Jean-Gabriel Gill-Couture
4a9bee73ad
- docs/guides/application-capabilities.md: the .with(...) menu, by-reference wiring, and a recipe for writing a capability; disambiguates the overloaded 'capability' (topology vs application) flagged in DX review; wired into the book. - Deprecate ApplicationScore, features::Monitoring, features::PackagingDeployment (+ doc-deprecate the ApplicationFeature trait) in favour of harmony_app + ComposeDeploy/.with(...) (ADR-026). Legacy examples now warn; harmony lib is warning-clean (internal self-refs allowed).
77 lines
4.2 KiB
Markdown
77 lines
4.2 KiB
Markdown
<!--
|
|
STATUS: PRIVATE DRAFT — not for publication.
|
|
Aphoristic "manifesto cut" of why-harmony.draft.md. Companion to the essay
|
|
"A Wonderful Future for Computers". Author controls if/when published.
|
|
-->
|
|
|
|
# Why Harmony
|
|
|
|
The first essay was about *where* computing should live. This is about who gets
|
|
to run it.
|
|
|
|
Decentralized hardware means nothing if only the giants can operate it. A
|
|
micro-datacenter in a village is scrap if it takes ten engineers and a year of
|
|
glue to run one service reliably. Infrastructure is the root of nearly every
|
|
software business — and the root is owned.
|
|
|
|
I've spent a career making distributed systems survive failed machines, bad
|
|
deploys, 3 a.m. outages. That robustness isn't magic — it's expensive teams and
|
|
hard-won patterns. It is real, and it is *rationed*. Only the well-resourced
|
|
get it.
|
|
|
|
That's the quiet injustice. Not villainy — an incline. Lock-in, pricing that
|
|
rewards scale, defaults that assume you're big: nobody had to conspire. The
|
|
ground is tilted, and everything rolls toward the same few owners.
|
|
|
|
So the conviction is stubborn and simple: put what I know into code instead of
|
|
another invoice, and the robustness that costs a platform team becomes
|
|
something every app just *has* — for free, by default, like running water.
|
|
|
|
**What we believe.** Computing should serve people, not skew toward owners.
|
|
We're not against the rich or the large — we'll serve a Fortune 100 and a rural
|
|
clinic with the same code. We just refuse to bake the skew in. We remove the
|
|
tilt rather than reverse it; a level field is already radical when the ground
|
|
is sloped. Robustness is a right, not a privilege — deploy something and have
|
|
it keep working shouldn't require an SRE budget. And freedom means you can
|
|
leave: if it only runs on one vendor's cloud, you don't own it, you rent your
|
|
own existence.
|
|
|
|
**What it is.** You describe what should run, as real code, and it converges —
|
|
the same way — wherever you point it. Ask the target for something it can't do,
|
|
and it won't compile: the mistake is caught at your desk, not at 3 a.m. A
|
|
deploy isn't done when a command exits; it's done when the thing actually
|
|
works, proven, the way a careful human checks before walking away.
|
|
|
|
**How we differ — without mysticism.** We didn't invent robustness; we stand on
|
|
a generation that did. What we found is a seam no one occupies. Tools that turn
|
|
metal into a cluster stop at the cluster. Tools that ship apps assume a cluster
|
|
exists. Tools that compose infrastructure live in the cloud and assume someone
|
|
handled the metal. The project closest to "infrastructure and app in one real
|
|
language" served only the big clouds — and its company didn't survive. No one
|
|
spans the whole of it — bare metal, to running service, to keeping it alive —
|
|
in one coherent, type-checked model that's honest about the physical machine.
|
|
Aimed first at the sovereign and the self-hosted. That seam is where we live.
|
|
|
|
**Where we honestly stand.** Early. The incumbents are years ahead on breadth,
|
|
and that's a real moat. We're small. Anyone claiming a young tool has already
|
|
won is selling something. So we won't win by breadth — we'll be undeniably good
|
|
at one thing: running real services on self-hosted and sovereign clusters, with
|
|
the robustness usually reserved for the giants. Our proof isn't a benchmark or
|
|
a funding round. It's a sovereign health-data hub headed to a community in
|
|
Cameroon — a place where depending on a distant cloud is a question of whose
|
|
hands the data is in. If it stands up there, the promise is real. If it can't,
|
|
no manifesto will save it.
|
|
|
|
**The ambition, kept humble.** This is less a company than the removal of a tax
|
|
— the quiet tax infrastructure levies on everyone with an idea and no platform
|
|
team. Software is one of our great levers for helping each other; the cost and
|
|
lock-in of the ground beneath it is a weight on that lever. Take the weight off
|
|
and you can't predict what people will build. That's the point. You don't pave
|
|
a road because you know who'll walk it — you pave it so they can.
|
|
|
|
Make the infrastructure sing — not for a few, not against them, but for anyone
|
|
who shows up. A paved road to freedom for all software.
|
|
|
|
If we get it right, most people will never think about it. The water will just
|
|
run.
|