Re: [Usability] User problems and practices with modern desktopsystems

[<john erling blad aftenposten no>]

> What I would like to have or like to design is a new filesystem, maybe
> based on a relational database, but not on top of an existent
> filesystem. The system may still run on a filesystem, but not the
> userdata, that what we now call "desktop" and the objects inside.

I would prefer if categorization and other indices about file content
was stored as file metadata, sill the metadata could be indexed by some
external means and stored in some relational database.

> >> This problem can be and should be tackled from both directions,
> > although
> >> automatically aggregating and searching information is useful it is
> > also a
> >> good idea to encourage users to organise their data little better.
> >
> > I don't think it is a very good idea to enforce to much manual
> > classification on the user. He simply won't use it. What I do
believe
> > is
> 
> But you do it already, while giving the file a name and looking for a
> right place in your folder structure.
> people who get into deep trouble finding their files are used to give
> long explaining filenames.

I've been observing that some statistical classification methods operate
a lot better if given even a marginal apriori knowledge. You could say
you use an existing hierarchy of directories as a hint for further
classification/cathegorization.

Let's say you have a "music" folder, then you know that all sound files
you find under this directory will also be music files. Now extend this
a bit and you find that you could have a "cv" directory that contains
all your various cv's. Now, turn it all around, if you look up a
specific music clip you know that another similar music clip is not
found underneath your cv-directory.

> > Directories like Documents, Pictures and so fort don't really add
much
> > information as the information already exist as suffixes, ownership
and
> > so forth.

Well, it add some information. :D

> > What Pictures will do is to collect those files in a common
directory.
> > But think about it, what if Pictures wasn't a real directory but
some
> > kind of meta folder. It could be a collection of all pictures on the
> > system. Of course it would be a huge folder. Now, what about some
kind
> > of balancing algorithm which added the directory instead if there
was a
> > directory with a lot of images? This is something that is more or
less
> > like locate.
> 
> What is the problem with a huge directory? there is only a search
> problem.
> If we add a button to nautilus hide everthing exept picture or docs or
> music files, we will have a similar result as we have with different
> folders. This will be one step to that what expect to have.

Search isn't very user friendly unless you are very familiar with it. My
idea of a balancing algorithm is something you could manipulate. Let's
say you have a music folder. When this grows large enough information is
extracted from metadata to build an artist level. When you traverse into
a folder with enough files this will again be divided into albums.

You could imagine that you was able to set one of several schemes for
how this would be done. In fact given a specific type of folder a scheme
could be predefined. A folder like "My Pictures" cold be balanced by
information from exif data.

> > But really, isn't this a special case of one single file type?
Perhaps
> > it could be generalized somehow?
> >
> > Maybe instead of Documents there could be something like "My files",
> > all
> > of which are owned by the current user? This would limit the number
of
> > files a lot compared to a complete traversal of the complete file
> > system!

This comes from an observation that "My Documents", "My Pictures", "My
Music" mostly are just some sorting on file types.

> > Perhaps a folder like "At same time" (in lack of a good word) which
> > contains files opened at the same time as whatever you have open
right
> > now. Most likely you would have a database with previous results
from
> > scan by lsof to make this work.
> > 
> I'm not sure if you mean the same thing, I avoid to use the name
> "folder", cause folder are strict hierarchically and  disjoint and may
> irritate.

Imagine hierarchies used for hiding unnecessary information, findingo ne
music clip among 5-6-7 000 files isn't easy but if you break it down to
5-6-700 albums or even 5-6-70 artist it will become manageable.

> I didn't say that it is mystic, but its a fundamental change. I would
> like to start doing it, but don't have the resource and knowledge, but
> maybe I find some time to do a simulation.

Similar things are done on several portals for news bureaus on the net,
so it isn't that mystic! :D

> Lutz

One idea I've played around with is to make a mountable directory that
takes some parameters that specify how to make an extract of files. In
its most simple implementation you could add a parameter locate=.mp3 to
get a directory of all your mp3 files as found by locate.

John