Thank you for your replies.
I want to clarify a few things.
The purpose of the human avatar in our program is to provide a more natural
way of navigating through Wikipedia and not to be an all-knowing bot. It is
supposed to answer simple questions and offer the full article. And no, it's
not an office assistant; talking to a humanoid virtual character is
certainly more natural than talking to an edit box. The program that we are
developing is a 3D platform that can be used for organizing presentations in
a manner that exceeds the abstract GUI that is present in today's software.
The problem we face is the lack of 3D content. If we could obtain written
information and 3D models/animations(scenes, objects, processes etc.) linked
together then things could get really interesting. Of course it takes time
and technology isn't yet ready(speech recognition has more to go) for
talking to computers the same way as talking to humans, but we can already
develop software, based on existing speech recognition engines, that should
be usable in a few years after the speech technologies mature.
I am perfectly aware that a technology has to be free in order to be
accepted by Wikipedia. Our program has nothing to do with the extra types of
information that Wikipedia could have attached to it's articles; we'd just
like to have the possibility of attaching 3d content to articles(and we'd be
willing to help accomplish this).
The concerns related to the number of possible contributors are realistic,
but we have to take into consideration that more and more people get to know
free software like Blender(thus the cost of creating 3d models is the time
spent by eventual contributors) and I am sure that some of them would spend
some of their time sharing 3d models. Building simple 3D models is not that
hard; Google SketchUp is easy to use even by someone who has never used a 3d
modeling program before.
There are many free 3d model repositories online, but they are very
fragmented, store multiple types of files and the models have all kinds of
licenses. I am sure a lot of people use pictures from Wikipedia and the same
could happen for 3d models.
I also believe that having this feature with few 3D models is better than
having no 3D content at all. I don't want to write more about the usefulness
of 3D content because it should be obvious.
The technical part is the real problem here. The only way to do it in the
current technological context is to have users install a plug-in for their
browser. Static models can already be visualized with existing plug-ins(some
of them free/open source) or Java applets(Java will be free soon). In order
to make this type of content optional(because it would require extra
software installed by the users), the web pages generated by Wikipedia
should contain a static 2d image if the browser doesn't have the required
plug-in installed or the actual content otherwise. This way nobody could get
upset. At first, only static 3D models should be supported(because it's easy
and it's possible to do it in a short time). If it proves useful(imagine 3D
visualization of molecules for chemistry articles) it could be extended with
animations, interactive presentations and so on.
What do you think ?
Regards,
Ovidiu Sabou