Quoting Hoeken <[log in to unmask]>:
> The continuous activity on this mailing list is very exciting to me
> and I love how the website keeps evolving. I'm wondering what the
> near future plans for reprap are? What immediate steps could someone
> like me take?
I'm going (shortly I hope) to put a coherent list of things that we need on the
website. It's perhaps also appropriate to say what we've got locally at Bath:
1. We have two conventional RP machines that we can use to make
parts and to test ideas.
2. We have mechanical design expertise (in my student Ed Sells,
who is a first-rate designer), and
3. I have a lot of experience of large-scale commercial software
development and of digital electronic design for control systems.
4. We have applied for funding for three years to do the project,
but that has not yet been approved.
> It seems that if we are to build this so that it can produce copies of
> all the parts that it needs, then we should start creating 3d models
> of all the necessary parts so that researchers across the world can
> replicate those parts. If we use the standard RP file format, it
> should be possible to get those parts fabbed at a commercial RP place,
> then use those parts to bootstrap the first reprap.
The ubiquitous format is STL files, but we maye use OpenRP as well. One thing
we're going to do as soon as possible is to lay down some standard specs and
define interfaces (soft, electronic, and mechanical) so that we can break down
the design problem into independent modules that others can address.
> Really, this project is very similar to an open source project. You
> can imagine that the RP unit as nothing more than a real world 3
> dimensional compiler, then you can also imagine how the parts the make
> it up are nothing more than files to be compiled and linked up. In
> this case, the reprap would be our compiler. It takes the object
> files, and prints them out in 3d.
A good analogy.
> Since everything is a file, we can start work on the project right
> now, over the internet. We just need a CAD program that is good
> enough and can handle the file types we'll need to give to commercial
> RP to bootstrap. Once we find one that is good enough we can improve
> it to suit our specific needs.
>
> Once we get a good CAD program we can start to design the system.
> There will definitely have to be a well defined feature set for each
> release. Since this is a physical machine, this will be just as
> important, perhaps even moreso than in software development.
As long as we use CAD systems that can intercommunicate (using STEP for
example), then it's not even necessary for everyone to use the same one. But
we do need one open-source CAD system to distribute with the system. STL files
are too low-level to act as a CAD-communication medium.
> It
> should be decided exactly what needs to be in each release (what
> materials it will use, the method of operation, etc.) Keep it simple
> at first, and add features once the previous features are proven to
> work. Once that has been decided then a Todo list can be setup and
> those who want to pitch in have an easy place to start.
>
> I propose that you set up a CVS system for all aspects of the system.
> That way all the files are available. I am pretty sure that I have
> access to a RP facility close to where I live, and I would love to be
> able to download a copy of this and play around and contribute back.
>
> So I guess i'll wrap up with a summary of what I would love to happen
> in the near future:
>
> 1. Settle on a 3d CAD system that meets our basic needs
> 2. Define the first phase requirements - as barebones as possible.
> 3. Start creating cad files to create #2
> 4. test, improve, test, improve
> 5. watch as we change the world.
That's more or less what we intend...
> PS. Keep up the great work!
Thanks!
Yours
Adrian
http://staff.bath.ac.uk/ensab
|