[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Ayatana] "fileless" paradigm





On Fri, Dec 17, 2010 at 7:23 AM, frederik.nnaji@xxxxxxxxx <frederik.nnaji@xxxxxxxxx> wrote:
On Fri, Dec 17, 2010 at 14:51, femorandeira <femorandeira@xxxxxxxxxx> wrote:

On Fri, 17 Dec 2010 13:37:12 +0100, "frederik.nnaji@xxxxxxxxx" <frederik.nnaji@xxxxxxxxx> wrote:
yeah, you're absolutely correct. The tools have improved a lot, but
none of the technologies has been installed into the default
distribution's UI in a generative place.
The first change we surely need to advance the semantic abilities of
our DEs is to add a tagging system to the "Save as.." dialog window.

Is it a problem with the tools, or is it something about how people think?

There are great differences between:
 - navigating step by step to a location and being required to "teleport" yourself there
 - recognizing your document when you see it, and being forced to recall some of its characteristics
 - always using similar steps to get to a document and having to come up with an appropriate search strategy each time
 - etc...

There is some research that shows that people like to have stable structures for their personal archives and that a strong physical metaphor ("one item is in one place and one place only") can decrease the cognitive load required to use such a system. Search is usually *only* used as a last resource tool when the user can not remember at all where he put that document...

True, but we don't want to say here that the user is incapable of managing a library, do we?
Often, in larger libraries, we don't know where something is, we can't remember.
Memory, as madbob already points out on his pages, is settled in the temporal lobe of the brain. It degrades with increasing temporal distance to the event.

Now imagine you have a library of photos covering a 3 year period. Are you sure you can locate all items by date, then by thumb?

Fred takes an average of 1 picture per day, he will have 365x3 pictures to go through in this case.
Fred remembers the photo was taken in year 1, winter, so he has about 150 pictures to look at via thumbs.
So much for spacial navigation combined with temporal sorting.

This is perfectly ok, it works for small libraries like Fred's.

Now imagine Jane has a new camera and goes on 4 vacations per year.
On each vacation, Jane takes several hundred shots.
After 3 years, Jane has 4300 photos she made all by herself, he Uncle's 200 wedding photos not even included.
As she tries to find "that picture with Auntie Mabel on that rock" from one of the vacations, she is confronted with 1200 photos to go through.
Now she complains: "darn, i wish i could just tell the computer to show me photos of Auntie Mabel".

Agreed, and compared to pro use cases, 4300 photos is still a tiny library.  For what it's worth, my wife is a professional photographer and will shoot 4000 photos at a *single* wedding.  Most of these photos will be deleted as she narrows them down to the very best, but even so, her current Aperture library is over 30,000 photos (representing about 2 years).

Now I'm not saying that a traditional file manager is suitable for professional volumes like this, but it would a shame if the "fileless" paradigm was even *less* suitable.  Ubuntu has a huge opportunity to become the platform of choice for pro photo and video.  IMHO, we need to make sure the "fileless" paradigm has the flexibility needed to accommodate pro users.

I need to get up to speed on the Ayatana "fileless" paradigm, but I'm hoping the Distributed Media Library can be used as a potential backend for media files - https://launchpad.net/dmedia

I think there are some good examples to follow in OSX in the way it allows a domain-specific database of files to be viewed through Finder in a fairly uniform way (you can find Aperture photos this way).  So the "fileless" paradigm doesn't itself need to support professional media libraries as long as it's possible to expose say a dmedia database with a plugin.
 
The reason why we need labeling is not because the other methods don't work.
Labels and tags are necessary, because our libraries are growing too big to be handled without topical, categorical or attributive management.
As long as we can not associate identifying attributes with storage item other than filenames, there is no chance at a simple approach to an intuitive management of larger libraries.



_______________________________________________
Mailing list: https://launchpad.net/~ayatana
Post to     : ayatana@xxxxxxxxxxxxxxxxxxx
Unsubscribe : https://launchpad.net/~ayatana
More help   : https://help.launchpad.net/ListHelp