I’ve had a post in my drafts folder for a few months now.
The title of that post?
“How to Code Your Website to Be as Instapaper-Friendly as Possible”
For a while there, I thought I had figured out how to do so, based on a lot of time spent on a lot of different websites.
I would save an article from TheOverAnalyzed to Instapaper, and then check the iOS and webapps to see if things like the author name, date, or other random context was parsed correctly.
If something was amiss, I would visit a website where I knew Instapaper was correctly parsing the thing that it wasn’t correctly parsing on my own site. After comparing the source, I could try emulating/incorporating what the other site was doing to see if that would help my site parse correctly as well.
I was this close to publishing that post, but then I noticed a nasty footnote parsing bug.
That led to a few long emails exchanged with Brian Donohue, Instapaper’s main developer (after Marco sold it).
Ever since, I’ve been testing my site against an experimental version of the parser (super top secret obviously), and I think I’m almost ready to publish that post. There are a few extra things we need to work out, but all in good time.
It looks like Brian and the rest of the Instapaper team beat me to the punch, because their new parser has been incorporated into the latest public release of Instapaper.
Instapaper is the best read-it-later/bookmarking app, period.