Medium: Live Data with Periodic Refresh

Fetch the current temperature from a free, no-API-key public source (open-meteo) and scroll it, refreshing every 5 minutes.

Full source: demos/medium/temperature.py

import asyncio
from scrollkit.app.base import ScrollKitApp
from scrollkit.display.content import ScrollingText
from scrollkit.network.http_client import HttpClient

URL = ("https://api.open-meteo.com/v1/forecast"
       "?latitude=52.52&longitude=13.41&current=temperature_2m")


class TemperatureApp(ScrollKitApp):
    def __init__(self):
        super().__init__(enable_web=False, update_interval=300)
        self.http = HttpClient()
        self.text = "Fetching temperature..."

    async def setup(self):
        self.content_queue.add(ScrollingText(self.text, y=12, color=0x00AAFF))
        await self.update_data()

    async def update_data(self):
        try:
            resp = await self.http.get(URL)
            temp = resp.json()["current"]["temperature_2m"]
            self.text = "Berlin: {} C".format(temp)
        except Exception as e:
            self.text = "Temp unavailable"
            print("fetch failed:", e)

        self.content_queue.clear()
        self.content_queue.add(ScrollingText(self.text, y=12, color=0xFFAA00))

Run it:

PYTHONPATH=src python demos/medium/temperature.py

What's new

• update_data() is the periodic data hook. The framework calls it every update_interval seconds (300 here) on its own async task — only if there's enough free RAM on the device.
• HttpClient abstracts the platform: on CircuitPython it uses adafruit_requests; on desktop it falls back to urllib. Same await http.get(url) call, same .json() response either way.
• We clear() the queue and add fresh content each refresh, so the display always shows the latest reading.

Blocking HTTP on CircuitPython

adafruit_requests is synchronous — the call blocks the event loop while it runs, so the scroll pauses briefly during a fetch. For a single small request that's fine. When you need lots of data, see the chunked-fetch technique in the hard tutorial.

Next: Hard — the full app.