New project: Aktions

[Zak Jensen]

Note: CC'ed the kde-quality and kde-usability mailing lists.

I guess I should first introduce myself, for those who do not know me.
I am Zak Jensen, a senior at Drexel University, and a long time
supporter of open source, the KDE project in particular. Some of you
may know me from kdedevelopers.org (where I have responded to some of
your blog posts), the kde-usability mailing list, or kde-artists. I
have been lurking on the kde-devel, kde-core-devel, and kde-quality
mailing lists for about 2 months now.

About 4 months ago, I proposed a replacement for the KTips dialog: a
mix between non-intrusive version of clippy, a plasmoid, and the
present KTips functionality.

http://lists.kde.org/?l=kde-usability&m=112609922114054&w=2
http://lists.kde.org/?l=kde-usability&m=112611125600245&w=2
http://lists.kde.org/?l=kde-usability&m=112611417629919&w=2
http://lists.kde.org/?l=kde-usability&m=112670485931144&w=2

And have also talked about it a bit on the kde-artists forums.
Unfortunately, the kollaboration forums have been relocated, and I
cannot find the original post at this time.

In any case, the idea has mutated from being an improvement of KTips
to a generic "action monitoring" service. The basic idea is to collect
abstract usage information from users, and analyze it. That analysis
could be used to target interfaces which should be investigated. The
information can also be used to pinpoint seldom used interfaces or
features.

It is important to mention at this point that we are providing an
abstraction to prevent the perception of our software as malignant
software. One way that we are doing this is by limiting the type of
information which the system will accept. We are limiting the input
information (in future versions) to abstractions like "typing",
"button click", and "toggle checkbox".

At the moment, we are finding out how the mechanism will work, or if
it will work at all. Future versions of our software will be built
with a proper outlook on security and privacy.

My senior design team wishes to integrate with an already existing
project, and I recommended KDE to them. To those ends, I was wondering
how interested the community is in the project. We are open to ideas
for features, feasibility, and recommendations on how to integrate
with KDE. We are not looking to be incorporated into the project at
this time. Rather, we require an audience to help us tailor and focus
our project, and we hope that there will be a few interested projects
within KDE.

Another portion of our project is defining a generic model for
graphical user interfaces which can be applied across several
platforms. If anyone knows of such a model which already exists
(excepting that defined by a set of libraries) details for that would
be greatly appreciated.

Attached are 4 documents (in the tar file) providing more detail for
our project. The best viewing of the html files occurs within
Konqueror, but they render alright within FireFox 1.5 as well.

Please forward any questions back to this thread or, if I  have posted
to a particular mailing list in err of its guidelines, to me
personally.

Thanks,

Zak
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Aktions.tar.gz
Type: application/x-gzip
Size: 63542 bytes
Desc: not available
Url : http://mail.kde.org/pipermail/kde-quality/attachments/20060315/82d693d9/attachment-0001.gz