Collaborative Design, The Wiki Approach

To get a sense of how this collaborative design will work,
pick a subject that you know quite a bit about. Look it
up in Wikipedia. Hopefully you can find something on which
you know a good bit more than Wikipedia currently records.

Try to edit the Wikipedia article so that
- it is free of your own biases
- it presents facts, not your opinions
- those facts come from written sources which you cite
- it is fun to read and relevant to the subject at hand

You will discover that this process is a load of work. While
you are doing it, you will note that other people will note
that your revisions do not accurately reflect *their* biases
and opinion, so they'll insert those, along with unattributed
heresay.

If you can't discern the difference between opinions and
fact, none of this will bother you, but on the other hand,
you also won't make much forward progress towards designing
anything real and so you are unlikely to affect many other
people. There is a big difference between a sum of the
world's knowledge and a sum of the world's opinions.

Separate note: I know of one team that is using a wiki to
store and revision engineering documents for a CPU design.
Within a company, where people hacking on the wiki can be
held accountable for their contributions, I think a wiki
could be a simple and powerful paradigm....

but we don't use it where I work, so I don't really know.

wrote:

There is a big difference between a sum of the
world's knowledge and a sum of the world's opinions.

Oh! Good one!

/kenw
Ken Wallewein
K&M Systems Integration
Phone (403)274-7848
Fax (403)275-4535

www.kmsi.net

wrote:
Hi again,

I was just browsing through Wikipedia recently -- I'd heard about it,
just never bothered to actually check it out before.

http://en.wikipedia.org/wiki/Wikipedia:Introduction

I was marvelling at how an encyclopedia can evolve just from random
visitors editing it and putting in their own 2 cents of opinion.

If I'm not mistaken Linux was built by such a collaborative effort.

--
Hop David
http://clowder.net/hop/index.html

wrote:
Well certainly, on many web bulletin board forum discussion sites, you
have to register to get an account before you can post. People who
abuse or misuse their privileges get them revoked. Likewise, this could
be the case for a CAD wiki.

The problem is that maybe for small projects - a bicycle perhaps, this
could work.

For huge projects, who maintains the interfaces?
For example, if one bit is a rocket engine Wiki, and another is
the fuel tank Wiki, how do you police inadequate propagation of changes
between the two that'll make the result not work when it's fitted
together.

Hop David writes:

wrote:
Hi again,

I was just browsing through Wikipedia recently -- I'd heard about it,
just never bothered to actually check it out before.

http://en.wikipedia.org/wiki/Wikipedia:Introduction

I was marvelling at how an encyclopedia can evolve just from random
visitors editing it and putting in their own 2 cents of opinion.

If I'm not mistaken Linux was built by such a collaborative effort.

Yes, but the prove of every line of code in Linux was how it worked. In
knowledge or even in building spacecrafts this is not as easy to test.


Jochem

--
"A designer knows he has arrived at perfection not when there is no
longer anything to add, but when there is no longer anything to take away."
- Antoine de Saint-Exupery

Hop David wrote:
wrote:
Hi again,

I was just browsing through Wikipedia recently -- I'd heard about it,
just never bothered to actually check it out before.

http://en.wikipedia.org/wiki/Wikipedia:Introduction

I was marvelling at how an encyclopedia can evolve just from random
visitors editing it and putting in their own 2 cents of opinion.

If I'm not mistaken Linux was built by such a collaborative effort.


But wasn't part of the key with Linux that a) it had a relatively small
starting point, and b) that there were a few key personalities that
provided shape and guidance?

And other collabrative efforts have stalled or worse; or, like GNU,
ended up running off in a different direction than the original vision.

/dps

David If I'm not mistaken Linux was built by such a collaborative
effort.

Linux had one guy doing all the integration for a very long time, who
had
a surplus of good taste in acceptable code, plenty of people checking
his
work for correct functionality, and public tolerance of a vast number
of
failures. Integration is, BTW, generally a thankless grinding job that
most engineers would flee from in a moment. Linus found it different
because he did not have supervisors requiring that certain code go in
because some customer obligation was higher priority than Linus' taste
in
acceptable code -- and yet he still had "customers", quite an amazing
situation.

