I’m trying to get puppeteer working on my Glitch project but running into some issues. When I try the standard installation command, I get an error saying my Node version isn’t compatible. My environment is running Node v16.14.2 with npm v7.20.6.
I thought about using an older puppeteer version like 19.6.1 that might work with my Node setup, but it seems that version is deprecated and won’t install properly.
Does anyone know a workaround to get puppeteer running without having to upgrade Node? Maybe there’s a specific version that works well with Node 16, or some configuration trick I’m missing?
Yeah, Node 16 works fine with Puppeteer. Just grab version 13.7.0 specifically - I’ve run this combo with Node 16.14.2 without issues. Use npm install [email protected] instead of letting npm grab the latest. On Glitch, you’ll need --no-sandbox and --disable-setuid-sandbox flags when launching since it’s containerized. If that doesn’t work, try puppeteer-core with a specific Chromium version. Gives you better control over the browser binary and usually fixes compatibility problems. Bundle’s a bit bigger but way more stable.
Had the same issue with Node 16 on Glitch. Set PUPPETEER_SKIP_CHROMIUM_DOWNLOAD=true and use puppeteer-core instead. This skips the auto Chromium download that causes version conflicts. You’ll need to set the executable path manually, but Glitch already has Chrome at /usr/bin/google-chrome-stable. Point puppeteer-core there and add those sandbox flags. Worked every time for me across different Node versions and dodged all the deprecated package drama. Performance was actually better too since you’re not downloading extra browser files.
Been there with Node version hell. Skip wrestling with deprecated puppeteer versions or hacking around Glitch’s limits - there’s a better way.
Hit the same wall running web scraping across different environments. The real problem isn’t just compatibility - you’re stuck maintaining everything locally.
I moved my puppeteer stuff to Latenode. They handle browser automation without the Node version headaches, dependency issues, or maintenance drama. Set up your automation visually and it runs in their cloud.
Built several scraping workflows there - zero version conflicts since. Way more reliable than Glitch’s resource limits too.
Check it: https://latenode.com