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

[<john erling blad aftenposten no>]

> Maybe sometime I write a bit confusing, sorry. My interests are:
> is there a possibility to get a more convenient "desktop" than today.
> 
> my mental model is invented by my own impossibility to handle my data
> well. These are adresses, bookmarks, email, written texts, downloaded
> texts and binaries like pictures, mp3 etc. These data are spread at my
> apple and my linux pc at work and my linux pc at home. And I often get
> into to trouble finding files, finding an appropriate place for saving
> data, getting rid of identical data or finding out on what file I made
> some special changes.
> Beside that I'm observing some people handling their data.

My point is, it seems likely there are a fairly small change necessary
to be able to handle various kinds of metadata from other sources, so
don't limit the solution to a single "desktop solution".

> My conclusion is that the current model of files and folders,
specially
> for user data, could or should be replaced.
> My idea is to save the same information that is implicit stored in the
> name of the folder may stored as metadata and used as a select
> criteria. Thats what I would like to show here:
> http://wkm.kunst.uni-wuppertal.de/~mueller/file.gif
> http://wkm.kunst.uni-wuppertal.de/~mueller/browser.gif
> 
> The second thing is to have more information about the data, as it is
> partially possible for images or mp3. For example: why don't save the
> text typed in a gimp image also in his metadata, Or the name after
> "dear Mr/Mrs ....". There are thousands of possibilities.

Don't focus on solutions for specific problem. What is the overall
problem to solve? I believe there are
  - Extraction and transformation of data to be used as metadata
  - Storage of metadata somehow in relation to the file itself
  - Creation of indexes so data retrieval is fast
  - How to search for the files by using the metadata

I'm also tempted to believe that the current folder metaphor isn't that
bad. It's the retrieval situation when you are searching for something
that sucks, and that can't easily be solved just by adding a search
engine.

Why? Simply by imagine a situation where you know that something exist
on your computer, you don't know where and you don't know any exact
phrase from the document. You sort of know what it should be and an
approximate date of creation. By scrambling together a few coarse limits
you could easily reduce several thousand documents down to 10-12
documents without to much fuzz.

Such fuzzy and probabilistic queries are usually not supported by
desktop search engines but they are often used in search engines for the
news industry.

> > I'm not quite sure but it seems like I'm focus on the retrieval
> > situation and how to find specific stuff while you focus on how to
add
> > such meta information? Could be we argue from different mental
models.
> 
> I would say that I'm focus on how to manage private data.
> 
> > Index			- list of descriptive words
> > Categories		- actual words (indices) in an index (metadata)
> 
> In my mail categories are defined by the user as a sort criteria.

You could use the sender's names, or an extracted person name from the
mail, or an automatic classification into private - work - spam.

Categories is strictly speaking only some specific grouping but I find
it convenient to use it in a slightly broader term as all types of words
and groupings that are meaningfull to use as a sort criteria or an
extract criteria.

> > Classification	- assigning categories to an item (file,
article, photo)
> >
> >> the metadata information must be stored in the "filesystem" because
it
> >> should be copied or moved with your file.
> >> try to move data with an emblem in nautilus, the emblem is lost. i
> >> would expect that such information even belongs to a file if its
moved
> >> by shell command, or copied via rsync for backup reason.
> >> With a flexible kind of metadata it may also be possible add
> >> information if the file was changed after copied. i often got the
> >> problem to find out what is never, the file at home or at work. the
> >> same problems appear with bookmarks, (yellow) notes, adresses etc.
> >> There are lots of solutions, but they individually depend on your
> >> software.
> >
> > What I was thinking of was to use the existing metadata solution
which
> > is an integral part of several file systems.
> 
> ?-)
> I don't know which file system supports other metadata than rights,
> owner and timestamp. only mac hpfs supports some more.

There are several and it becomes more and more common. Even if some
implementations place such metadata in a separate file structure the
available tools for the file system will support file operations that
keeps the relationships between the metadata and the file itself.

Note that metadata in a file system context is often refered to as
extended attributes.

Without checking I think Ext2/Ext3 keeps the extended attributes in a
separate directory while XFS assigns it to the file itself. Even FAT
file systems supports extended attributes.

I did a quick check and extended attributes are now supported by Linux
for ext2, ext3, jfs, xfs, RaiserFS. From other sources even more patches
for other file systems can be found. 

John