Why are Makers still fooling around with toys and mash-ups and not
making serious things? (short answer; like early computer hackers
lacking off-the-shelf media to study, they're still stuck reverse-
engineering the off-the-shelf products of existing industry to learn
how the technology works and hacking is easier than making something
from scratch)
Why are Makers rarely employing many of the modular building systems
that have been around since the start of the 20th century? Why do so
few tech-savvy people seem to know what T-slot is when it's ubiquitous
in industrial automation? Why little use of Box Beam/ Grid Beam when
its cheap, easy, and has been around since the 1960s? Why does no one
in the world seem to know the origin and name of the rod and clamp
framing system used in the RepRap? (short answer: no definitive
sources of information)
Why are 'recipes' in places like Make and Instructibles most about
artifacts and rarely about tools and techniques? (short answer;
knowledge of these are being disseminated ad hoc)
Why is it so hard to collectivize support and interest for open source
artifact projects and why are forums like Open Manufacture spending
more time in discussion of theory rather than nuts & bolts making?
(short answer; no equivalent of Source Forge for a formal definition
of hardware projects -though this is tentatively being developed- and
no generally acknowledged definitive channel of communication about
open manufacturing activity)
Why are Fab Labs not self-replicating their own tools? (short answer;
no comprehensive body of open source designs for those tools and no
organized effort to reverse-engineer off-the-shelf tools to create
those open source versions)
Why is there no definitive 'users manual' for the Fab Lab, its tools,
and common techniques? (short answer; no one has bothered to write it
yet)
Why is there no Fab Lab in my neighborhood? Why so few university Fab
Labs so far? Why is it so hard to find support for Fab Lab in certain
places even in the western world? (short answer; 99% of even the
educated population still doesn't know what the hell a Fab Lab is or
what the tools it's based on are)
Why do key Post-Industrial cultural concepts remain nascent in the
contemporary culture, failing to coalesce into a cultural critical
mass? Why are entrepreneurship, cooperative entrepreneurship, and
community support networks still left largely out of the popular
discussion on recovery from the current economic crash? Why do
advocates of Post-Industrial culture and economics still often hang
their hopes on nanotechnology when so much could be done with the
technology at-hand? (short answer; no complete or documented working
models to demonstrate potential with)
Are you, as I am, starting to see a pattern here? It seems like
there's a Missing Link in the form of a kind of communications or
media gap. There is Maker media -thanks largely to the cultural
phenomenon triggered by Make magazine. But it's dominated by ad hoc
individual media produced and published on-line to communicate the
designs for individual artifacts while largely ignoring the tools.
People are learning by making, but they never seem to get the whole
picture of what they potentially could make because they aren't
getting the complete picture of what the tools are and what they're
capable of.
We seem to basically be in the MITS Altair, Computer Shack, Computer
Faire, Creative Computing, 2600 era of independent industry. A Hacker
era. Remember the early days of the personal computer? You had these
fairs, users groups, and computer stores like Computer Shack basically
acting like ad hoc ashrams of the new technology because there were no
other definitive sources of knowledge. This is exactly what Maker
fairs, Fab Labs, and forums like this one are doing. And the magazine-
dominated media for personal computing at the time was all about DIY
programming in BASIC or Pascal and hacking compared to the
contemporary computing media which is all about new products and the
elites and corporations who make them. Again, this is just like Make
magazine. As yet there is no media yet showcasing and reviewing
fabrication tools the way PC Magazine does PCs or 'market' magazines
like Computer Shopper collecting small dealers of stuff, but this is a
readily anticipated future development. And you had visionary books
like Computer Lib evangelizing the technology, culture, and it's
future potential. And that's exactly what Neil Gershenfeld's book FAB
is.
There are a lot of parallels here to the early personal computer era,
except for a couple of things; there's no equivalent of Apple (yet..),
no equivalent of the O'Reily Nutshell book series, no "##### For
Dummies" books. Now, I'm not exactly sure that having an Apple in this
field would be a good thing. Apple and companies like it then were key
to opening the door to a very open and personal Information Age. The
nature of personal computing the industry was pointing toward before
Apple was one that looked like MiniTel and laptop Videotext keyboard
terminals plugged into TVs. Apple came out of the sub-culture and
turned it mainstream, making its technology socally relevant and
ubiquitous, even if ultimately doing so for the sake of creating a
'mainstream market' to exploit for profit. But it also helped
establish a traditional Industrial Age hegemony based on turning
personal computers from tools people understood into mass-produced
appliances the public used and yet never really came to comprehend and
so remains dependent upon a corporate hegemony for. One might argue
that this was a necessary evil by virtue of the complexity of the
technology and the failure of public education to embrace and
disseminate new knowledge effectively. But the end result was that the
promise of social liberation information technology originally
promised was superseded and forgotten. I think this is cautionary
history. Do we want a Fabber that no one ever opens up or truly
understands the working of? Or is this, on some level, still a
necessary evil for the sake of realizing ubiquity? 40 years later,
Grandma still isn't writing code and still has trouble finding the on-
switch. What exactly does that mean for these new tools?
Then as now, there was this gap in the parallel dissemination of
knowledge with the technology. Disseminating the power of computing
very quickly went from being about disseminating knowledge to putting
mass produced products in people's hands at a 'reasonable' cost,
largely because the communication of knowledge proved so much more
difficult for an increasingly exclusive sub-culture composed largely
of misanthropic nerds to pull-off than selling the technology by the
novelties of its turn-key applications and 'style'. (pay no attention
to that man behind the stylish ABS clamshell case...) But the end
result of that strategy is clear; Silicon Detroit. Real progress in
personal computer development is now reduced to a snail's-pace because
of the drag of hegemonies. Moore's Law isn't translating into
productivity gains, let alone social liberation.
I think the folks at MIT may have clued into this issue early on in
devising the concept of the Fab Lab. But their solution is typically
academic-minded. The Fab Lab is an ashram. There's nothing wrong with
that. It's a good model. However, it's ability to communicate is
limited by proximity. So you have to make a lot of them to spread the
message. And that is limited by the population of experts since it
doesn't leverage their accumulated knowledge that well. This is being
done remarkably swiftly for what it is, but if it was in competition
with an Apple Fabber -and it may soon be- it would lose to PC style
market cultivation over true education. The key limitation here is
that Fab Labs don't don't network their own knowledge coherently and
they don't publish.
The Make magazine and blog publishes. Thing is, they don't quite know
what they're doing or where it's going. They're surfing a meme with a
YouTube model, letting it carry them rather than directing it in any
particular way. It works extremely well as a medium for the exchange
of incoherent hacker-style knowledge and it's much more culturally
accessible because of its visual media, but it's not that good at
communicating knowledge in an organized way and as the sub-culture it
embodies becomes more technically sophisticated, it will inadvertently
produce barriers to its own accessibility. Eventually, fewer and fewer
articles and recipes will make any sense to John Q. Public. But take
note of something. Who is their parent publishing company? O'Reily.
Someone upstairs may be watching and waiting for some sign that it's
time for some strategic partnerships and "Epilog Laser in a Nutshell"...
Thinking on this, I'm struck by the question of what, then, is the
functional role of an open manufacturing movement in the context of
all this? What's the Missing Link? What collectively should the
activities of this movement coalesce around for maximum cultural
impact? The answer seems to be organized knowledge. If we are to
preclude the outcome of an Apple Fabber degenerating the movement into
another Silicon Detroit the key is in the effective dissemination of
knowledge over and ahead of the dissemination of products. And that
would suggest that what's missing in this movement is publishing. The
systematic creation and dissemination of a large body of knowledge in
the most accessible forms possible. Fab Labs have lead the way in this
task, but they don't publish and so they don't leverage their
dissemination potential that broadly. Make and the like publish, but
they don't publish much that's coherent. They're 2600 for Makers.
This brings me to the notion that, perhaps, the effective role of the
movement is that of a knowledge generating and publishing engine with
the objective being to use the generation of media as a means of
achieving ubiquitous accessibility for the technology through the
accessibility of the knowledge in parallel with the technology. in
other words, technology disseminated merely through the dissemination
of products is a process of encryption that limits limits its
liberating potential. But technology disseminated through the
dissemination of knowledge through media is a process of decryption
that enables its liberating potential. We need to think not just about
the cultivation of the technology itself but the forms -media forms-
in which we disseminate it and what cultural impact that has. So lets
imagine a Fab Lab -or the Fab Labs collectively- as an ashram that
publishes. A movement as an engineering laboratory whose output isn't
products but media about the technology it cultivates.
