[Bug] image cache problem

[umbrella-h]

The image in my profile README
is not updated immediately as expected.
(Checked that the file in Cloudinary is updated successfully)

Clues:

• Cloudinary cache
• GitHub readme image cache
• browser cache

[umbrella-h]

Browser

• https://stackoverflow.com/questions/49547/how-do-we-control-web-page-caching-across-all-browsers

Cloudinary

• https://support.cloudinary.com/hc/en-us/articles/202520852-I-ve-replaced-an-existing-asset-with-a-new-file-but-my-site-or-application-still-shows-the-old-asset-Why-is-that-
• https://support.cloudinary.com/hc/en-us/articles/360001208732-What-URL-conventions-are-invalidated
• https://cloudinary.com/documentation/image_upload_api_reference

[umbrella-h]

GitHub

• https://docs.github.com/en/authentication/keeping-your-account-and-data-secure/about-anonymized-urls#removing-an-image-from-camos-cache

• https://dev.to/jcubic/github-action-to-clear-cache-on-images-in-readme-5g1n

[umbrella-h]

Adding invalidate: true in Cloudinary upload parameters
Somehow influenced image scraping
(Not every job.)
so I removed it temporarily.

[umbrella-h]

Cache-control across user browsers is related to adding header parameters to the image response, which might be hosted by GitHub.

I tried to add metadata in README...
But it showed no effect on the image, so I removed the metadata settings.

Remain unsolved.

[umbrella-h]

Besides, I faced this problem while running this action.

Going to try to show my avatar via Gist
rather than Cloudinary & profile README.

[umbrella-h]

Adding invalidate: true in Cloudinary upload parameters Somehow influenced image scraping (Not every job.) so I removed it temporarily.

It still happens

[umbrella-h]

The cache problem is bypassed with Gist, but the image-scraping problem still happens.

For a hot fix,

1. use RMagick to check the percentage of white pixels in the image, and retry if needed.
2. Or simply set a delay before scraping