By Joel Dare - Written July 3, 2026
Yesterday I made a change to my blog. A simple change to my subscription page. I added a bit of “social proof”. A users very nice comment to me. I did what I always do to push it live:
git push
I write my blog with plain markdown. There is no build step, not on my side anyway, it’s just text. Then I rely on GitHub to take that text and build it into html for me. This used to happen automatically and quickly.
But, yesterday, it never went live.
I logged into GitHub and saw that the deploy step had failed. I don’t have custom actions here. This is GitHubs default behavior. Just another temporary GitHub failure, I figured. I re-ran it manually and it worked the second time.

Then, today, I made another minor change to a different site. It never deployed.
So, again, I log into GitHub and take a look. Same problem. The action failed at the deploy step.
This is not some custom action I wrote, it’s GitHub’s own internal GitHub Pages build process.
This is getting really frustrating.
It’s probably time that I start to look at other solutions. A local build step and a push to a VPS or shared hosting account would probably happen in a few seconds. The build burden would move to my own machine. It could be easily automated with a simple deploy.sh bash script.
I’d also consider a deploy to AWS, GCP, or Cloudflare but these services feel like overkill and unknown costs for my tiny blog.
Now I’m unable to deploy this article. It looks like they’re having an ongoing outage. I’ll have to try again later.
P.S. Happy birthday Mom.
JoelDare.com © Dare Companies Dotcom LLC
through formrobin.com