Searching vs Filtering
It struck me today that the way I use a search engines is often not to search, per se, but to find something I know I've visited in the past.
The difference between searching and filtering is not exactly mind blowing, however, I generally like to use the best tool for a given job and in this case I was not. Search engines provide a ton of value to be sure, but they are decidedly not great at finding things you've already visited. If you've ever tried to remember specific keywords just to get Google to surface some article you read a while back you know what I mean.
Yesterday I discovered HistoryHound, a macOS desktop app that reads all your browsing history from all your browsers and builds a search intex from it. The real killer feature is that it will download the full webpage and use the contents of the page to build the index, not just the title of the page. This means searching becmes much easier—titles aren't always what you expect or remember.
The ideal tool would do the following:
- Index browsing history based on full-text as well as titles
- Distill a minimal, content-mostly copy fo the web page and archive it
- In case the site goes down or changes. Automatically submitting your browser history to archive.org might also be an option, but offline would be faster.
- This is similar to a number of browser plugins or iOS's reader feature that remove most irrelevant elements of a page to focus on the content.
- Keep the archive forever
- Allow programmatic access
HistoryHound does some of this currently:
- ✅ Index browser history
- ❌ Archive pages
- ❌ Keep pages forever
- ❌ Programmatic access
So I'm still on the search for the ideal tool. Or perhaps HistoryHound actually does do some of this but I'm unaware of it. I sent an email to their support site with a few questions. Fingers crossed.
The big sticking point is programmatic access. For common tasks I use commands, not clicks and I want to incorporate history search into my workflow.