This is sort of what I was thinking about with the Vajra Maker
Incubator concept. It might run an open Fab shop, but it's key job is
to gather and cultivate open technology -and the people who invent it-
while producing media about what its doing and how its doing it -which
would be how its residents earn a living. The community creates a
haven for this intellectual activity -distributing the high cost of
tools- and pays for itself on the publishing royalties -which is much
the same model as the traditional artists commune which pays for
itself by creating an environment conducive to those artists'
creativity and then gets a cut on the sale of what they produce to
keep going. Corporate industrial research laboratories do much the
same thing, only they're objective is to output patents -an
intellectual and technological real estate scam. This community would
be outputting the same thing, intellectually, only it's in the form of
open technology which people make a living from based on the sale of
the media that conveys it rather than controlling and exploiting the
use of it -which is one of the common ways Linux and other open
software is 'monetized'. This seems to parallel Cory Doctorow's and
others vision of the Maker Monastery as a bastion of EcoTech and heart
of an Outquisition movement. (though I'm liking the ashram analogy
much better. Much less Gothic in aspect) Self sufficiency isn't
possible for such a haven early on given the available technology. So
it either needs a lot of extremely wealthy sugar daddies to support it
indefinitely as a gift -and life preserver- to future generations,
bootstraps from a very low-tech agrarian model, or it plugs-into and
exploits the outside economy in a way their least interferes with its
social objectives -and publishing seems to suit that. Maybe any
combination of all three would work.
Now this notion isn't necessarily limited to some community in one
place. As Make demonstrates, thanks to the Internet our ashrams can be
virtual. Living in proximity to tools one can't afford alone is a big
help, of course, but the key here is the community network and the
development and publishing models. Not all makers are particularly
good writers, illustrators, photographers, and videographers. The
production of marketable media will likely often demand collaboration.
And not all these people need continuous access to workshops. Also,
some people will personally invest more in personal workshop
facilities simply because they need more privacy for the sake of their
own productivity. And, of course, for this to be effective as a
movement it has to be able to function across many locations around
the globe.
This is where that word ToolBook comes in. This is my imagined 'brand'
for the publishing cooperative that this community would coalesce
around. It would be like the 'Nutshell' and 'Dummies' books series
name but would also be the publishing co-op company's name, ultimately
a multi-national corporation. The basic 'deal' of the community is
that, collectively, it owns a series of resources -workshops, live-in
villages, data centers for web hosting and such, media production
facilities, supplies warehouses, etc., under the ToolBook Publishing
company name with individuals sharing ownership through stock in that
company -much like a Kelsonian Community Investment Corporation. The
individual maker in this community applies for facilities for any
particular development project which is intended to culminate in the
production of one or more forms of media whose sales -or ad revenue-
will recoup the expense of development. Now some things might be
really small and simple so all the maker is looking for would be
publishing assistance akin to LuLu.com, for which he receives a
royalty on unit sales. If he must collaborate with writers or artists,
they too would share those royalties. Other projects may require a lot
of facilities and thus represent a greater investment decision based
on the maker's reputation and the collective opinion on the value and
importance of the project among the whole community. So here is where
peer-to-peer organization an social credit come in with projects
needing increasing communal support and collaboration the more
elaborate they are and a formal community-wide project submission,
evaluation, and management scheme, web based. And if the community
member is living in a community-owned village then they are sharing
some portion their royalties with that local community by way of
'rent' relative to their other forms of input to the local economy of
that village. A diverse collection of media would be produced by this
co-op; print and eBook media in casual to college-level forms,
streaming and packaged video, blogs, events, talks, etc. And it would
go beyond media to include project kits, small production or made on
demand products, one-of-a-kind pieces, and so on. Individual community
member activities could also split-off into industrial production -
with both an external for-profit component and an internal community
support component- particular with the production of stock materials
for other makers to use. Villages might also generate products for
their local economy, such as surplus produce from small farm operations.
How does this model seem? Nonsense, or maybe feasible?
Eric Hunting
erich...@gmail.com
This is a great document; it's useful the way you organized the questions
and short answers together.
I'm not sure there is any one project that people will support, but it is a
good question to ask how we can support a broad range of efforts if we come
up with better ways to organize and share manufacturing-related information.
That's what's got me fired up in part about Social Semantic Desktop ideas,
but one can use many tools to do this, like Wikis or databases.
I especially liked your analogy: "We seem to basically be in the MITS
Altair, Computer Shack, Computer Faire, Creative Computing, 2600 era of
independent industry. A Hacker era. Remember the early days of the personal
computer?"
It's not exactly the same, obviously, because of the diversity of
possibilities in manufacturing, but it certainly has that flavor.
Also, on: "40 years later, Grandma still isn't writing code and still has
trouble finding the on-switch. What exactly does that mean for these new tools?"
What's important to realize is that often systems are in tiers. There is the
1% dedicated core, the 10% serious users, and the 89% casual users. The
serious users often become the dedicated core, and sometimes casual users
become serious users. So, a system only understandable by the 1% may still
be a very open system which is adapted to all sorts of needs -- like the
person setting up an easy to use GNU/Linux box for their grandma which only
runs email and a web browser. Similarly, one might expect that only some
small percent of the population may want to really get all the details of
fabbing, but they would help everyone around them.
Anyway, this doesn't do justice to your contribution, which I will have to
read in more detail later, but I wanted to make an early reply.
--Paul Fernhout
Eric Hunting wrote:
> In following and participating in the discussions on open
> manufacturing and peer-to-peer organization and as a result of
> exploring the possibilities of local New Mexico Fab Lab development as
> well as my own personal project ideas, I've started to notice
> something. There seem to be a number of re-occurring questions that
> come up -openly or in the back of peoples minds- seeming to represent
> key obstacles or stumbling blocks in the progress of open
> manufacturing or Maker culture. And it seems that they share something
> in common. A 'missing link', if you will, in the mechanisms of
> cultural development. Here are a few of these questions that stand out
> for me;
[...]
> We seem to basically be in the MITS Altair, Computer Shack, Computer
> Faire, Creative Computing, 2600 era of independent industry. A Hacker
> necessary evil for the sake of realizing ubiquity? 40 years later,
> Grandma still isn't writing code and still has trouble finding the on-
> switch. What exactly does that mean for these new tools?
[...]
> Thinking on this, I'm struck by the question of what, then, is the
> functional role of an open manufacturing movement in the context of
> all this? What's the Missing Link?
[...]
Thanks for your post, Eric, it raises excellent questions and delivers a
very good analysis of the situation!
I'm not so sure about the ToolBook answer, however. What brought PCs into
the mainstream wasn't O'Reily, it was the IBM PC with it's modular and open
architecture. So I think Nick's answer might be more to the point:
Nick Taylor wrote:
> As I clumsily attempted to explain to Marc recently, for contagion you don't
> need more documentation, you need less. In fact if you can do it with a
> picture, don't have words at all; if you can do it with a video, skip the
> pictures.
>
> In fact what you really need is lego.
>
> If you look at the inside of a non-laptop PC these days... they don't look a
> hell of a lot different to how they did 15 years ago... and to me they look
> a lot like lego. Everything plugs together. It's RISC applied to physical
> things.
>
> I think the way to increase uptake by vast armies of grandmothers out there
> is to create systems that have reduced-instruction-sets (aka simple building
> blocks) rather than increased documentation.
That leaves the question how a modular, flexible, extensible system of open
building blocks for physical things could look like? What's the best
starting point? Are there already promising initiatives in this direction?
Any ideas?
Christian
--
|-------- Dr. Christian Siefkes --------- chri...@siefkes.net ---------
| Homepage: http://www.siefkes.net/ | Blog: http://www.keimform.de/
| Better Bayesian Analysis: | Peer Production Everywhere:
| http://bart-project.com/ | http://peerconomy.org/wiki/
|------------------------------------------ OpenPGP Key ID: 0x346452D8 --
If the meanings of "true" and "false" were switched, then this sentence
would not be false.
I would say so. I've been doing work on the interoperability and
compatibility of hardware. The plan here is to make it so that each
part has some metadata as to what the ports geometrically allow as
well as what they parametrically allow or are "rated for". Now, this
isn't building blocks like legos, but it allows a variety of unique
shapes to be thrown together, as long as standard interfaces are
respected. This is in part the original motivation for the emails on
recipe representation and interoperability that I've been sending out
in the past.
Here are my answers to the portions I think are missing.
