nepsilon

If you can predict when you're going to get new content, just use an HTTP header to tell until when the content can be seen as fresh. Then the client will know when to call the API again.

If not, ask the client to provide an HTTP callback endpoint, and ping it to notify of new fresh data on your side. Using WebSocket is also an option.

So this has nothing to do with Django, other than being built with it? It took me a while to realize that.

These are 2 valid concerns. Using localStorage to cache the URL already looked up would help for the pages you visit the most (gmail.com, twitter, etc). Having a whitelist/blacklist of site {not,} to lookup might also help. Let me know if you think these 2 features would actually help. I could implement them.

Thanks for your comment!

I've about 7 years experience coding various software in Python (data intensive, web apps, network app, etc). Today when coding in Python I mostly go to the official Python and Django doc directly, that I have installed locally. I open it about 2 to 8 times a day (4-6 hours span) depending on how well I'm familiar with the library I'm dealing with.

For more general questions not covered in the doc I Google about twice to 4 times a day in average. I usually go to stackoverflow.com directly to search.

When coding HTML/CSS/Javascript I clearly lookup caniuse.com about 20 times a day to ensure consistent browser support and learn about corner cases.