SubscribeFiles aren't children. The requirement for unique names is to let the computer uniquely identify them.Actually, it's not; it's to let the computer unambigously and conveniently refer to them, which is exactly the same as the function of children's names with respect to children.
As for why it's a bad idea, teach someone who has never used a computer before why they can't name all their pictures of their son Billy "Billy." I mean, the pictures are of one person, Billy, not separate people named Billy01, Billy02, Billy03, etc. You would be surprised as to how counterintuitive this is.Anybody who can understand why all the Harry Potter books aren't all simply called "Harry Potter" can understand this.
When you have photos in an album, they don't even have names. Imagine if, before you could put photos into an album, you had to name each of them uniquely!Typically, nobody actually has to name their photos anything at all. Typically, the camera gives them some meaningless but unique name perhaps based on a date, and when they're uploaded they get put into a folder with another possibly meaningless name. In this instance the filesystem is doing the minimum work required to keep all the photos distinct and accessible.
Okay, tell me how to do that using Word.With Word, you don't even need separate patch files; you just turn on its inbuilt version control stuff and all the patches get saved right inside the same file as the base document.
Sure, if you want to do work a computer should be doing all day. The idea that I'm going to spend hours creating folders full of shortcuts, and then spend additional hours maintaining them? Not. Computer's job.If the computer in fact has as much understanding of the material it's organizing as you do, it can auto-index it and save you a lot of time; that much is true. But I don't see why a reasonable objection to manual indexing should make auto-indexing on top of a database any better or worse than auto-indexing on top of a filesystem, if the filesystem is capable of providing the kind of indexing facilities actually required.
A database is a much better idea.
posted by teece at 8:33 PM on February 16, 2006