On Fri, Jan 30, 2009 at 1:26 PM, Eric Hunting <erich...@gmail.com> wrote:
> And if the community member is living in a community-owned village
The division/distribution of ownership cannot be glossed over. We
must address this sticky issue to discover how to differ from a
'regular' city.
> then they are sharing some portion their royalties with that local community
> by way of 'rent' relative to their other forms of input to the local economy of
> that village.
We must devise a careful and precise description of how and when/why this
'rent' will be collected, and how those funds flow to the collective
growth or we will likely emulate current 'tax' systems that stifle
growth while centralizing control.
I have incomplete ideas about how to address both of these problems,
but won't go into that unless there are signs that someone besides
myself is interested in such issues.
Sincerely,
Patrick
> That leaves the question how a modular, flexible, extensible system of open
> building blocks for physical things could look like? What's the best
> starting point? Are there already promising initiatives in this direction?
The single biggest barrier to modular design for common platforms is
probably "intellectual property." If it were abolished, there would
be no legal barrier against many small companies producing competing
modular components or accessories for the same platform, or even big
companies producing modular components designed for interoperability
with other companies products.
What's more, with the barrier to such competition removed, there
would be a great deal of competitve advantage from designing one's
product so as to be conducive to production of modular components by
other companies. In a market where the consumer preferred the highest
possible degree of interoperability and cross-compatibility, to
maximize his own freedom to mix 'n' match components, or to maximize
his options for extending the lifetime of the product, a product that
was designed with such consumer behavior in mind would have a leg up
on competing products designed to be incompatible with other
companies' accessories and modules. IOW, products designed to be
easily used with other people's stuff would sell better.
Imagine if
Ford could produce engine blocks that were compatible with GM chasses,
and vice versa;
or if a whole range of small manufacturers could produce competing
spare parts and modular accessories for Ford or GM vehicles;
or such small companies, individually or in networks, could produce
entire competing car designs around the GM or Ford engine block;
or many small assembly plants sprang up to put together automobiles
from engine blocks ordered from Ford or GM, combined with other
components produced by themselves or a wide variety of other small
companies on the Emilia-Romagna networked model.
Under those circumstances, there would be no legal barrier to other
companies producing entire, modularization-friendly design platforms
for use around Ford or GM products, and Ford and GM would find it to
their competitive advantage to facilitate compatibility with such
designs.
--
Kevin Carson
Mutualist Blog: Free Market Anti-Capitalism
http://mutualist.blogspot.com
Studies in Mutualist Political Economy
http://www.mutualist.org/id47.html
Anarchist Organization Theory Project
http://mutualist.blogspot.com/2005/12/studies-in-anarchist-theory-of.html
> The single biggest barrier to modular design for common platforms is
> probably "intellectual property." If it were abolished, there would
> be no legal barrier against many small companies producing competing
> modular components or accessories for the same platform, or even big
> companies producing modular components designed for interoperability
> with other companies products.
The only legal barrier to producing compatible hardware is the patent
system, and patents generally don't cover interfaces, but rather the
operating principles behind the components themselves. So that reduces the
change from "overthrow the government" to the much more feasible "ban
patents on interfaces". Please don't fall into the "intellectual property"
trap because it doesn't exist as such. See
http://www.gnu.org/philosophy/not-ipr.html
> or if a whole range of small manufacturers could produce competing
> spare parts and modular accessories for Ford or GM vehicles;
These already exist, there's even a name for it: "the aftermarket sector".
You can even buy whole aftermarket engine blocks:
http://www.jegs.com/pt/Engine+Block
> What's more, with the barrier to such competition removed, there
> would be a great deal of competitve advantage from designing one's
> product so as to be conducive to production of modular components by
> other companies.
> In a market where the consumer preferred the highest
> possible degree of interoperability and cross-compatibility, to
> maximize his own freedom to mix 'n' match components, or to maximize
> his options for extending the lifetime of the product, a product that
> was designed with such consumer behavior in mind would have a leg up
> on competing products designed to be incompatible with other
> companies' accessories and modules. IOW, products designed to be
> easily used with other people's stuff would sell better.
This may be true over time, but large companies that can afford to develop
new technology desire to lock in the consumer to their brand to extract
maximum profit. So they sell the next big thing encumbered with
incompatible interfaces. And so we are left with dregs like "Memory Stick,
then MagicGate Memory Stick, then MagicGate Memory Stick Duo, then Memory
Stick Pro and Memory Stick Pro Duo and now M2" even though eventually the
market settled on SD and microSD. Now almost all professional-grade Sony
cameras accept SD cards...
http://en.wikipedia.org/wiki/Vendor_lock-in#Sony
But they still make cellphones with incompatible chargers, and people
still buy them because people are idiots. By and large we make decisions
based on price, style, and exclusive "killer app" features first;
compatibility and other feel-good stuff second.
Please don't fall into the "intellectual property"
> trap because it doesn't exist as such. See
> http://www.gnu.org/philosophy/not-ipr.html
I generally put the term "intellectual property" in quotes, precisely
because I do not recognize it as a legitimate form of property. But
I still use the term because, unlike Stallman, I believe there's a
commonality among all the members of the genus that--for me--is more
important than the differences between them. That commonality is that
they're all artificial property titles in the right to perform some
activity, whether it be to arrange letters of the alphabet or musical
notes in a certain order or to configure physical matter in a certain
order to produce some widget. And I believe the economic effects are
similar for all of them.
several years ago I wrote a wiki page that was very similar in content to
eric's message. copypasted here from http://fennetic.net/machines/foo for
your convenience:
I hate writing editorials. I am also very lazy. Therefore, I will
present this article in an outline format. Please read it slowly and
consider each point carefully. There are no layers of filler and is very
rich in content. If you are very motivated and as excited about this
project as I am, please chime in. Warning! Rambling follows, to some
extent. Perpetually under construction, as is the rest of this site.
First, the negatives.
There are lots of people out there repeating the same mistakes over and
over and over because either:
they can't find the information they are looking for because:
few people have distributed the data about their failures/successes
it is buried in a mailing list or newsgroup archive
it is buried under unrelated information on a webpage
they don't know the information is available in the first place because:
lack of cross-referencing and inter-group organization
lack of self promotion
intentionally keeping the info secret or charging a fee to access it
(this means YOU plans-hoarders!!!)
they think nobody has tried to do what they are doing before
The broke-poor home machinist community has no direction, no common
goals, little group cooperation, and not that many achievements under
it's belt. Compare to the Linux developer community.
HF, Homier etc will keep beating away at their slave laborers in
chinese prisons, exporting cheap crap to the rest of the world while
our industrial common sense and infrastructure dwindles
shop classes are being shut down in most school districts, kids never
ever get to see inside a real factory or a traditional crafts shop
(least not where I'm from) old fogeys and hand techniques are looked
down on by the modern mindset
people thing you need a degree to know how to tie your shoelace. "self
taught" means "high school dropout" to most employers.
oil prices are rising and the big companies ain't doin sheeit
gawd I'm starting to sound like a republican
Now, the positives.
You can give a man a fish, or you can teach a man how to fish, or you
can teach five hundred men how to fish. That's what this wiki is for -
giving people fish and teaching lots of people how to fish.
Sick of getting ripped off? Hate cheap stuff that never works right?
Want to invest your time in something educational, and also useful in
its own right?
Making your own tools yields a high quality product (the tool) for
very little money, and it gets easier as you go along. The skills you
learn apply to more than just toolmaking.
Having the right set of tools (foundry, lathe, mill, CAD/CAM) allows
you to make anything you could buy, and you can make it much better
from scratch than what you could afford to buy in the first place. You
can turn crappy items into very nice items with a few deft strokes on
the lathe.
You can afford to have more tools since they cost almost nothing to
make. You can never have enough tools.
When you make your own tools, you aren't afraid to modify them to suit
the task at hand. You can do things that people who are stuck in the
consumer mindset are unable to do. You can turn that old lathe into a
shaper for an odd part, or add a nice little bracket here for your
gizmo, or totally redo half the machine when it doesn't meet your
expectations.
When you have the right set of tools, and aren't afraid of modifying
things, you begin to see everything as a tool. You are no longer
afraid to modify your precious consumer item, since you know how to
make a new one if you need it.
Technical schools teach you how to obey orders, but forget to teach you
how to think independently and be creative. Making your own tools
exercises these long forgotten mental functions, and it is deeply
satisfying.
