Quote:
Originally Posted by Linwe
Ok, I think I posted about this once before a long while ago without any real resolution (primarily, I think, because it wasn't a repeatable issue) but here's something I'm getting quite frustrated with now. I'm open to any ideas and help.
...
|
For a problem I can't see for myself, I'm largely limited to speculation.
- FFF makes certain assumptions about story/chapter URLs and if they don't match exactly what's in the cache, it's a problem. A few years ago, a user had issues with a browser security feature changing the URLs. But that was on Firefox.
- The default browser_cache_age_limit is 4.0 hours. FFF ignores cache entries older than that. And, depending on the browser, it can sometimes be difficult to get the browser to re-cache the page.
- Some sites just mark their pages as no-cache. It seems unlikely ffnet would do that for only some pages, but I don't put anything past cloudflare--and they seem to be the dominate web server on the net these days.
- On one occasion, there was a corner case where our cache reading code didn't compute the hash correctly. We were eventually able to pin down a specific URL that could replicate the problem where I could see it and fix it.
The only way I know of to get any further is scanning every cache entry and looking for the matching URL.
I have code to do that, but only as a debugging tool. If you are prepared to try purpose-built debugging versions,
open an issue on github and we'll take it up there.
Personally, I use
use_browser_cache_only:true for ffnet--I assume direct access is never going to work and only use cache. Then I use
open_pages_in_browser:true in case I forgot to pre-cache the pages. But I use Firefox as my normal browser.