[hackerspaces] [Open Manufacturing] Re: Possible HackerSpace Hardware Inv. Software

Sam Putman atmanistan at gmail.com
Wed Jan 20 20:18:30 CET 2010


On Wed, Jan 20, 2010 at 10:45 AM, Bryan Bishop <kanzure at gmail.com> wrote:
> On Wed, Jan 20, 2010 at 12:32 PM, Sam Putman wrote:
>> So when I click on "share", something should happen? other than this:
>
> No. Right now the only thing that should work is "packages". In
> particular the best example is the "lego" package:
>
> http://heybryan.org:8081/package/
> http://heybryan.org:8081/package/lego/
>

Okay, well you have one user clamoring for the "share" functionality
as of now. Please don't ask me to write it myself :-)

>> b/c I still have that MakerBeam YAML file somewhere, and there's more
>> documentation coming.
>
> I don't know if I remember that or if I'm getting it confused with
> something else. It's not just YAML files that makes everything work..
> look around in the lego example and see how there's a catalog of parts
> (data.yaml), but also tools to help figure it all out (lego.py) and do
> "sense checking". To package something up you're going to have to
> start with a skeleton- like the lego example- and work from there.
>

So I'm going to need to write some python to make this work?

I am not thrilled at that prospect, offhand. Why can't I describe the
relevant data using YAML?

>> If I can't figure out how to use SKDB, or at least be convinced that
>> it's on its way to useful, I can't very well use SKDB to package the
>> MakerBeam project.
>
> To be honest, you don't have a reason to use skdb because you're
> buying all of your parts. It's easier for you to just go manually
> order parts from your suppliers. For projects with parts from multiple
> suppliers and potential replacement parts that may or may not be
> compatible, that's where skdb starts to get useful.
>

This is a serious misunderstanding of the project.

Aluminum extruders are not generally in the screw manufacture
business, nor do either of these types of suppliers generally do
stamping, plastic extrusion, injection molding or any of the other
specialized types of manufacture used in bulk manufacture of MakerBeam
components.

Also, wherever possible we are offering parts files for construction
using our available rapid-fab technologies. An example would be the
beam itself: if you need a small section of MakerBeam, with light-duty
tolerances for strain and proportions, you can take the STL file for
the beam itself and print one.

Or take the brackets. Currently our Alpha brackets are being cut on a
ShopBot with a vacuum hold-down system. We're in the process of
getting some of the brackets stamped, and working on laser-cutting the
rest. That's three ways to make a given bracket, depending on how much
you want to invest up-front, the choice of material, and available
machinery.

We need the "buy" and "make" buttons. That's why SKDB is interesting to us.

>> What needs to happen to change that? I have a test project: several
>> objects, including a screw fer crikey's sake, which have various
>> materials, file formats for description, and dependencies. It needs to
>> be packaged up somehow; if I can't do SKDB I'm just going to accept an
>> inferior but gorrramit functional alternative like Thingiverse or an
>> awkward tarball.
>
> So what's stopping you from typing in MakerBeam dependencies into an
> SKDB package? If you have already done this, then I have completely
> forgot and need to be linked again to these files. :-)
>

Because you have provided no documentation other than idiosyncratic
example files, basically. I made a YAML file, as I've indicated, and
saw no work done in return. I have no idea how to write up
dependencies in a useful way, and I'm not a python programmer.

>> I also volunteered to work on packaging the periodic table of
>> elements, and was told that this would be a waste of time. A
>> perspective I don't share, given that absolutely everything material
>> is built from said elements, but there you have it.
>
> Who told you it was a waste of time?
>

So there's no confusion, this was the exact quote:

"anyway none of this gets us any closer to having the entire ASTM catalog
of alloys and polymers available. might as well just start writing up
random facts about your favorite alloy. "

If this is how Ben feels about the project of documenting the periodic
table, I can't expect a lot of support in doing so.

>> and displays no function at present.
>
> Are you on Windows or something? That would explain a lot.
>

I run a modified BSD distribution with a proprietary user interface
called Cocoa.

I can produce a terminal on command. However, it's been two full
decades since the 1980s and a program without a usable graphic
interface is strictly weak sauce. I'd prefer that it be web based,
since a) it's better and b) you consistently tout this as a
Thingiverse killer, but a reasonably well-done Jython skin would do.

Furthermore, Windows still commands the largest installed user base.
If Windows users can't use it, you have a pretty big case of fail.

> Dunno, but I can't do everything on my own. It's really just been me
> for the past 6 months.
>
> - Bryan
>

Well I can't write your program for you, for a variety of reasons.
What happened to Ben, wasn't he a big part of the SKDB team?

I can work on packages, but I'm hesitant to put the work in without
some sense that it's all turning in to something usable.

Is there a roadmap? Because if it's just you, then documenting what
the project is supposed to become, in detail, would seem to be a
crucial element in getting SKDB off the ground.

I see no alternative to SKDB at present. I remain concerned that it
will never condense out of the vapors.

-Sam.


More information about the Discuss mailing list