Creative people who keep their results to themselves are called
crackpots, are greedy, and are boring. Creative people who share their
results with others are known as scientists or artists, and become the
backbone of a community.
[[Dave_Gingery]] only takes you so far in the development of your
machine shop. Many people want plans or at least ideas to help design
more advanced machines. Dave started off with looking at reprints of
old tool catalogs for inspiration for his designs. However, there is no
reason we can't also use modern designs and materials for our
inspiration.
I hope this page can take people who have read Dave's books a bit
deeper into the process of designing, refining, and evolving new and
better machines. This will help them to think critically about how a
design was engineered, what tradeoffs are present in the design, how
they might do it better, etc..
Most plans out there are not open to peer review, and suffer from the
same mistakes every time a person builds to their specifications. This
sad situation could be avoided if the design were open source and the
plans could be modified by the people who acutally execute them.
Any design will have errors, false assumptions, or sub-optimal
comprimises built into it. No design is perfect, regardless what they
say.
The more people to critique and improve the design, the better. If a
person uploaded a solution to a problem in an accessible, centralized
database every time they encountered one, nobody would have to make
that mistake again.
This wiki was created to allow designs to be "open source" so that the
builders and users of the designs could modify, critique, and improve
upon them. By attempting to be a centralized database, I hope that we
will share our mistakes with others so that they don't have to repeat
them.
Look at the Linux developer community - they've shown it's possible to
beat "the man" at his own game. (Add "We should too" to the end of each
subsection)
Untold thousands strong, they have clear ideals and express their
values openly, although not all of their values are the same by a long
shot. They are actively working to manifest those ideals in the world
via software, grassroots organizing, and ethical business practice.
They help each other with little regard to the time or resources
involved, share freely, and actively shun bad behavior. We should too.
the Linux crew have a much more complicated task ahead of them. We
have an advantage, since one man can understand foundry, bench work,
machining, engineering, and electronics, and know most every nook and
cranny of each subject. One person cannot hardly hope to understand
and be up to date on the latest versions of the kernel, X, GCC, and
Gnome, and proficient at coding for all of them. Well, no mere mortal
can at least.
Linux became a success because each programmer shared his results with
the rest of the community. They didn't have to re-write everything
from scratch each time someone wanted an alternative operating system.
Any one mistake was progress for the whole community, since the fix
was included in the next development release. The source code also
served as a centralized repository for information about the
particular project they were working on.
Code forks are not allowed in the kernel, X, GCC, or Gnome. This is
to keep the information and development efforts centralized, so
people don't have to look in ten different places for a fix to their
problem, or make the same mistake each time. Multiple versions of the
same thing are bad for a collective, non-hierarchical effort. This is
especially so when the design is difficult to develop.
Later on in the development of Linux, the developers adopted an
informal set of standards to make information exchange easier. (The
changelog, automake, version numbering schemes, library naming
conventions, distributions) Sometimes they even worked to industry
standards described by standards organizations for the "final product"
(POSIX compliance, X windows architecture, EMC complies with RS274
G-codes.)
The online machining community should adopt a set of standards to
describe what we are doing. We need to give names to common devices
that are being developed. It isn't practical to describe the
construction of a machine every time you mention it. There's a reason
they invented names.
"CNC foam-cutting mill" means any one of twenty different designs.
Some are made out of 2x4's and drawer slides. Some are made from iron
pipe and aluminum castings. Some use routers, some use dremel tools,
some use hot wire cutters.
"Hexapod" is trademarked; "stewart platform" doesn't convey
information about bearings, actuators, feedback electronics, general
configuration, or even intended purpose.
There's no direct way to compare two machines in the same class. Nobody
ever advertises plans for "CNC mill- removes 3 pounds of mild steel a
minute while holding tolerance of .001" but even if they did, such a
performance description is still inadequate since it depends on
workpiece geometry, cutter geometry and materials, overall machining
setup, etc..
Performance is usually judged based on prior experience or a
manufacturer's reputation. There is no prior experience in the hobby
field, and even if there were, it's usually for a new design with no
reputation. (Or a bad one.)
It's not our fault. Industry and government standards organizations
have been ignoring this issue for a long time. Cars have miles per
gallon and top speed ratings described up front. Shouldn't machine
tools have chips per kilowatt-hour and top hogging speed ratings?
Well, you get my point, I hope.
Conclusion: Education should be free. Designs should be good. Standards
should be standard. Hobbies should be fun. Everything should be easy.
[[fenn]]
The lego analogy is apt. Unfortunately, we may not yet have a means
for a universal lego until nanoassembly becomes a viable table-top
fabrication technology and we can treat families of molecules and
atoms as fundamental modular parts. But there should be much more
employ of modular building systems in the Maker community than we
commonly see, which is why my list of proposed books includes a whole
set of these. Current Maker knowledge trends favor electronics and
computers and are a bit deficient in mechanical and structural
systems, perhaps because there is so much less available media for
them. But what's also interesting about the lego analogy is the way
lego designs instruction manuals. Lego is intended for free and
spontaneous play and structures that play through the architecture pre-
encoded in lego blocks (as is the case with most modular systems,
computer components in particular) but when you want a specific model,
there are kits and instruction manuals for these and they have a very
sophisticated design to them that allows them to transcend age and
language. Japanese model kits also demonstrate this. These sorts of
things will be key models for how to compile Maker knowledge in a
structured way, which is why I describe ToolBook media content as
being very visually oriented. So, yes, I completely agree with you on
the cross-cultural multi-lingual potential of a visual information
approach.
I am a little concerned, however, about over-doing the visual focus
because I often think it excludes the visually impaired. But we are in
a highly and increasingly visual culture -though I have worked on
designs for personal computing platforms for the blind to help
compensate for this. I've long been advocating a design called Mozart
for a portable computer based on an iPod-like form factor and WiFi
peripherals that is based entirely on an all-audio user interface
combining the expected speech synthesis with 'audicons',
'soundscapes', music, and other audio interface elements. Physical
interface would be palm-held chord keypad and earphones. The goal of
this design was not only to make a comprehensive computing platform
for the blind but also a practical low-cost portable computer suited
to field scientists and journalists based on the elimination of system
overhead used for display -the most expensive and power-consuming
parts of a portable computer. High performance audio capability
capable, through clever design, of most of the functional capabilities
of visual environments costs, in dollars, hardware space, and power,
an order of magnitude less. And it's a largely unexplored area of user
interface design that, long term, is going to be very important to
'immersive computing' -computing within VR environments. It could have
beat the OLPC to the under-$200 portable computer by a decade, but all
I ever got out of it was a free pop-rivet gun in the NASA/Emhart
design competition -but I digress...
Eric Hunting
erich...@gmail.com
> TOPIC: ToolBook and The Missing Link
> http://groups.google.com/group/openmanufacturing/t/4205d64009a98fce?hl=en
> =
> =
> =
> =
> =
> =
> =
> =
> ======================================================================
>
> == 1 of 2 ==
> Date: Sat, Jan 31 2009 5:01 pm
> From: Nick Taylor
>
>
> Well that was interesting Eric, nice one.
>
>
> My driving thesis - the thing that gets me out of bed in the err...
> "morning", is memetics. The short answer to all of your early
> questions is
> memetics. It's not organised, it's ideas jumping from one head to
> the next
> and deliberately being mutated along the way.
>
> There's a bit of an explosion going on at the moment because Clay
> Shirkey's
> ideas about cognitive-surplus are actually happening, and the
> attention
> economy is specifically friendly to radical memetic-mutation.
>
> And it ain't organised, and probably won't be because of the
> radical-mutation component.
>
> What improved organisation would do though, is assist one of the
> cornerstones (I prefer the word vector, but that's the wrong word) of
> virality - and that is fidelity of transmission. As far as bringing
> it to
> your grandma goes, I think another "vector" is more important, and
> that's
> contagion.
>
> As I clumsily attempted to explain to Marc recently, for contagion
> you don't
> need more documentation, you need less. In fact if you can do it
> with a
> picture, don't have words at all; if you can do it with a video,
> skip the
> pictures.
>
> In fact what you really need is lego.
>
> If you look at the inside of a non-laptop PC these days... they
> don't look a
> hell of a lot different to how they did 15 years ago... and to me
> they look
> a lot like lego. Everything plugs together. It's RISC applied to
> physical
> things.
>
> I think the way to increase uptake by vast armies of grandmothers
> out there
> is to create systems that have reduced-instruction-sets (aka simple
> building
> blocks) rather than increased documentation.
