Tuesday, 18 May 2010

WebDiff

I haven't Googled this yet, but here's the thought while it's still in my head.

A server sends a user a HTML webpage. The user reloads it later, and only a tiny bit has changed - say, a text advert.

With WebDiff, it only sends the difference between the last page and the new one, so saving 100's of kilobytes in big documents.

The Diff is calculated using the Unix tool diff.

This only works because HTML is raw text.

Cute, huh?