaks web dev: to sunset a feature means retiring it on purpose — switching it off with care while it still works, rather than letting it rot until it breaks. Nobody’s mad at the sun for going down; it’s just how the day ends with dignity. That’s the spirit here.

• Polls & PredictionsTwitch built these into the platform itself, better integrated than ours ever was. There is no shame in losing to the house.
• Hype Train trackerThe train now lives where trains belong — on stream, native, animated by Twitch rather than mirrored by us.
• Raids tracker, Milestones & Bits leaderboardLoyalty numbers now live where they matter: in the Crew Quarters dashboard, tied to real accounts instead of parallel bookkeeping.
• The analytics suiteFour dashboards of charts almost nobody opened. The truth is the stream’s own analytics do this job; ours mostly duplicated them with extra latency.
• Chat replay & stream health pagesDeveloper tools that escaped into production. They’ve been returned to the workshop drawer where they belonged.
• The game page & the troll pageIn-jokes that outlived their moment. Some things should be stories, not URLs.

There’s context worth having here. Much of what powered these widgets is being contributed upstream into InnoBot, the streaming toolkit from Ingenious Mind Engineering — meaning the capabilities survive as reusable modules even as their standalone pages go dark. Consolidation isn’t deletion. It’s the same energy, pointed at fewer targets.

If one of your favorites is on that list and it stings a little — good. It stung to write. Tell me which one and why. Sunset decisions are reversible when someone makes the case with feeling. That has literally happened before, and the person was right.

No Map

Here is the section I’ve started and deleted eleven times, because every version sounded like a press release, and press releases are how developers lie by accident.

The truth is this: this website has no five-year plan. No north star. No grand vision slide. There is no document in our drive titled n1hawk.com — end state, and there never will be, and I want to explain why that’s a feature of how we work rather than an absence.

We tried roadmaps. Early on, there was a lovely plan with quarters and milestones, and within eight weeks reality had eaten all of it: a stream format changed, a game exploded in popularity, a widget everybody ignored suddenly mattered, a widget everybody loved quietly died. Roadmaps written for a community are forecasts, and communities refuse to be forecast.

So the site now runs on something simpler: a loop. Fly. Watch what the crew actually uses. Fix what annoys them first. Build the smallest next thing the stream genuinely needs. Kill whatever stopped pulling its weight. Repeat. No ceremony. The direction of this site is set, week to week, by the community that streams into it — which means the honest answer to “where is this going?” is toward wherever you all push it.

One more myth to retire while we’re here. Every few weeks someone says some version of there’s still more left to add, isn’t there? You must know where this ends. We don’t — there is no end goal, no final version, no finish line, no done. And that’s exactly why your ideas matter here: if you can see something we’re missing, don’t keep it to yourself. Tell us. When there’s no map, the crew’s eyes are the sharpest navigation we have.

Some of you will find that liberating. Some of you will find it unambitious, and you’re allowed to. But I’ve watched too many projects die of vision poisoning — suffocated by plans so grand that nothing small ever escaped them — to worship at that altar again. This site will earn its future one useful Tuesday at a time, same as the stream does.

There’s a distinction worth drawing while we’re being honest: we are developers, not architects. We don’t draw blueprints for a living — we build. And yes, let’s be bold about it: some of what ships here, nobody asked for. We added it because living inside this site every night teaches you things a suggestion box never hears. But there’s a line we won’t cross, and it’s this one: we will never make two websites the same. No cloning features because another streamer has them, no communities stamped from a template by the dozen. A site isn’t a checklist — it’s a portrait of one person and their crew, and a portrait painted from somebody else’s photograph is worthless. Which is exactly why the direction can’t come from us. It has to come from the person it’s for. We can swing the hammer all day — only you know what house you want to live in.

That’s the whole secret. Sorry it wasn’t sexier.

What I Need From You

A builder without feedback is just a guy guessing loudly, so here’s my open request list — the actual asks, numbered, no marketing gloss.

One. Fair warning: this site is built by a small team in our spare time — evenings and weekends, between real life — and it’s still in beta, so some things will break. No need to report every wobble; odds are it’s already on the whiteboard. But when something breaks badly enough to ruin your visit, tell us through the contact page and be specific. “Site broke” burns an evening nobody has; “the gift redeem bar looks wrong on iPhone” takes ten minutes anyone can spare. Specific bug reports are love letters, promise.

Two. When a sunset hurts, say so, and make the case. I listed the dead above precisely so you can argue with the list. The strongest argument wins, and I have reversed these calls before.

Three. Tell me what you’d use that doesn’t exist yet. Not features you’ve seen elsewhere — things you would genuinely open on a Tuesday. That’s the entire input side of the loop now.

Four. Patience with the edges. The store goes fully live the moment Stripe credentials land in the workshop. The gift-card pipeline is wired and waiting on its service. These are known, tracked, and closer than they’ve ever been.

And one promise in return, since asks should be symmetrical: I will keep writing these. Shorter ones usually. Whenever something big shifts, you hear it from me here first — straight, unpolished, at whatever hour the work actually finishes.