IdeaLogging. It's totally about knowing something exists. A rationale or development direction for WalaCode.

The bookmark model is flawed for at least two key reasons: Linkrot and local storage.

The thing is, lists of URLs get broken rapidly. The address scheme that drives the web requires first that the named address portion of a URL resolve to the right physical machine, and secondly that the URL still points to an existing file or chunk of database. Few such addresses can be counted on for more than a few months or years. Even if the right server is still taking requests, files are frequently renamed, deleted, or changed beyond recognition - and database driven sites are often models of ephemerity.

On the other hand, if you know something is out there under a certain heading, somewhere on the Internet, there's a good chance you can find it. Google and its lesser cousins may be starting to choke on search engine spam, but they're still remarkably functional - and they're adding capabilities like maps and fulltext book search that massively expand their search space. Also, projects like Wikipedia and Everything2 are almost frighteningly resource-rich these days. So why not keep track of search terms, titles, authors, subjects, etc. instead of addresses to specific resources?  After all, other people with a lot more time, money, and phDs are working on the search/address problem.

As to local storage, well, what's the point of keeping what you know to yourself? And anyway, what are the chances that the only time you want to access this stuff will be when you've got your personal machine handy?
<[[BrentNewhall]]> There's another side to local storage, though:  distrust of permanent connection.  Too often, I've wanted to watch a music video, but have not had an internet connection.  Even wired broadband has this problem; ISPs go down, home routers fail, etc.

I don't want to have to rely on the internet for certain key pieces of information.

<[[Brennen]]> True that. I think my setup will eventually include a local mirror of stuff like this wala, and some scripts to merge in local changes - I'd rather be editing the whole thing in VimEditor anyway - but that's going to be dependent on having even an occasional 'net connection on my personal machine, which at the moment I don't.

I used to edit everything on p1k3 locally and run a utility (I forget which one, but I think it worked much like rsync) to mirror the changes on the webserver. Then I was in the alternate universe of a college campus with always-on broadband for four years, and got lazy.