History
In the early stages of the Evernote Service development the Web Clipper was just an easy way for developers to get content into the development notebooks quickly. It also helped us test our HTML sanitizer which makes sure no nefarious javascript makes it into users notebooks. First there was a command-line version, then a little clipper that just sent the URL’s to Evernote but didn’t work on secure sites. As we approached the final weeks before the beta launch we decided that the Javascript Web clipper was going to be part of the core service offering. We debated a little, each of the native clients support clipping and screen shots so it wasn’t clear that we really needed another way to clip content, but there was a strong lobby for not wanting to leave the web when you’re capturing web content.
Ultimately we decided that we needed a really good browser based clipper because a key part of the service launch was letting users capture from anywhere.
The version we released did a lot of what we wanted: one click capture, easy install, instant feedback, image and HTML document capture, capture selections of text and images or the whole page. We also support clipping from password protected sites like web-based email clients (which turns out to be kind of tricky).
Launch
Probably 15 minutes after we launched we realized the Web clipper was going to get a lot more use than we planned. In many ways it was just what we’d hoped for, users merged their personal photos and notes with information they gathered from around the web; Evernote helped them bring it all together. The Javascript clipper isn’t entirely bug free, each browser is different and mileage still varies. Also, there are a lot of really bad HTML pages out there and cleaning them up into our own internal variant on XHTML can be tough. For the first month or so I’d get a list of errors logged when the clipper could not save the clipped content from remote web pages. This list of different sites has provided a great source for testing and bug fixing.
Clipper Revisited
With this release the Web clipper has finally become really useful (at least for me and I hope for you).
Tagging and filing the clips as you clip them makes it so much easier to find pages later. I’ll clip sample code from articles, blog posts from Google Reader, recipes and put them straight into the correct notebook.
Some older features also seem more useful. Whenever you clip content, Evernote sets an attribute so you can view all your web clips and then filter by tags, notebooks or other criteria. If you are feeling a bit more adventurous, you may want to try filtering based on source URL. In the search box type: sourceURL:http://www.nytimes and you’ll be able to see all the notes you’ve clipped from the New York Times.
If you put clipped pages into public notebooks or email a note from a clipped page, your friends will also be able to access the source URL.
More to do
I’ve still got more features to add to the web clipper. At the top of the list is “search as you type” tags. Clipping from sites with frames is still a little flaky espacially on Safari. Lastly, I’ve been toying with the idea of recommending tags based on the clipped document’s similarity with other documents you have.