Getting started
Download the JSON bundles and use them in JavaScript or Python.
The whole map is published as plain JSON files. No account, no key, no rate limit beyond normal web courtesy. Pick the bundle that fits your job.
The bundles
| Bundle | URL | What is in it |
|---|---|---|
| Everything | https://platformontology.lol/data/ontology.json | All 844 capabilities with every field, plus the taxonomy |
| Light index | https://platformontology.lol/data/index.json | One short row per capability: id, name, kind, group, headline, reach, status per platform |
| Taxonomy | https://platformontology.lol/data/taxonomy.json | Labels, icons and descriptions for kinds, groups, lifecycle stages and browsers |
| Android | https://platformontology.lol/data/platform/android.json | The 361 native capabilities, with the Android block under platform |
| iPhone | https://platformontology.lol/data/platform/ios.json | The 361 native capabilities, with the iPhone block under platform |
| Web | https://platformontology.lol/data/platform/web.json | The 483 web capabilities, with browser support under platform |
Every bundle starts with the same small header: name, version, built, license, homepage and count.
Start small
Load index.json first. It is much smaller than ontology.json, and it has enough to build a list or a search box. Fetch the full bundle only when you need details.
JavaScript
const res = await fetch('https://platformontology.lol/data/index.json');
const { count, capabilities } = await res.json();
console.log(count); // 844
// Tracking capabilities that reach broadly
const broad = capabilities.filter((c) => c.kind === 'tracking' && c.reach >= 3);
console.log(broad.map((c) => c.name));Python
import json
from urllib.request import urlopen
with urlopen("https://platformontology.lol/data/platform/android.json") as r:
data = json.load(r)
print(data["count"]) # 361
# Android capabilities that need a runtime permission prompt
for c in data["capabilities"]:
if c["platform"]["grant"] == "prompt":
print(c["name"], "->", c["platform"]["permission"])The shape of one record
In the full bundle each capability is one object. In a platform bundle the platforms field is replaced by a single platform block for that platform.
{
"id": "mob.app.associated-domains",
"name": "Linking an app to a website",
"kind": "capability",
"group": "app-ecosystem",
"space": "mobile",
"plain": "Proves the app and website are owned together...",
"platforms": {
"android": { "s": "y", "v": "6.0", "api": 23, "permission": "..." },
"ios": { "s": "y", "v": "9", "permission": "..." }
},
"limits": { "headline": "...", "can": ["..."], "cannot": ["..."], "reach": 1 }
}Read the data model for what every field means.
Build the bundles yourself
Clone the repository and run the build. It validates every file first, so it also works as a check on your own edits.
git clone https://github.com/ever-just/platform-ontology.git
cd platform-ontology
STRICT=1 node scripts/build-data.mjsThe bundles land in dist/. There are no dependencies to install.
Link back to a capability
Each capability has a stable id. Use it in your own data so you can always find the page again. For example, the page for mob.app.associated-domains lives under capabilities in the app group, and the simulator opens it at /lab/#mob-app-associated-domains (dots become hyphens).