Comparing Firmware Development Environments

All posts written without LLM assistance unless otherwise noted.

Loading graph...

---
config:
  xyChart:
    width: 900
    xAxis:
      labelFontSize: 12
  themeVariables:
    xyChart:
      plotColorPalette: "#b342a0"
---
xychart-beta
    title "Zephyr Build System Benchmark"
    x-axis "Platform" ["Ubuntu 24.04 6.8", "WSL2 24.04 5.15", "WSL2 24.04 6.1", "macOS Sonoma*", "Windows 11", "VMWare (no AMD-V)" ]
    y-axis "Build duration in seconds (lower is better)" 0 --> 1680
    bar [664.44, 754.642, 723.323, 741.547, 1393.901, 1655.560]

About a year and a half ago, I decided to take a different approach to setting up a Zephyr environment for a new project at Intercreate. Instead of using my trusty VMWare Workstation Linux VM, I opted for WSL2. I was curious to find out: Would hardware pass-through for debugging work reliably? Would all of the tooling dependencies be supported? What about build system performance?

Not only did everything go smoothly, but since then, many colleagues have also moved from native Linux or traditional VMs to WSL2 and seen great results.

Continue reading at the Interrupt…



© 2024 by JP Hutchins. Published under a Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license.


Discussion

Response comments are pulled in from the fediverse, Bluesky, and GitHub. I try to moderate them, but I can’t guarantee the content of other people’s replies. Expand to read them and join in.

Show discussion Hide discussion

From the fediverse & Bluesky

No responses yet. Reply from Mastodon, the wider fediverse, or Bluesky and it will appear here automatically (after a bit).

To join in, reply from your own account to @crumpledpaper.tech@web.brid.gy on the fediverse, or crumpledpaper.tech on Bluesky.

Comment with GitHub

Made with Astro and lots of OSS.

🏳️‍⚧️