Global Browser History
Posted on November 17th, 2003
I have lots of bookmarks. Many hundreds. Far more than can reasonably fit into the Bookmarks/Favorites menu of my browser, even with the best efforts toward foldering and subfoldering. I searched high and low for alternatives, and until recently was using Active PHP Bookmarks.. But the author of that program has terminated the project.
Now I have no qualms at all about using antiquated computer hardware. Software is different. I took another look at the world of bookmark managers, but nothing really stood out. There are a few nice web based programs, but like the one I'm trying to replace their inherent limitation is that you have to do the work. From the moment you decide, "I should save this page for future reference" to the moment you have it saved, there's an interim of varying length where you select which folder or subfolder to put the page in, what description to give it, what title, and so forth. Too much effort.
Browser-based bookmarks have their own problems. I use multiple browsers in multiple locations— each would like nothing more than to clutter up with bookmarks until I had some here, some there, and none where I really wanted.
One solution would be to do away with bookmarks all together and just use Google. Possibly, but not ideal. The other day I was hot on the trail of a certain site I knew I had come across several months before, but damned if I could remember where. Google is good, but not good enough to say, "Find me that page I looked at a couple months back, the one with the really cool diagrams."
Google is impersonal. Its business is connecting you to the information you want, not remembering where you've been. Browsers handle that with their History feature. But that mostly sucks because it's specific to that one browser; it's short term; and it only tells you where you went, not necessarily what you found when you got there.
So here's the idea: a global history daemon that stores every page you visit in a database. It would attach to your browser as a proxy. Every page you visited would be remembered. This memory would stretch back months or years, not just a week or couple days. It would have a front-end for searching, and like Google you'd be searching the content of each individual page. It would be a mini-Google, only representing the slice of the Internet that you've personally travelled.
IBM researchers have already done something like this. There's something called the WBI Development Kit for Java. It looks like it hasn't gotten much attention since 1999. WBI is referred to as "a way to quickly and easily build intermediary applications." Just what I need! I've got a stream of incoming data that I want to view in its normal form (via web browser) but also log as soon as it comes in. All I need to do is introduce a slight detour in the road.