I recently decided to start collecting archives of articles that I read on the web. In the past, I have depended on memory and Google but either my memory is not what it was or Google is not what it was or the web is just getting too big.
The first step was choosing the application or system that would hold these articles. I had a list of candidates in mind and quickly assembled a list of requirements for any system:
- Mac and iOS versions
- Over-the-air (OTA) sync between Mac and iOS
- Ability to edit or, at least, insert new records on iOS
- Solid search capability on either platform
Not a huge list of requirements, but it was surprisingly hard to find such a system.
I looked at Yojimbo (no OTA sync; read-only iOS app), DevonThink (no OTA sync; extremely sketchy reviews of iOS app) and a straight-up PDFs-in-Dropbox approach - which mostly works except that turning a web page into a PDF on iOS is either very awkward or impossible.
Lastly, Evernote. I have tried to love Evernote in the past but found it frustrating as a general solution. This is mainly because the iOS client is unable to edit anything but plain text notes and doesn't support the shared notebooks feature from the desktop version. That said, Evernote does have OTA sync so it seemed worth pursuing.
Inserting a web page into Evernote on iOS isn't a piece of cake. You can install the "Clip to Evernote" bookmark in mobile Safari but it's a little tricky and doesn't always catch exactly what you want. Secondly, the rendering of a web page in Evernote isn't always totally faithful to the original. I'm not sure why.
Everyone knows Instapaper, right? You save web pages for reading later. There's a website and an iOS client for Instapaper and support for the service is built into many, many Twitter clients and RSS readers. The Instapaper bookmarklet even correctly detects the article you're reading in Google Reader. Instapaper is great.
The second thing that's great about Instapaper is that it does a reformatting for readability. It generally does a great job of extracting the main body and you can add new parser configurations for your favourite sites if it doesn't.
So here's a service that I can insert a URL into from just about anywhere I'll encounter one on Mac OS X (Safari, usually) or iOS (Safari, Twitterrific, Reeder) and it makes it pretty and readable for me.
But what about having my own archive? One thing Instapaper doesn't do all that well is searching through your archived articles.
Well, it turns out that the Instapaper site has a couple of very useful "Sharing" options. Whenever you "like" an article in Instapaper, you can set up one of five actions for the server to take: post to Facebook, Twitter, Tumblr, Pinboard or Evernote.
Turning on auto-share to Evernote means that my workflow now goes like this:
- Find interesting article
- Save into Instapaper
- Read it at some point
- If it's good, hit Like in Instapaper
- Article automatically gets stored in Evernote
The nice part about this is that what gets stored in Evernote is not the original HTML but Instapaper's pretty-printed version of it.
At the moment, I'm just letting these articles pile up in Evernote's Inbox. Maybe eventually I'll sort them but probably not - Evernote has great search capabilites. What I want to build here is a place to go to find that reference to the day Sony stopped making floppy disks or that great line about the iPad 2 being considered a supercomputer in 1993.
My favourite thing about this is that the capture workflow is deliciously non-annoying and it separates capture, reading/reviewing and storage/retrieval into three distinct places: capture wherever I find the link; read and review in Instapaper and storage and retrieval in Evernote. All synced, all the time, everywhere. Wonderful.
My thanks to Ian Betteridge for his help in figuring this out.