Many of us use the Web as our main source of inbound information, but not our only one. The one thing we learn very quickly is that the Web is fragile and forgetful. Bookmarking things is not the most optimal way to help us get back to the information we value or might value in the future. As we know, having a link to a Web page doesn't come with insurance that the page will be there, or be there with the same information it had previously.
That forgetful Web created a market for tools that can capture and store that information for our later use. In past columns, we have looked at services on the Web that can do that. But because they are on the Web, are they always going to be there, or will we always have a connection to the Internet to get to the items? That thinking helped spawn software that lives on our desktop or laptop. "Junk drawers"-as some call them-are where we shove all sorts of things, from full Web pages, to video or audio (metadata is needed to help find/refind it), notes, documents, etc. All of those things are stored locally on our hard drive(s).
Having the information stored locally is comforting (well, if you are at the machine that has the repository). Locally stored information needs to have good retrieval. The offerings in tools include native text search, robust metadata parsing, folders and/or tagging. In many of the tools, the ability to get to exactly what you are looking for is quite good, which makes them great tools to have at one's beck and call. What is amazing about having those tools running for some years is the ability to easily turn inklings of "I think I remember seeing that" into "I have that here," pulling up an article from seven years ago in seconds.
The key to having things from years past within easy reach is having a tool you are feeding over those years. Most of those tools will provide really quick results using the features to aggregate research for reports, articles or quotes for customers. But feeding the services requires a bookmarklet that you can put in your browser toolbar or a desktop droplet (what an icon on the Mac is called) where you drag a URL and it captures the page. Many of the tools will also pull information in using RSS feeds (what would we ever do without them).
I stumbled onto those services in 2004 when a friend mentioned DEVONthink (by DEVONtechnologies). Then author Steven Berlin Johnson praised the tool in a blog post in 2005. (In his recent book, Where Good Ideas Come From, Johnson focused most of a chapter on it.) DEVONthink is a Mac-only tool, and I do know two or three people who have a Mac just to have DEVONthink running. But, there are other similar services that have surfaced since DEVONthink started in 2002.
On Mac, those include Yojimbo by Bare Bones Software, Together by Reinvented Software and Evernote by Evernote. On Windows, the ease of capturing web pages doesn't seem as popular, but it exists to various degrees in Microsoft OneNote, CintaNotes, and Evernote among other more pure note managing tools.
At various times, I have used all those Mac tools, as well as OneNote. I started with DEVONthink in 2005 or so, and realized its value quickly. But in 2007, I started writing a book and found that the (then) DEVONthink lacked a tagging capability that I wanted. So I moved to Yojimbo because it had that tagging functionality. Early versions of Yojimbo didn't scale, though, so I switched to Together, which served me well until DEVONthink Version 2 arrived with tagging. Moving between the applications was as simple as taking my existing repository and importing it with software provided tools.
Why did I end up in DEVONthink? Well, not only can I find, with relative ease, exactly what I am looking for in the nearly 14,000 objects I store in it, but DEVONthink also employs artificial intelligence (AI) to show similar resources in my repository and the calculated strength of that relationship. That one feature, on top of easy search and tagging at scale, turns DEVONthink into an absolutely genius tool at my fingertips.