Motivation

As noted in a previous entry, writing a bookmark manager had been on my mind for a while. I haven’t had time to really sit and write down what I was planning on doing. The idea itself wasn’t very clear to start with and the existence of online bookmark managers (del.icio.us, spurl, hyperlinkomatic, I’m probably missing others…) doing more or less what I thought about doing didn’t really make me feel the need to press forward with this endeavor.

Recently, however, posts on the subject have appeared in weblogs that I read. First, it was Buzz Andersen announcing that he just released a beta version of a Cocoa del.icio.us client, Cocoal.icio.us. Then, Jonathan “Wolf” Rentzsch posted about what a better bookmarking system could be like. These two posts provided the one-two punch that woke me up from my dogmatic slumber, er, complacency. Ahem, too much Kant reading in my previous life I’m afraid…

Anyway… This post is the first part of a two-part article meant to provide a place for me to put down some ideas on the topic of bookmarks from a knowledge management perspective. My purpose is to sort out my ideas and see wether it is worth it for me to spend some of my free time writing a bookmark manager. To this end, in this first part, I will look at the state of the art for bookmark management by providing an overview of what I think the key issues are and possible solutions to these issues. Part 2 will elaborate on desiderata for a knowledge management tool based on bookmarks and will extend the topic into the realms of social networks and what, if anything, they can contribute to an ideal solution.

This post will certainly be updated as time goes. This is but a first attempt at organizing my ideas so please bear with me as I try to work this out.

Bookmark management SOTA

Jonathan Rentzsch posits that bookmarking revolves around three operations, namely: creation, maintenance and use. This is an interesting point of view on the subject. That post also made me discover HistoryHound. My big beef with bookmark managers (and this seems quite a commonly held opinion) is that bookmarks are usually inefficient since it is so hard to find them again after you’ve accumulated a sufficiently large number of them. The crucial idea here is that I don’t use bookmarks (and I don’t think I’m alone in this case) as they were initially intended to be used i.e. a small number of frequently consulted pages. Rather, whenever I come across an interesting page, which I might or not have the time to read at that precise moment, I bookmark it so that I can find it again later. In a way, I use bookmarks as a knowledge management tool, a brain dump.

The big issue is obviously how to find the information again when I need it. I could categorize my bookmarks so that I can find the information I’m looking for more easily the next time I go hunting for something that I remember having encountered before and that I might have bookmarked. However, as noted by Rentzsch, browsers rely on hierarchies to organize bookmarks and those are too arbitrary to be really useful: a bookmark might fit in several different places of the hierarchy but can only (without having to jump through hoops) be filed in one spot. Also, I might have filed a given bookmark under a given concept in the hierarchy but this concept might not be the one I’m thinking of when I’m seeking to retrieve the information: I end up looking in one spot of the hierarchy when the information is really somewhere else. HistoryHound is of help here. No need for categories anymore: just put the bookmarks wherever and let HistoryHound index their content. Retrieval is just a matter of using a search engine on your bookmarks’ content. Full text indexing is definitely a must when it comes to knowledge management via bookmarks.

Another issue is the fact that bookmarks need to be available wherever and whenever you need them, regardless of the currently used browser. This obviously disqualifies “native” browser bookmarks and one needs to turn towards one of the aforementioned online bookmark social softwares: using one of theses services makes your bookmarks available from any browser as long as you can go online. Moreover, these bookmarks services also provides assistance when it comes to organizing the information: they all provide ways to assign tags/keywords to bookmarks to facilitate their later retrieval. Since these tags need not be placed in a hierarchy, part of the rigidity associated with traditional bookmark organization tools disappears.

In my quest for knowledge management, spurl has taken a slight lead: it seems to do all that del.icio.us does (you can even import your del.icio.us bookmarks and synchronize your spurl’ed bookmarks with you del.icio.us account) BUT also provides full-text search on your bookmarks. Spurl, thus, more or less manages to combine HistoryHound’s usefulness with online services’ availability and flexibility. However, spurl doesn’t let you re-index already stored bookmarks. This means that the bookmarks that you imported from del.icio.us or from your browsers are not indexed. This also means than once the index is created, the information is sealed: it won’t evolve with the page at the other hand of the link. Moreover, contrary to del.icio.us, spurl is not RESTful. It is therefore more difficult to integrate it with other knowledge sources such as a personal weblog, though this issue is alleviated by the fact that you can keep your spurl’ed bookmarks in sync with a del.icio.us account. Spurl would therefore take care of the bookmark storage and retrieval while the sync’ed del.icio.us account would provide an interface to other services.

My Own Summer (Shove It) from the album Around The Fur by Deftones

Related posts:

  1. First Adium / Twitter integration thoughts in 1.4 beta 1 It’s quite funny to see that, as I finally got...

Related posts brought to you by Yet Another Related Posts Plugin.