>
But I think the key thing is that its hard to pay attention to the
architecture of the night club when your on the dance floor.
Certainly, dominant social systems filter reality for their own
purposes. Culture is, essentially, a mechanism for encoding human
behavior which, I've long suspected, evolved as both a means of social
cohesion and as a means of non-violent competition between social
groups for geographical resource control. (I've long been of the
opinion that the History = War proposition is idiocy) But I suspect
that most of the time there need be no active attempts at suppression
of cognition. The smartest means of social control are those you don't
have to do a lot to maintain. People caught up in activity at a
certain level have their attentions confined by the parameters of that
activity. So just as an individual termite has no cognition of the
architecture of the larger colony structure it is contributing to the
creation of, we often have a hard time seeing the Big Picture because
we're too deep inside it and too pre-occupied by the activity it
imposes on us to step outside for a broader look. This is why so many
big social and environment problems seem to have an insidious
character to them. They seem to sneak up on us and then suddenly
become a crisis and we fumble around trying to figure out who knew
what and when. why we didn't see it earlier, and craft allegorical
stories to explain them after the fact.
The attention of a society is sort of frog-like. It it isn't moving -
and moving enough to be notices- it doesn't exist. Or as an elderly
venture capital broker once told me when I asked why it was so hard
for a good idea to get investment unless it practically had no need
for it, people only jump on the moving ox-cart. You can have the most
elaborate and beautifully crafted ox-cart in the world, but until it's
actually moving no one can tell which direction it's going. But when
it looks like it's going where they want to go, they will all pile on
no matter how rickety it might look -which explains a lot about
economics in general.
Right now the Maker movement is caught up in and driven by a lot of
excitement. It's the first thing to come along to give society a
positive feeling about the future since the start of the Personal
Computer revolution. And it's very hands-on for most of its
participants. Frankly, most everyone in the MIT-derived Fab Folk group
are exasperatingly, yet also very happily, in over their heads in the
activity they are doing. So they are lucky to catch their breath, let
alone think about the long term or the Big Picture. And that's not
necessarily a bad thing. I think it's wonderful. It makes me envious.
I'd like to be out there surfing too. A novice surfer makes decisions
in a small temporal window. He doesn't have the situational awareness
to see the whole system around him. So he frequently overlooks a lot,
makes mistakes, and gets surprised. And the smart person treats that
as part of the fun. (as long as surprises like tiger sharks remain
relative rare...) That window of situational awareness grows with
experience, and right now no one in the movement has a lot of
experience. They're all learning from each other's examples because
you can't 'teach' and surf at the same time. We need a better
mechanism to record and disseminate the experience so it accumulates
collectively rather than in individuals. A collective experiential
feedback loop. A way to exploit the power of Metcalf's Law. That's
what networks are for and that's why Maker blogs have worked as well
as they have so far. But they convey too much noise in the loop's
signal. They are communicating freely, but not encoding knowledge and
experience well.
As for me, I suppose I'm still on the beach watching things with
binoculars. It's easy for me to see the Big Picture but, like Mycroft
Holmes watching the street scene from a third floor window, I can only
deduce what's going on and deductions aren't facts until tested. Alas,
all too often they never get tested. You can end up just debating the
logic behind the deductions indefinitely. Well, that's why the books
were about Sherlock, after all...
Eric Hunting
erich...@gmail.com
On Feb 1, 2009, at 4:12 AM, openmanufacturing group wrote:
> Hi Eric,
>
> <<
> A 'missing link', if you will, in the mechanisms of cultural
> development.
>>>
>
> questions, and
At the same time, imagine the difficulty in conveying that information
visually. on the other hand, data structures, while not text, are a
way of formatting text and codifying a system of the representation of
types of information, even executional information for running on
hardware (like programs for a computer). So I'm not sure video is the
answer either- imagine if each one of your 14 trillion cells had to
watch millions of videos before they could make their own parts and
components :-). Yikes.
> It would be quite cool to compile a whole Civilisation-Seed of these videos
> - and though assuming that people have the technology is something of a
> leap of faith... maybe it's less of a leap of faith than assuming that they
> have the brains to understand the text version rather than the video
> version. Because I know I struggle with text.
Yes, text is pretty terrible. Video is pretty bad too. I am not sure
that a videotapped civilization seed would be useful though. It would
be better than what we presently have (comparatively little), but
you'll have to convince me.
> It would be an interesting project - I'd be up for any programming that
> needs doing etc.
Really? How about you and I talk sometime- either on the IRC channel
(#hplusroadmap on irc.freenode.net) or over the phone, or over some IM
protocol of your choice. There is a lot of programming work that I can
give you related to these projects- it's kind of why I'm here.
> As a total aside, this civilisation seed reminds me a bit of the occult
> thing - where if you make your own instruments (daggers etc) you're
> supposed to start from zero - mine and smelt your own metal rather than
> buying it, and if you can make your own tools to make better tools etc...
> then so much the better.
What? Do you have some references on this "occult practice"?
> I find this idea quite fascinating - an genotype for an entire civilisation
> - which includes every point in the path from caveman to five minutes
> ago.
What's interesting about biological genomes is that not only do they
encode information, but it's also functional or executional ("function
is structure")- making for very high signal to noise ratio, which we
aren't going to be able to see in text and video simply because these
symbols that we communicate with are hardly 'grounded' to the reality
of the fabrication technologies.
I am not sure that lego-like interoperability and compatibility must
wait until nanotech picks up. For instance, consider molecular
machinery that we're building today- the same problems that apply to
interoperability and compatibility that we experience here in the
macro world applies just the same down there. Problems of growing an
interoperable mechanical playground include forcing too much of an
ontology on other individuals. It's conceivable that you can make an
elaborate specification of different interfaces and things that
certain designs must do in order to meet some pre-defined standards,
but then you're generally supressing bottom-up innovation, maybe you
hold a gun to every designer's head and force them to use your
ontology and interfaces for instance. That's not quite feasible or on
my todo list for that matter. I suspect there is a way to encourage
compatibility, standardization and so on without top-down feartactics
like guns pointed at heads, perhaps through increased information
transfer (open source hardware). A few weeks ago I was flipping
through some documentation on ANSI B16 standards for pipe fittings, as
well as the USB 3.0 specs (pg. 76), wondering if I could quickly
parametrically model the interfaces as a way to encourage the
diffusion of the use of those standards. This way, designers would opt
to incorporate those technologies when it becomes as simple as
drag-and-drop or importing a few extra CAD files that are widely
distributed.
> I am a little concerned, however, about over-doing the visual focus
> because I often think it excludes the visually impaired. But we are in
> a highly and increasingly visual culture -though I have worked on
> designs for personal computing platforms for the blind to help
> compensate for this. I've long been advocating a design called Mozart
> for a portable computer based on an iPod-like form factor and WiFi
> peripherals that is based entirely on an all-audio user interface
> combining the expected speech synthesis with 'audicons',
> 'soundscapes', music, and other audio interface elements. Physical
> interface would be palm-held chord keypad and earphones. The goal of
> this design was not only to make a comprehensive computing platform
> for the blind but also a practical low-cost portable computer suited
> to field scientists and journalists based on the elimination of system
> overhead used for display -the most expensive and power-consuming
> parts of a portable computer. High performance audio capability
> capable, through clever design, of most of the functional capabilities
> of visual environments costs, in dollars, hardware space, and power,
> an order of magnitude less. And it's a largely unexplored area of user
> interface design that, long term, is going to be very important to
> 'immersive computing' -computing within VR environments. It could have
> beat the OLPC to the under-$200 portable computer by a decade, but all
> I ever got out of it was a free pop-rivet gun in the NASA/Emhart
> design competition -but I digress...
I was once working on an interface for the blind once as well. It was
a chorded keyboard but also with an 80+ character plastic mechanical
braille interface. It didn't get very far, but I was going to use some
of the parts I had accumulated from my even earlier days investigating
wearables, mainly in the interest of walking around with a keyboard
with a giant buffer for typing out text all day long without a laptop.
No, human readable text can algorithmically be generated from the
machine-readable code, so nobody never necessarily has to go trolling
through weird markup languages or weird source code for things unheard
of.
> I think humans often need examples and chat and so on to come to grips with
> something. This oral tradition that we're recreating has subtleties to it
> that go beyond the information that's there as face value I think. The
> comments under the "official bit" in the online PHP manual are often more
> use than the official bit for example.
Right.
