Holly has discovered the personal data warehousing problem:
| As an MA candidate merely working on one paper for one course, the digital packrat’s nest that has become this project’s home on my hard drive makes me wonder how generations of PhD candidates were able to slog through the stacks of notecards that must have littered their tiny grad-student hovels from wall to wall, “room” to “room,” and floor to ceiling, not only to find the ones they wanted when they wanted them when it came time to TYPE, but just to find their way to the all-important coffee pot. |
I don’t have a solution for the piles of PDFs all over my hard drive, but I do think the cards work well (and I’m a CS grad student!). I keep my bibliography in BibTeX (a tool, well, similar to EndNote), but I keep my notes for authoring papers on 3×5 index cards. Why?
- Cards are portable in a way that computers are not. Even my Powerbook.
- Cards can be rearranged in ways that notes in the EndNote DB cannot.
- The act of making the cards helps me remember the content in ways typing a few words into the EndNote DB does not.
- The act of making cards helps me think about how to summarize the thesis of a paper, section, or paragraph, which is useful for relating other people’s work to my own.
- Cards can be shared (with ones supervisor, say) easily in a meeting, shuffling through them when discussing ideas.
BibTeX is a lifesaver when it comes to including citations in a paper, but I don’t use it for managing the writing process; perhaps EndNote is particularly useful in that respect. I suspect it depends on how one likes to write. I haven’t quite figured out how I like to write yet, so I started with a simple system that’s been tried and tested.
On another note, I wonder if you’re familiar with the paper Prototyping for Tiny Fingers? I assume you have access to the ACM Digital Library at Columbia. This came to mind because you reached straight for a piece of software in which to build your simulation—is that a requirement? Would it be easier to get to the meat of the exercise—the testing—if you weren’t constrained to developing your simulation on-line? And wouldn’t it be easier to adjust the simulation over time as you discover flaws with it? Food for thought, anyway.
And lastly, is there a possibility that a constructivist simulation environment for learning how to filter email would be one in which the application itself supported exploration on the part of the user—perhaps by having context-sensitive help and appropriately named and placed “wizards”? Or, perhaps it provided you with an “email filtering game” that you could use to test your filter-writing abilities (a way to practice)? Since email filters are, effectively, regular expressions against semi-structured text, learning how to use them (without learning what “regular expressions against semi-structured text” means) might be done best when you’re playing a game and not worried about mucking up your own email. Consider FrontPage: you can fiddle all day with it, but it isn’t “live” until you want it to be live. Seems like a similar notion might apply here.