On Tue, 2009-08-18 at 17:40 -0400, David Zeuthen wrote:
> 1. what problems we want to solve for the user
Here some use cases I think that could be solved with some of those
shiny new technologies - but I'm not involved with any of tracker,
zeitgeist, couchdb, whatever, so I can't tell if they actually try to
solve them.
1. Tim is doing a video collage for his friend Foo. Therefore he wants
to browse all media files (audio, video and photos) related to his
friends Foo or Bar or himself.
1.1 Tim wants the oldest photo of Foo he got
1.2 Tim searches a specific photo he shot on his holidays that were
between date X and Y
1.3 Found a photo he did some funny things with in the GIMP but now he
would like to get the original photo back (and lets assume its
still there, somewhere on his hard disk)
1.4 After showing the collage his friend, his friend asks him about some
movie clips Tim used. His computer could easily tell him which
other video clips has been used to create this one.
2. Kim is a computer science student and his professors at the
university provide a pile of pdf files including presentations, scripts
and exercises. While working on a paper he is not sure about a algorithm
he needs but recalls one of his professors did a really good
explanation. But who was it? And in which of those hundred pdf files on
which page is it?
3. Tim sends (via mail or IM) Kim a good script that Kim didn't know
about. A day later he wants to look into it again but forgot where he
put it and how it was called.
[I've had talks like this a lot after sending someone a file via IM:
"How is the file named you just [yesterday/..] send me? I cant find it
anymore]
The problem that is to solve here is to improve effectiveness. You got
to interact with a several different programs and collect information
on your own - but your computer could do it for you.
--
Florian Ludwig <dino phidev org>
Attachment:
signature.asc
Description: This is a digitally signed message part