notes on database fiction

i’ve been reading traversals by dene grigar and stuart moulthrop, which is about experiencing and trying to preserve early interactive fiction. i was really struck by the chapter on uncle roger, an early work of database fiction by judy malloy.

malloy first published uncle roger on the whole earth ’lectronic link (the well) as a sequence of lexias, short text extracts with some metadata. the idea was, you’d copy these lexias and their metadata into your own database software as they were released on the well, and from there you’d be able to navigate through this story by searching through the metadata and retrieving different entries.

i really like how minimal this system was. publishing these entries agnostic of any specific database software - just an extract of text and some keywords - which readers could enter into their own systems as needed. most systems now need a bit more infrastructure than this. i’m writing this entry in markdown and exporting it into html using a pandoc batch command. even this is more apparatus than the original version of uncle roger. but i want hyperlinks and css styling, and there’s no way i can have those just using txt files. even twine, in comparison, is a whole lot more apparatus.

why focus so much on the formatting? traversals points out that some much electronic literature has become inaccessible, even less than a decade after it was first created - things like uncle buddy’s phanton funhouse or patchwork girl, which relied on proprietary software.

but on the other hand, i replayed reagan library a couple days ago, which is basically a collection of html pages with some javascript to store variables. (the beautiful navigable panoramas unfortunately did not last so long seeing as they relied on quicktimevr. one workaround is to download the panorama files as you play and open them in another program.)

and last night i fiddled around with the web version of uncle roger, which relies simply on hardcoded links between different html pages for each lexia and so remains completely accessible.

my takeaway from this is, the more simple the digital medium being worked with, the more likely they are to outpace obselescence. text files will probably be around as long as there are computers.