[Edu-sig] RE: articles of possible interest

David Scherer dscherer@cmu.edu
Sun, 30 Apr 2000 14:41:03 -0400


Stephen Figgins writes:

> Has anyone metioned POO yet?  Maybe I missed it.
>
> http://www.strout.net/python/poo/doc.html
>
> I haven't played with poo (ick!) but it might make a good starting
> place for those interested in Python Moo development.

and then:

> Ah, and one more, MOOP, this one found lurking in the vaults.
>
> http://www.accessoft.com/moop/
>
> It has a more pleasant acronym.

The acronym may be the biggest difference - MOOP seems to be based closely
on the POO codebase.  I've played with POO (though not with MOOP - I
couldn't find a place to download it!), and I mentioned that in my first
post on this topic.

The POO code would undoubtedly come in handy in writing a system like I
describe, but the existing system has some serious shortcomings as a first
experience for novice programmers.  It tries to be friendly to people who
already know LambdaMOO, which has the side effect that you need to
understand both "@set foo to 42" and "foo=42".

> That said, I find news forums a more interesting place to meet others
> and learn about programming, and it has the added benefit of being
> archived.  In a chat or moo, the conversation is often lost, and
> because of when you logged on, or what room you are in at the moment,
> your questions might not reach the right eyes.

A MOO would complement archived forums, not replace them.  A newsgroup is by
far a better place to say, "I'm working on an X, and I need to do Y with
module Z.  I keep running into A; does anyone have any experience with
that?"  A MOO is a better place to say, "Hey, look at this!  How does this
thing work?  I wonder if I could use that for X?  Anyone here know where I
can find a tutorial on Y?  What have you been working on lately?  Is there
any way I can help out with that?"

http://www.cc.gatech.edu/~asb/papers/convergence.html is one of Amy
Bruckman's papers, from before "Moose Crossing."  I think it makes this
argument more effectively than I can.
http://www.cc.gatech.edu/fac/Amy.Bruckman/thesis/4-collaborative-learning.ht
ml#4.1 is also a vivid example of how supportive a MOO environment can be.

> And if I am to take the time building interactive objects in Python, I
> think I would like them to be more visible to the world.  What if we
> thought of the web itself as a giant moo?  With each site as its own
> sort of room?

It seems to me that there's an important difference between a MOO and the
web.  The web technology is built around a "publisher/reader" model, and
MOOs are built around a much more fundamentally interactive one.  From the
first paper I linked to above:

| If [VR environments are] "interactive," it is in the limited sense that
| most hypertext systems are interactive: there are multiple paths
| through the material, and the system has a limited ability to react
| to the user. However, the ways in which the system reacts are designed
| by the artists and engineers who constructed it and not by the users.

There's also a big difference in the level of social contact involved.  Two
people looking at the same web site don't ever come in contact with each
other, or with the site's author.

Even at sites that provide forums, the vast majority of visitors lurk.  For
example, there are 164 people subscribed to edu-sig right now, and others
reading the archives.  Since 2/1, there have been roughly 420 posts by 63
people.  (Hi, lurkers! :)

Dave