[dingaling]

I've been trying to design something along these lines now that Firefox aggressively prunes history with no user override ( it keeps barely five days worth on my laptop ).

However the main challenge is that much of the web is now SSL, so a light caching proxy isn't feasible unless it does MiTM which I don't want.

My current idea is to log DNS lookups but that doesn't, of course, provide page content data.

[maha1979]

The Fetch backend works pretty much like a crawler, what Google sees, it sees...private content is not indexed, e.g. your gmail mails, but the titles are indexed so there is still some usefulness in the search. We may add functionality for the user to sign in to these services so we can index them. E.g. the user may add his google account ao we can index mails etc. Would love to hear your thoughts on this.