>> Really? How about you and I talk sometime- either on the IRC channel
>> (#hplusroadmap on irc.freenode.net) or over the phone, or over some IM
>> protocol of your choice. There is a lot of programming work that I can
>> give you related to these projects- it's kind of why I'm here.
>
> Ok - we'll need to synchronise time-zones. It's about midnight for me here,
> which is a little late for anything coherent, although I am largely
> nocturnal.
CST: http://www.timeanddate.com/worldclock/city.html?n=24
>> No, human readable text can algorithmically be generated from the
>> machine-readable code, so nobody never necessarily has to go trolling
>> through weird markup languages or weird source code for things unheard
>> of.
>
> How would you generate this from machine-readable code?
>
> http://www.youtube.com/watch?v=BLfXXRfRIzY
wow, that's 286 seconds of my life i'm not going to get back. This is
pretty much what I think every time I see an instructional video. Besides
the pointing finger of authority and some pictures of a soldering iron,
this 11MB of video conveys a small number of "facts":
- use a 30-40W iron
- "tinning" is putting solder on the iron
- flick the excess tinning solder onto a piece of paper
- twist wires together before soldering
- touch the solder to the wire, not to the iron
- a good joint should be shiny silver
how long did it take to read that list? twenty seconds?
and it's searchable, editable, random-access.
now at just under 300kB, here's some text-based soldering info; there
are a couple diagrams
http://www.ladyada.net/media/common/soldering.pdf
you get tons more info, can skip over parts you dont care about, and
there's even part numbers! I could easily see a document like this being
generated from some computer-frobnable data format.
most of learning how to solder is building muscle memory through hands-on
practice; in this respect videos won't help at all.
Honestly, worst-case scenario? 3D animation. The information on
mechanical movements can be used in tandem with the 3D models of a
generic person as well as the tools. This answer though is a bit of a
hack, and the actual reality of the situation will develop in a more
"right" way.
On Sat, Feb 14, 2009 at 1:20 AM, ben lipkowitz wrote:
> you get tons more info, can skip over parts you dont care about, and
> there's even part numbers! I could easily see a document like this being
> generated from some computer-frobnable data format.
This is also the basis for many generated content as it is: the type
that you see in spam farms on the web (linkfarms, etc.), in
randomly-generated video game content (which sometimes tends to suck,
it's true), and so on.
> most of learning how to solder is building muscle memory through hands-on
> practice; in this respect videos won't help at all.
Maybe that brings up a good point. Actual social diffusion of skills
usually requires one-on-one instruction, and that's where we get the
ancient apprenticeship model that for some reason has been dwindling
in the past few decades. I certainly want to encourage hands-on fine
motor skills and the building of muscle memory, but I'm not sure if
that should occur with computers and the systems we are building- and
in fact I am probably underqualified here since I am unable to recall
any scholarly references on fine motor skill learning via computers.
Has there been any research in that area? Anyone?
On Sat, Feb 14, 2009 at 2:53 AM, Nick Taylor wrote:
> You aren't the audience for these videos. For someone who's never soldered
> anything before, and never seen anyone solder anything before and is
> basically a bit scared of trying, your searchable, editable, random access
> list is utterly useless.
The 3D animation models of the physical manipulation procedure would
be able to help there. Another advantage of having this information
encoded, now, not only is that it can be converted from a very small
dataset into 3D animation for the user to view, but it can also be
used to do fine motor control of machine hardware (theoretically- it
would require me a full demo I suspect to show the idea methinks).
> On Fri, Feb 13, 2009 at 11:16 PM, Nick Taylor wrote:
> Honestly, worst-case scenario? 3D animation. The information on
> mechanical movements can be used in tandem with the 3D models of a
> generic person as well as the tools. This answer though is a bit of a
> hack, and the actual reality of the situation will develop in a more
> "right" way.
like one of these?
http://images2.wikia.nocookie.net/uncyclopedia/images/b/b9/Office2001to2003.PNG
But everyone loves Toolie, the talking hammer!
"it looks like you're making an improvised explosive device. would you
like help?"
> On Sat, Feb 14, 2009 at 1:20 AM, ben lipkowitz wrote:
>> you get tons more info, can skip over parts you dont care about, and
>> there's even part numbers! I could easily see a document like this being
>> generated from some computer-frobnable data format.
>
> This is also the basis for many generated content as it is: the type
> that you see in spam farms on the web (linkfarms, etc.), in
> randomly-generated video game content (which sometimes tends to suck,
> it's true), and so on.
well, there'd be more of a one-to-one mapping of data to prose, not just
random information splatter. perhaps zooming text interfaces could come
into play here (like code folding)
>> most of learning how to solder is building muscle memory through hands-on
>> practice; in this respect videos won't help at all.
>
> Maybe that brings up a good point. Actual social diffusion of skills
> usually requires one-on-one instruction, and that's where we get the
> ancient apprenticeship model that for some reason has been dwindling
> in the past few decades. I certainly want to encourage hands-on fine
> motor skills and the building of muscle memory, but I'm not sure if
> that should occur with computers and the systems we are building- and
> in fact I am probably underqualified here since I am unable to recall
> any scholarly references on fine motor skill learning via computers.
> Has there been any research in that area? Anyone?
flight simulators, surgery simulators... seems like all the haptic
simulation is for medical stuff. it's 2009, where's my teledildo!
i've found that one-on-one instruction can work as well over a computer as
in real life (perhaps better, since the person can't play dumb and get you
to do it for them) but it's much harder to build up a social commitment to
actually finish what you started. also there's the problem of the learner
not having the requisite tools to do the job.
> On Sat, Feb 14, 2009 at 2:53 AM, Nick Taylor wrote:
>> You aren't the audience for these videos. For someone who's never soldered
>> anything before, and never seen anyone solder anything before and is
>> basically a bit scared of trying, your searchable, editable, random access
>> list is utterly useless.
I guess not. My point was that it takes about one second to look up and
down a list, but if i need to go back through a video and search for
something again - perhaps some offhand comment I wasn't sure I heard right
- it can take even more time than the length of the entire video! When
learning something I have to check multiple times to "make sure I'm doing
it right" because of my pitiful human short term memory buffer, so
anything you can do to cut down on the search time helps tremendously. A
table of contents for the video would help. Youtube doesn't differentiate
between "high production quality video" that has been edited for speed and
clarity, and "I was bored so I soldered some wires together on my webcam"
maybe we should abolish this mailing list and switch to youtube videos?
> The 3D animation models of the physical manipulation procedure would
> be able to help there.
Not for general instructional videos. Humans are amazing at inferring
general principles from a single example. It's the specific cases, like
assembly instructions, that we have trouble with, and where 3D animation
would help most.
> Another advantage of having this information encoded, now, not only is
> that it can be converted from a very small dataset into 3D animation for
> the user to view, but it can also be used to do fine motor control of
> machine hardware (theoretically- it would require me a full demo I
> suspect to show the idea methinks).
yep that's the idea.
I don't know how to make zooming text work (at all)* or here, but
zooming instructables have been on the todo list for a while now, or
not 'instructables' per-se but at least zooming packages so that
instructions for making something can be called forward when you go
"wtf is that" while reading over instructions for some new tool you're
making.
* probably some linguistic grammar voodoo magic involving to an author
constructing a sentence and adding more information under the hood
than is explicitly written in the sentence.
>>> most of learning how to solder is building muscle memory through hands-on
>>> practice; in this respect videos won't help at all.
>>
>> Maybe that brings up a good point. Actual social diffusion of skills
>> usually requires one-on-one instruction, and that's where we get the
>> ancient apprenticeship model that for some reason has been dwindling
>> in the past few decades. I certainly want to encourage hands-on fine
>> motor skills and the building of muscle memory, but I'm not sure if
>> that should occur with computers and the systems we are building- and
>> in fact I am probably underqualified here since I am unable to recall
>> any scholarly references on fine motor skill learning via computers.
>> Has there been any research in that area? Anyone?
>
> flight simulators, surgery simulators... seems like all the haptic
> simulation is for medical stuff. it's 2009, where's my teledildo!
I probably sound like I'm contradicting myself. Many months ago I was
talking about Skinner's baby-in-a-box and wondering about social
parenting by computers, and here I am saying now that I'm not sure
fine motor skills can be taught digitally, or even if they should. I'm
so confused, aren't I?
> i've found that one-on-one instruction can work as well over a computer as
Yes, but what about for the acquisition of fine motor skills? It
requires them to figure out what the words actually mean in terms of
physical movements that there are rarely words to describe, simply
because we aren't that pedantic about motion: "conical motion about
the third reference axis to swivel the nth pre-established reference
point". I guess it could work.
