First off, I think it's safe to say I've blown my "write here pretty much every week" goal from the last post.
There is a file on your computer. You want to open it. What information about that file do you need to open it?
Depending upon your OS and the current year, you may need to know:
- The name of the file
- The directory the file is in
- The disk-drive the directory is on
Plan-9 is the only OS that I can think of where you don't have to know something about which disk-drive the directory is on. With most Unix-based systems, the "Which disk-drive" and "Which directory" questions get melded somewhat. But, if you consider a machine that has two CD-ROM drives and a disk mounted in each... then you'll have to be fairly crafty if you want to eject both disks, swap which drive they are in, and still have the same path names.
I was initially thinking that we've progressed somewhat in our interfaces. I was thinking that first we progressed beyond needing to know which disk-drive. I was thinking that we then progressed into not needing to know which directory it was in (because of visual browsers). But... I was wrong. We're still entirely locked into that mentality.
I was thinking that we no longer need to know which disk-drive. But, that was entirely wrong thinking. I was remembering drive specifications on AmigaDOS and DOS and Apple II's. But, we're still pretty stuck in those conventions.
Most OS-es try to make this easier by plopping lots of stuff on your Desktop. But, that's not much of an answer.
In addition to the above, we often need to know these days which computer the file is on. Much of this stuff is mitigated with 'Recent Files' menus and 'Restore Network Drives On Login?' buttons and search tools. But, it all seems very hackish.
Am I advocating a globally unique name-space again?