Also, a wiki where you need to get a membership has been tried:
Nupedia.
This was the failed predecessor of Wikipedia. Look it up on Wikipedia,
there should be a good article on what went wrong.

The world is still feeling out the best ways of utilizing the great
surplus of professional talent brought about by widespread higher
education and computerization. It appears that the vast majority of
people do not understand how hard things get done (small numbers of
dedicated people make ludicrous sacrifices to drive things to
completion), and assume that somehow lots of people making small
contributions can do as much as small numbers giving it their all.
Unfortunately, nothing has changed about the mythical man-month.

Hop David wrote:


wrote:
Hi again,

I was just browsing through Wikipedia recently -- I'd heard about it,
just never bothered to actually check it out before.

http://en.wikipedia.org/wiki/Wikipedia:Introduction

I was marvelling at how an encyclopedia can evolve just from random
visitors editing it and putting in their own 2 cents of opinion.

If I'm not mistaken Linux was built by such a collaborative effort.

I'd have to say that in some ways you are.
Linus was a big part of what made Linux happen.
In that he pretty much determines what goes in, and what stays out,
and who is responsible for which bit.
It certainly wouldn't have happened if anyone could add source without
regard to an overall plan, or even vision.
That way lies madness.

the Artemis Society, precursor to the Moon Society, is a wiki org for
designing the first civilian flight to the moon.
It works, for designing something. It's a beautiful thing to regard
when people write a detailed plan together.
When it comes to building flight hardware, which requires money and
manual labor, wiki may not work too well.
One problem, a la Gregory Bennett, is that the wiki organizer
acquires a valuable knowledge base which is salable on the open market
and he goes to Bigelow Aerospace. :-) Wiki remains a valuable
amateur technique and can get communal ideas organized on paper but can
it manipulate materials to build real things?

Well, while surfing around I came across a free wiki website:
www.pbwiki.com

It allows you to sign up for your own wiki, just as easily as you would
for a webmail or blog account.

Editing privileges are passworded, so that you have to distribute the
password to others in order for them to be able to edit it. This can
then keep things exclusive enough to stave off the chaos, etc.

I guess the key distinctions with a wiki is that it exploits ease of
input to encourage the buildup of information from a wider body of
people. Unlike a blog, which is a linear/serial record, the wiki gives
a topographical presentation of inputted information and is mainly
geared to showing you the 'latest and greatest' info, although
obviously you can look back at history.

I've been reading about SVG (Scalable Vector Graphics) lately, since
it's about to be released for MS Internet Explorer and Mozilla Firefox,
and I think that the advent of this would be critical to the
realization of wiki-CAD. Right now, when you login to your
webmail/blog/wiki you can merely input and format text through a text
editor. But as soon as SVG becomes a standard baseline feature in
browsing the internet, then it stands to reason that the various
webmail/blog/wiki service providers would improve their editor GUIs to
permit graphical editing/input.

(Gee, I wonder if Usenet wouldn't similarly adapt to SVG, since it's
all encoded in text format anyway. Sci.Space.Tech could one day be
bristling with vectorized blueprints and drawings. Obviously, posting
image binaries directly in here is considered in appropriate and
bandwidth-wasting, but would the more compact vectorized SVG format be
considered permissible?)

Note that I won't include the existing use of Macromedia Flash or Adobe
Illustrator, since these are largely closed formats which limits the
way in which graphical editors can be designed to produce them.

Once SVG opens up the web to facilitate 2D CAD, then it would stand to
reason that 3D CAD will soon follow, since the 3D-manipulation exists
inside the editor, and the output will always be a 2D projection in the
same old format anyway.

So conceivably, we could soon have full-fledged CAD-wikis allowing for
mass collaborative design.
Gee, and once you have wiki-CAD, perhaps you can incorporate a physics
engine into your web-based editor, and then suddenly you could have
simulation-wikis. Maybe the wiki approach could then be applied for
videogame design purposes, since videogaming is a hyperactive market.

Perhaps all of this would greatly streamline and accelerate the
rapid-prototyping process for engineering, so that concepts could be
ever more rapidly developed and actualized.

Disruptive technology? Comments?