On Sunday 07 September 2003 07:24 pm, Manuel Amador (Rudd-O) wrote:

> Putting them into folders has the already discussed disadvantages. 
> Filling metadata is much easier.  Just let the computer make up the
> "Folders".

Putting a file in a folder: drag and drop.

Filling in metadata: select category, it doesn't exist, create category, 
then select it, now create another category, since having only one 
category per file is no better than the traditional hierarchical 
filesystem. Now select a third just to be safe. Make sure that all 
three categories differ in their classification.

I somehow think the first is much easier. No matter how easy you make 
the user interface, you still need to attach at least two pieces of 
metadata to the file. No one's going to do it for the user. Certainly 
the computer won't, not if you expect the user to be able to find it 
again.

Pre-existing categories to select from isn't going to work. My real life 
experience says so. Is that music new age, ethnic or jazz? I have 
several albums where that's very hard to discern. What pre-existing 
categories are you going to give me for my shell scripts? Believe me 
when I say that at least half of my scripts won't fit your categories. 
So the user is going to stuck creating their own, and thus slowing them 
down, or they're going to use the all-purpose "general" category.

> with current data volumes, it's not only valuable, it's time!  How
> many times have I tried to collect all songs of a particular album?

It's more than songs. Every time I hear someone offer a practical use of 
Storage or similar database filesystems, it's always songs. I guess I'm 
just an old fart or something, because there's a lot more data on my 
system than just songs.

-- 
David Johnson
___________________
http://www.usermode.org

 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<