Websites and file nomenclature

Silversprung's entry "Advice" brings up a good point. Websites and hypertexts that involve hundreds or even dozens of files can get cumbersome to administer if one does not have organized system of directories or filenames to help one keep track.

Keep the nomenclature so the web author can keep track of things. Readers/endusers will experience the site in terms of its linking, not in terms of filenames or directories, so these don't need to correspond. I suppose it would be convenient if they did, but I've never been so fortunate.

Perhaps this is counterintuitive, but I find that a fairly arbitrary system of classification affords the most flexibility in later editing. Here's why:

In establishing your nomenclature, you're essentially setting up a system of classification for your files. To keep things orderly, the general system of classification has to be consistent. In other words, if you're first classifying people by residence in your neighborhood, residence must always be the first classification, even if gender or age might be more important in the case of one or another character.

For instance, Silversprung mentions a set of residences in a neighborhood in what I assume will be a kind of narrative. These residences, I assume, will be peopled by families and perhaps other social units. If there are fewer than ten residences, I might start as follows:

00 = meta-information
01 = first residence
02 = second " . . . . . and so forth.

Then, within each residence there may be an oldest to a youngest party, a fattest to a thinnest, a reddest to a bluest, or whatever. I might designate that relationship as a-z. If so, the second reddest, fattest, or oldester person in the second house might be 02b.

How many lexia pertain to each person or each house? (Pertain could mean many things here, I guess, especially since lexia will inevitably refer to multiple characters. Perhaps lexia belong to their narrators, perhaps they pertain in some other way.) If so, we might have the third lexia delivered by the second fattest in the third house as 03b03.htm

The smallest category can have as many letters as you like, but in my experience it's convenient if every level of classification retains a fixed class of letters, so that I always know that the 3rd digit refers to characters' credit rating, or whatever.

Of course, this kind of system is grossly arbitrary. But I find that has some advantages. Stories evolve greatly as they're written, and you don't want your organizational matrix to either be inappropriate or to have to be revamped with each fresh edit. Also, you'll find that your theme will probably resist being subdivided into lexia. It almost has to, since the reader will reassemble something akin to your thought precisely by extrapolating between lexia.

-- A related point.

I map estories early and often. Several motives exist:

1) Outline form doesn't properly represent the relationships between lexia in hypertext, or not normally. Both sequence and hierarchy operate differently in associative relationships and in linked texts.

2) IMO, the structural difficulties of hypertext are WAY more difficult than those of a printed text that will be read in a single sequence or with the understanding of a single sequence. One's dependence on some kind of pre-writing expands accordingly.

3) Different linked presentations seem to have sufficiently different structures that I can't find one single mode of graphic representation that I am happy with for all of them. Therefore, after 10+ years, I still find myself feeling through each new project for a mode in which I can represent the structure.

However, there are a few practices that have worked frequently.

I generally start by feeling around at a map, representing temporal relationships spatially, either by movement from left to right or from top to bottom. Generally I represent some abstract distinction between lexia by the other dimension -- character, place, event, topic, or whatever. If I need more distinctions, I may make the drawings of the lexia different colors or shapes.

Storyspace used to be fine for this kind of thing -- far better than for the ultimate fictions themselves, I found. But probably no one has an ancient enough system to use it. MS Visio is about the one Microsoft product that I've enjoyed using, and it handles this kind of thing pretty well.

On the other hand, some crayons and a large sheet of paper or a few papers stuck together work just about as well as long as one doesn't mind wadding a couple dozen papers up first and if one won't lose the things.

I've known people who used 3 by 5 or 4 by 6 cards and a big bulletin board. I never did this, but I'd bet it works superbly. The cards can be moved back and forth, and one can pencil in mini-outlines of content for each card/lexeme. As I write this, I'm picturing little stretches of colored yarn stretching between lexia to represent links, but I'm not sure I would go through that much.

As a third point, you will probably want to have some idea how the characteristic screen of your presentation will look. This will presumably evolve right along with your concept of site structure, since the screen's visual aspects will typically be analogous to the abstract relationships of the lexia onsite. So I find that shifting between an evolving map of my site and an evolving picture of my screen usually helps me organize what I'll do.

Best of luck. I'll be interested in hearing how everyone does this, since this is something I've stumbled through with no established instruction, and I have no idea which of these ideas are or are not general in the industry.