[hackerspaces] Idea: things.hackerspaces.org now that Thingiverse is not ok anymore

Matt Joyce matt at nycresistor.com
Wed Sep 26 20:57:58 CEST 2012

Just a quick update on thingiverse.


Curious if anyone has any concerns not addressed in there.

On Tue, Sep 25, 2012 at 7:32 PM, Yves Quemener <quemener.yves at free.fr> wrote:
> On 09/25/2012 12:17 PM, Matt Joyce wrote:
>> In case anyone is into discussing it.
> Hello Matt
> Your article triggered a lot of thoughts.
> I wonder if the open hardware community is not trying too hard to mimic the
> free software movement without considering the differences between hardware
> and software.

I'd say they are making a mistake mimicking the Free Software
movement.  They should be mimicking other successful Open Source
software companies.

> I think that the open source licenses were the core tool in making the free
> software movement the success it is today. The GPL especially made a very
> good job at using the specifics of the software world and using them to make
> new kind of collaborations possible and profitable to the whole community.

I'd disagree that the GPL makes new kinds of collaboration possible.
The reason BSD and Apache licenses are used a hell of a lot more than
GPL in business is because of the avenues of collaboration that are
cut off by GNU / GPL.  At the very heart of the the fundamental
disconnect between commercial interest and free software lies GPL.

> Maybe does the open hardware movement need to make its own set of rules,
> adapted to the hardware world?

That would be very interesting.  A hardware equivalent to the Apache
license would be really cool.  As would an Open Hardware Foundation to
administer it.  Maybe folks at the OHS this week will discuss such

> As I see it, there are two different kind of open hardware today :
> - The kind that has its blueprints freely available but that needs to be
> assembled from bought machined parts. The reprap, the makerbots.
> - The kind that can be generated totally from raw materials and open
> hardware : most objects on thingiverse.

Okay.  Not a very formal definition.  OSHW has some specific formal
definitions.  I kinda feel like they might need some revision.

> The analogy with software works great for the second kind. Software is cheap
> to copy and does not require manual work. Even if printed parts require a
> lot more time, electricity and a few cents of plastics, the analogy can work
> well.


> But the first kind is a very different beast. It is similar to the situation
> where the free software world was when neither linux nor gcc existed. One
> can still find examples of free software that require proprietary content to
> work correctly, but they are not the norm. In the open hardware world, most
> of the useful things are not 100% printable or CNC-able (not even 50%).
> Until we are able to print electrical engines, board, chips, metal axis,
> etc. this will be the situation.

I think there's two sides to this point.  GCC came about as a natural
solution to the fundamental necessity of compilers.  When compilers
were new, and C was new, there was no immediate need for such open
source software.  Companies worked very had to develop that technology
at that time and they needed to fund that work.  So they sold their
compilers either as stand alone or part of another product.  Later, as
compilers became a fundamental building block of software, open source
and free software democratized that technology and allowed for market
driven innovation to focus on emerging technology.

Markets of scale arguments are another thing entirely.  In the case of
a hardware fabricator a small shop simply cannot compete with the
foxconns of the world.  And so they cannot rely on their having
invented a product to ensure their success.  A bigger fish can simply
out pace them and sell cheaper.

On is a predatory threat, the other is merely a discussion of when
does technology become a target for open source?

> What is the best way to deal with that?
> Makerbot used to be a facilitator for the duplication of the first kind of
> hardware : selling prepared kits or even completely mounted machines, they
> made duplication of printing machines affordable and not time-consuming.
> With their change of policy, they now only focus on helping the duplication
> of the second kind of open hardware : the printable ones.

I think that's true to a point.  I think they can certainly continue
to drive research in rep rap and other open source kit communities by
releasing their work staggered, or as component designs.  I hope that
they do.

> This is still useful (and I still support them) but that means that
> suddenly, they will probably stop doing efforts on the first kind of open
> hardware. A free software analogy would be Linus Torvalds saying that v2.0
> of the kernel would become proprietary but would still be a full POSIX OS
> and would support GNU with dedication. It would have angered many people,
> but would not have prevented the development of new OSS tools. (Actually
> this is a bit dishonest comparison because Linus does not have to run a
> manufacturing business to release Linux)

I disagree.  By the time Linus wrote his Linux OS, Unix like operating
systems were plentiful.  But, in the earliest days Unix cost money.
When the technology is still emerging it's sold as a commercial
product.  But later one, as it became a fundamental necessity of other
products it was standardized and turned into first BSD then minix and

3D printing isn't "new" technology persay.  But like Unix it will be
an emerging technology for decades, as the hardware and methods
develop.  At some point the ubiquitous need for 3D printing will drive
a legitimate open source adoption.

RepRap began the first attempt to do that.  Makerbot launched an
attempt to build a successful commercial entity on top of that effort.

The way I see it, Makerbot is the redhat to reprap's proverbial linux.
 And when it comes right down to it.  Redhat made linux more than just
a computer science hobby kit.  It made it a defacto standard and an
open source one at that.  However even today parts of redhat are
special sauce and closed source.  Still their contributions to open
source are vast.

Sure it's harder copy and distribute a printer.  But in both cases a
return on investment in the open source technology was needed to drive
development and adoption.  In both cases a commercial entity rose up
and was able to pursue a hybrid open / closed source model.  I'd say
there is a clear parallel.  Not much of a difference at all.

> Isn't there a way to find a license that would allow duplication
> facilitators of the first kind of open hardware to make a living? I know
> that open-source types frown upon "non-commercial" clauses, like in the CC,
> but couldn't something like that (maybe with a short time limit) help
> makerbot make a living while being a community actor ? I could imagine a
> clause saying "until <this date+one year>, only non-commercial duplication
> is allowed".

Open source doesn't frown on non-commercial clauses.  Free software
does.  These are two very different things.

You can be open source and still proprietary.

Heck I actually agreed with this approach in my post.

> This would prevent carbon-copy competition. This would not "protect profit
> generating innovation", as you put it, but this part seems antinomic with
> what the whole open-source/hardware movement is about.

As I said, I agreed in my post.  It is interesting discussion though.

> Yves


More information about the Discuss mailing list