> in real life (perhaps better, since the person can't play dumb and get you
> to do it for them) but it's much harder to build up a social commitment to
> actually finish what you started. also there's the problem of the learner
> not having the requisite tools to do the job.
Right.
> maybe we should abolish this mailing list and switch to youtube videos?
Maybe we should abolish this mailing list and generate Second Life
content from the text. (or, preferably not)
The PC was never truly intended by IBM to be a 'personal computer' as
we use the term today. It was long regarded as a peripheral for
mainframes and minicomputers serving a business computing market. A
new kind of 'smart terminal' able to off-load some of the scalar
processing overhead of the larger computers in order to increase
efficiency. Bear in mind, this was a time when most big companies
still thought non-business PCs could never get beyond hobbyist use and
mainstream personal computing would be defined by VideoTex terminals
you plugged into a TV -as were then being developed by the likes of
RCA. Apple was much closer to the contemporary PC concept despite much
less open hardware. (right to the present day...) For a very long time
-well into its emergence as a personal computer- IBM treated the PC as
an experiment. Only with the introduction of the PCjr did they
actually acknowledge the existence of a mainstream consumer computer
market -and you can see in that design the utter nonsense executives
believed about that market. Very little of the knowledge needed to
turn the PC into a mainstream consumer product came from manufacturers
who considered software development their exclusive right. Outside of
the realm of hobbyist hackers, (whom the PC manufacturers themselves
invested heavily in the media demonization of) you couldn't even get a
basic education in computer technology or programming below the
university level. The very idea of a software industry independent of
hardware manufacturers was itself radical. The true breakthroughs came
not with the introduction of any piece of hardware but with the
introduction of key pieces of independently developed software, such
as WordStar, VisiCalc, and Lotus 1-2-3, the games of Infocom and
Sierra On-Line, whose creators initially depended upon a repurposing
of mini-computer software development (even Zork started as software
on PDP-11s) experience and independent exchanges of knowledge often
gathered through a reverse-engineering of the technology. (which
manufacturers of PCs, then unable to understand the emerging nature of
the software industry, often tried to suppress as unwanted competition!)
Modularity of hardware enabled rapid growth of this new industry
through the power of an industrial ecology. But that consumer level
market for the technology was ultimately created by the media culture
it inspired, disseminating the awareness of the technology and the low-
level knowledge of its use, feeding back into the industry through
entrepreneurship cultivated in the end-user community and enabled by
the free availability of knowledge the manufacturers would not,
themselves, disseminate out of the traditional compulsion for monopoly.
You can't learn how to use a CNC, laser cutter, rapid prototyping
machine, etc. from their users manuals alone. You can't learn about
what you can potentially make with them from those manuals either. Fab
Labs were not invented by any of the companies that manufacture any of
their tools. The very term 'fabber' did not originate with any maker
of rapid prototyping machines. These concepts were beyond their
imaginations, as is the Maker movement they observe today with
detached bemusement. Without a media culture to bridge this knowledge
gap and thus create a feedback loop of outsider evolution of the
technology, these new technologies of independent industry can't get
beyond hobby activity. Modularity hierarchically encodes knowledge but
doesn't communicate it. At its high level of sub-system components, a
child can assemble a PC in minutes because, at a certain level of
topological simplicity, they become intuitive in their assembly. But
that child doesn't necessarily learn anything about how the lower
levels of the technology works. They can drill-down into that
knowledge, but not through the artifact itself because it can't
communicate much of its encoded knowledge itseld. Only through the
available media related to components at different levels. In other
words, the knowledge encoded in the hardware can only be decoded
insofar as the artifact exhibits the characteristics of a 'spime'
rather than a 'product', which implies that it exists within a media
information space in parallel to its physical component hierarchy.
Electronics has long had this to a degree because, in order for its
industrial ecology to function, intimate knowledge of component
performance must be circulated in order for any component product at
any level in the food chain to have a market. You can build a
knowledge of electronics out of, literally, sales brochures! This is
not yet the case with these new digital machine tools. Mechanical
engineering never evolved this capability because the products it has
long dealt in never evolved industrial ecologies. We haven't yet
gotten to the car assembled at the dealership. It never evolved beyond
'products' to 'gizmos' as Bruce Sterling would put it. That's why we
have this knowledge gap -much like when the first PCs emerged and the
computer industry's industrial ecology was still nascent and just a
side-show to the ecology of the general electronics industry.
This isn't to say that modularity is not extremely valuable. But we
have to bear in mind that it is a hierarchical encoder of knowledge,
not a communicator of it. It eases end-user manipulation of the
technology at high levels and enables industrial ecologies. (that
aren't necessarily open) But it doesn't teach. In fact, it inherently
insulates the user from lower levels of knowledge if they do not
actively pursue it. Like a graduate from some 'brand name tech
certification' course who can reformat your hard drive or setup a
router but can't actually fix anything or tell you how an IC is made
or how a CPU actually works. Like the PC revolution, I feel the open
industrial revolution must coincide with a mainstream media culture
about it to achieve a mainstream industrial literacy akin to computer
literacy. Otherwise it's just nerds huddled in a dark corner
cultivating a hobby and the Apple iFabber coming 10 years late, too
expensive, and un-openable.
Eric Hunting
erich...@gmail.com
> TOPIC: ToolBook and The Missing Link
> http://groups.google.com/group/openmanufacturing/t/4205d64009a98fce?hl=en
> =
> =
> =
> =
> =
> =
> =
> =
> ======================================================================
> == 1 of 3 ==
> Date: Wed, Feb 4 2009 3:44 am
> From: Christian Siefkes
(quotes clipped for brevity)
> Thanks for your post, Eric, it raises excellent questions and
> delivers a
> very good analysis of the situation!
>
> I'm not so sure about the ToolBook answer, however. What brought PCs
> into
> the mainstream wasn't O'Reily, it was the IBM PC with it's modular
> and open
> architecture. So I think Nick's answer might be more to the point:
>
> Nick Taylor wrote:
>> As I clumsily attempted to explain to Marc recently, for contagion
>> you don't
>> need more documentation, you need less. In fact if you can do it
>> with a
>> picture, don't have words at all; if you can do it with a video,
>> skip the
>> pictures.
>>
>> In fact what you really need is lego.
>>
>> If you look at the inside of a non-laptop PC these days... they
>> don't look a
>> hell of a lot different to how they did 15 years ago... and to me
>> they look
>> a lot like lego. Everything plugs together. It's RISC applied to
>> physical
>> things.
>>
>> I think the way to increase uptake by vast armies of grandmothers
>> out there
>> is to create systems that have reduced-instruction-sets (aka simple
>> building
>> blocks) rather than increased documentation.
>
> That leaves the question how a modular, flexible, extensible system
> of open
> building blocks for physical things could look like? What's the best
> starting point? Are there already promising initiatives in this
> direction?
>
> Any ideas?
> Christian
I would be more than happy to get into a deeper discussion on this to
see if there is interest in actually planning and developing the Vajra
community project, perhaps as a logical physical host to ToolBook. I
just don't want to confuse people by getting them too intermingled.
Eric Hunting
erich...@gmail.com
On Feb 5, 2009, at 4:52 AM, openmanufacturing group wrote:
> =
> =
> =
> =
> =
> =
> =
> =
> ======================================================================
> TOPIC: ToolBook and The Missing Link
> http://groups.google.com/group/openmanufacturing/t/4205d64009a98fce?hl=en
> =
> =
> =
> =
> =
> =
> =
> =
> ======================================================================
> == 3 of 3 ==
> Date: Wed, Feb 4 2009 7:03 am
> From: Patrick Anderson
I think this is a fundamental divide in much of what we've been
talking about. Are you talking about just making learning materials
for people, or are you interested in some industrial functionality
that can also provide that? Ben and I were talking in another thread
recently about ways to do information diffusion, so maybe this common
thread should be discussed there instead (since the points have more
thoroughly been hashed out).
> Labs were not invented by any of the companies that manufacture any of
> their tools. The very term 'fabber' did not originate with any maker
> of rapid prototyping machines. These concepts were beyond their
When fablab first started showing up, it was from the MIT class on
"How to Make (Almost) Anything [except these tools with which you make
stuff]". Oh well. Yes, I can certainly understand how the dynamics of
modularization and community allow systems to be taken into new and
unheard of directions, but we're already talking about stuff like
that.
