Hacker News Reader

No worries, glad that worked! I have support for offline mode, and even though it's set up to try and get resources from the network first, it seems to often pull from the local cache even when it shouldn't. I'll have to try digging into that issue some more

View on HN

The biggest difference maker for me is using hashed filenames for non-html resources. That way even if the html caches too long it's at least a coherent experience, and if anything else caches too long who cares, the user doesn't see it anyway.

I don't know what your pipeline looks like or if you're already doing that, but if not it's worth a look.

Reply on news.ycombinator.com