the mechanics of sharing physical design information

[Michel Bauwens]

Here´s a proposal by Jon Kuniholm of the Open Prosthetics and Shared Design Institute:

I'm particularly interested in the mechanics of sharing physical design
information, and I'd like to get some wiki-based discussion going on the
specifications for an open source software suite of tools to make this
easier, including CAD software.

Thanks for your consideration.

Jon Kuniholm

[Jon Kuniholm]

I think that the way that I have found my way to this network is a perfect example of some of the problems that face us in using available tools for collaboration on design. While I found out about you all from the P2P Foundation Wiki, it appears that your Ning network may be the most functional tool that you have for generating interaction, and I was unable to get a login to help with the wiki. The Open Prosthetics Project Ning group is also our most successful site for attracting users. There are other tools that are better suited to specific tasks associated with design collaboration. The linear nature of blog and forum posts, for example, don't lend themselves to organized collective thinking or to collaboration.

I'll paste below the wiki entry that I did in raising this question on my wiki (avoid the broken links below), as a straw man for getting the conversation started.

As you may have gathered, we are using a collection of free and open tools to make this happen, including the [http://openprosthetics.wikispot.org/ wiki], the [http://www.openprosthetics.org/ main site], the [http://openprosthetics.ning.com/ social network], the [http://groups.google.com/group/openprosthetics Google Group], our Google Code page for the [http://code.google.com/p/myopen/ Myopen Project], our http://www.instructables.com/group/openprosthetics/ [Instructables page], and probably 15 more we forgot. It’s not perfect. As we said in the blog post on Ning, "if you are a master programmer and looking for a challenge, we’d love for someone to help us build a fantastic Sourceforge-type project management and collaboration tool with social networking capability, as well as the next generation of collaborative open source CAD tools to drive this and other open design projects."

There's already stub for a page on the [http://openprosthetics.wikispot.org/Open_Source_CAD_Program CAD problem], which would be great, but is likely a much more challenging problem that is a separate issue. That said, as Mako Hill pointed out to us, perhaps an open source CAD program could make some headway if it offered a collaborative aspect in the same way that Google Docs does, causing users to overlook its limitations. I would call the availability of a free and open CAD package a near second to the problem of overall project management, organization and communication between participants.

First, not much is going on towards fixing this problem. The discussion consists of vague pleas [http://openprosthetics.ning.com/profiles/blog/show?id=1492079%3ABlo... like this one] without good documentation. Let me first say that I think that solving this problem is one of the first things I'd try to solve if we had the resources to do it, because it is a solution to the meta issue of how collaboration on open physical design for any underserved market happens, and represents a tool for the core mission of our non-profit, [http://www.shareddesign.org/ the Shared Design Alliance]. It would also increase the impact of so many of the other things we're trying to do. In order for the overall mission to succeed, and for us to facilitate efforts similar to those of Open Prosthetics on behalf of some other underserved market, we'll need to have an integrated package of tools that solves as many of our problems as possibl

[Michel Bauwens]

Hi Jon, our wiki is a web of trust, this means you have to ask an existing member for access, so you simply need to fire off an email to me and I will send you id/pw ,

Michel

[Michel Bauwens]

Interesting email note from eric hunting:

We are discussing a chess game with an extremely large number of different pieces. There are no artifacts or systems which have only one possible design and the nature of this design determines the fabrication methods and thus the potential minimum of scale of production facility and the nature of the infrastructure that is needed to support it. Design and fabrication method are interdependent. Design limits the range of possible fabrication method. Fabrication method limits the range of possible design. When a product concept is completely new and there is no established production facilities for it. design leads fabrication. When a product is an incremental improvement of existing products with an existing investment in production facilities and infrastructure or when other logistics limit fabrication method options (like choosing to minimize facility scale and production overhead), fabrication leads design. In the pathology of product development you eventually realize a reconciliation between these two sides determined by cost and competition -how much a company is willing to invest in new production capability, and hence novelty in design, for an edge in market share. By extension, fabrication method determined the logistics of production which, in turn, influences the architecture of an overall industrial infrastructure. However, where that infrastructure is already established in some form, it feeds back on production logistics in the manner of 'cost effectiveness' or 'practicality', which ultimately trickles up into limitations upon design. Thus we arrive at the cybernetics (in the classical sense) of an industrial infrastructure. Design can thus influence an overall industrial infrastructure, or it limits can be dictated by it.

[Michel Bauwens]

From Bryan Bishop, via email:

Hey Jon, glad you brought up this topic. I'm presently sitting in the
Automated Design Lab - quite literally on some rather comfortable
chairs - while the computers are humming away generating CFGs,
graphical depictions of designs of various artifacts from a repository
that I've been throwing together with an open source group that tends
to post to that openmanufacturing mailing list that Michel mentions.
As you are aware, the CAD interchange formats suck immensely, and they
do not really capture all aspects of design as they are now.

There's some data sets that you can go look at *right now*:
http://heybryan.org/~bbishop/docs/

In particular, there's the repo/ directory which contains an XML
description of the interconnection of parts of various consumer
products. This is a direct rip from the MST
repository.designengineeringlab.org website. The prettyrepo.zip file
might be more useful since it's significantly smaller. There's also
the biobricks.zip file which is quickly approaching some design
sharing format in general, thanks to some new developments from the
diybio.org mailing list and on the IRC channels.

What we've been figuring is that we're going to be using YAML metadata
to specify the overall information about designs and provide an exact,
computational, specific way of representing and specifying all of the
important details about interconnections and other properties of the
design. Anyway, as you say, you're interested in the mechanics of the
sharing, perhaps not so much the how-the-knowledge-is-represented
issue.

So, you might be interested in the mechanisms that we've been using
with the community at large -- various git repositories. This is one
repository that has been used for the biotech toolkit and bioreactor
project, which I consider a subset or child project of the larger
ideas flying back and forth among us:

http://heybryan.org/gitweb.cgi
http://heybryan.org/biotech.git

There's also a somewhat nonworking ikiwiki + git interface to the
dataset to help facilitate web-users to peak into the data set in some
wiki-ish interface that they apparently love so much.