Quote:
Originally Posted by tompe
With "flat" tags you cannot have things like sub-sub-sub-folders. So the standard tag implementation is not a superset to folders with respect to representational power.
|
yeah, I see the benefit of being able to create the proper solution space for your documents using folders. I suspect the something simpler can be creates with the cloud & tag concept. Simply allow the creation of sub-tags. Just considering it quickly, tags would seem to be easier in that all you are doing is building the look-up tables and indexes. We actually have that after a fashion with the shortcut links in Windows. It's a way to keep a file accessible in multiple folders as would be needed. The problem is, creating links and adding them to each folder is problematic w/o a mouse of some sort. A tag hierarchy could be similar but would allow a bit quicker linking when the content is added to the system, especially given the state of HID for data input on readers. Of course books properly tagged by publishers and authors to begin with would make everything a lot easier as we build out libraries. I see Calibre as a huge step in that direction.
You know a lot of books belong in multiple areas...but I might now see a bit of a circular argument here. Not sure, but I am basically considering what is easier for the hardware to manage quickly. I just have an intuition it is tags/sub-tags with a proper index and database system which might be faster. Even if the folders are built in assembly. I sense the tags would be easier for people to grasp and retain over time. I know people tend to lose track of folder structures they create.