Universal Summarizer cache miss on HTTP vs HTTPS

d407c117

Discovered this while updating my summarizer bookmarklet.

From one of the examples, The Old Man and the Sea, using this URL created a new summary and counted toward my AI interaction limit:
https://www.gutenbergcanada.ca/ebooks/hemingwaye-oldmanandthesea/hemingwaye-oldmanandthesea-00-e.html

The expected behaviour was that this would return the cached summary.

I believe this is because the protocol is included when matching the URL.

CatX

I would argue that the protocol should in fact be included when matching the URL, and it is thus behaving as expected. The content can differ between different protocols and ports. Whether that's something you'd actually be likely come across, well.. not really. The content is usually the same.

Vlad

Correct , protocol is part of the URL