Electronics datasheets from sales brochures is a very good point. And
in fact, digital media now allows us to have that information in a
computational format that lets you render it on screen, or in a video,
or in a song (ok, computer generated songs are pretty bad in my
opinion), and so on. So we're practically talking about the same
thing- but in most cases, when we talk about media, what I imagine is
something catastrophically terrible happening, like people drawing
over electronic datasheets and this information not being capturable,
and being lost in the same way that the current Maker culture's work
is being lost in the signal-to-noise of markup and natural language as
a poor medium of the communication of design and tool information.
> This isn't to say that modularity is not extremely valuable. But we
> have to bear in mind that it is a hierarchical encoder of knowledge,
> not a communicator of it. It eases end-user manipulation of the
> technology at high levels and enables industrial ecologies. (that
> aren't necessarily open) But it doesn't teach. In fact, it inherently
> insulates the user from lower levels of knowledge if they do not
> actively pursue it. Like a graduate from some 'brand name tech
> certification' course who can reformat your hard drive or setup a
> router but can't actually fix anything or tell you how an IC is made
> or how a CPU actually works. Like the PC revolution, I feel the open
> industrial revolution must coincide with a mainstream media culture
> about it to achieve a mainstream industrial literacy akin to computer
> literacy. Otherwise it's just nerds huddled in a dark corner
> cultivating a hobby and the Apple iFabber coming 10 years late, too
> expensive, and un-openable.
Originally in your email you were talking about electronics and the
datasheet media from sales brochures as being useful throughout the
industry, but then you mention here about the low level tech support
people who know how to format a hard drive, but don't know how
anything about ICs- so it seems like the problem is pervasive, even
give some of the media that currently exists today. To me, this is an
indicator that something new has to be developed, something easy and
extensible, but something right and fit for our day and age. (Unless
you can elaborate on this, maybe that discrepancy that I point out
doesn't actually exist.)
Not at all, you're quite welcome here- nice of you to drop in. I hope
you stay for a while. What I am starting to see is that overall these
projects are of the type where there's a core group in it for the long
haul- you seem to have been at it for at least 15 years- and then
there seem to be others, perhaps funneled in by the magazine culture,
that get on at an entry point, and get off the ride at some
pre-defined exit point, for some capability or specific skill that
they were interested in. From the looks of it, we have no problems
finding people that are in it for the long journey, which is
increasingly hard for me to define in terms of industrial capability
and growth, but nevertheless I am confident that this group is it.
Have you read some of the archives, maybe we should talk sometime?
> I just stumbled into this thread. I'd like to introduce myself to the
> group by way of commenting on the discussion, interesting to me
> because I have been struggling and learning as a participant in this
> movement for the last 15 or so years. I resonate to the breadth and
> multiple levels of themes raised by Eric's initial questions, even
> though I can probably only comment on a few of them. And, I second
> Paul's emphasis that the diversity of manufacturing interests and
> possibilities very much complicates the conceptualization of what I
> take it is meant here by open manufacturing. It's also the case that I
> aspire to produce at least one version of something like an 'Apple' of
> digital fabrication and will try to avoid shameless promotion.
I have to admit that while ShopBot has been around for a while, I've
unfortunately not been very in tune with what ShopBot's interests are
when it comes either to the Apple of open manufacturing, or to
encouraging this community, open source hardware packaging
(datasheets, schematics, etc.), or its relation to fablabs providing
one of the most expensive tools. I saw a tent over at the Austin Maker
Faire 2008 but didn't get a chance to say hello.
> While appreciating the issues of culture, communication, and education
> that are a part of the theme, my first reaction was a more pragmatic
> question: Is the concern with the design, manufacture, and making of
> things, or with creating the tools that do the production. At what
> point should the focus be on the tools; at what point do we take some
> tools that work adequately and change the focus to the new kinds of
> production and distributed manufacturing that we would like to
> accomplish with them?
And is distributed manufacturing the final destination? And what does
it mean exactly? Sometimes I have heard of people using it to mean
business-to-business supply chain management so that one part is
manufactured in one location, and another two parts are made in
another location, i.e. off-loading to the grid. And then, more
commonly I have encountered it to mean compressing as much capacity
into a single location as possible, which may or may not be on that
grid of distributed jobs (somewhat like emachineshops). As for the
tools that do the production, that's industrial automation :-) which I
would very much like to see go open source and participate in the
design, manufacture and support thereof.
> My interest is in both. I started off very much focused on the tools
> -- having myself stumbled into the amazing capabilities of digital
> fabrication and becoming fascinated with finding ways to turn big,
> expensive, industrial CNCs into something that people can use. The
Looking over the shopbot website, I'm not seeing those big CNC
machines (not that this is a bad thing- there's some elegance in size
reduction of course). Just wondering if there's any five, six-axis
stuff that I'm missing out on.
> first small ShopBot that I showed off to a local woodworking club 15
> years ago was made almost entirely with "off the hardware store shelf"
> parts. And, with respect to their structural components, the tools
> could cut themselves out. They were self-replicating -- sort of. Our
Did you get a chance to look over some of the posts on self-replication?
http://groups.google.com/group/openmanufacturing/msg/e4c375acce772250
> plan was just to make plans and a parts list available to people (back
> in the Compuserve group days) and help them make their own tools, from
> which they would help their neighbors make tools. I naively/
> romantically thought that as the capabilities of the tools became
> known to people there would be an explosion of activity building them
> and then starting to make all sorts of things with them.
So, as you mentioned, you got people knocking on your door to get a
fully assembled machine for whatever their jobs were. It's interesting
that this happened, rather than something like a fablab or techshop or
menshed at the time. I wonder if that can be changed? For instance,
one topic that comes up is the idea of having package (or fabber
media) maintainers that would work to make sure parts and machines are
interoperable, that they have their standards properly written up,
designs and schematics put together, and the dependency information-
such as what tools are required to make this tool/artifact- so that
this information can at least remain in circulation rather than dying
off each time somebody "gets off the train" at some certain point when
they decide they are done with it all and just want to be happy making
signs or whatever.
> Well, the first reality check was the self-replication. We wanted to
> get to 4'x8' so that standard sheet goods could be cut and machined
> (boats were my interest) and it turned out that cutting-yourself-out-
> of-plywood did not scale. We were just never able to get enough
> rigidity. So, it was back to the hardware store for the structural
> components too. Utility strut became our modular material for
> structural elements (a large scale 'Erector Set') and it turned out we
> were able to use the strut edges as the rails for motion components,
> with patio door rollers as the bearings. And, the tools worked fairly
> well.
Re: rigidity, have you looked into hexapods and Stewart platforms?
http://fennetic.net/machines/?hexapod
> But the real problem with the original DIY idea was that nobody was
> interested. That's not to say that there were not a few who wanted to
> build their own CNCs back then, but many wanted a smaller tool for
> metal work, or wanted to build a plasma CNC, or they wanted a bigger
> tool, or didn't like our design. It turns out that to make something
> like a CNC tool affordable in the absence of available, volume
> produced, modular components it needs to be vertically integrated. Our
> parts worked well together for our tool and did serve as the basis for
> modification and development, but they were not a good starting point
> for people who wanted to build very different types of tools. There
> were too many different needs, priorities, and ideas for much of
> anything to be standardized -- probably still the case.
Hm, I would think that it would be the case that the same tools that
you were using to make your original ShopBot would be the same tools
that would be needed to make the modifications or different designs.
It's true that it doesn't scale as the number of designs and number of
people wanting to make custom components increases, but that would
just be a sign of it becoming popular and in demand. As for modular
components, that's kind of the idea of "kits" or letting people get in
on at the entry points that they want, and off when they want- or for
some of us, we're never going to want off. And it's interesting that
you mention that the ShopBot customers weren't interested in kits- you
had to basically just give them a drop-in machine. I guess it's still
somewhat a kit, except the assemly is being done on your end. Not a
big deal I guess, but it does hinder the progress on any type of
modifier kits that might have been otherwise produced to replace
specific components. Have you had any modders yet make kits for
retrofitting ShopBots to do new things? Is it always custom
modification requiring specialized knowledge?
> The surprise was that there were some people who wanted to make things
> and they started banging on our door. They were cabinetmakers and
> signmakers. They had already been sold on the concept of CNC by the
> big industrial guys. While they had shops that could benefit from
> digital production, they could not afford an industrial tool and were
> not comfortable with investment that was so large they would have to
> organize their business around a machine. They wanted a tool oriented
> to a small shop and to use by a couple of people (as in the PC vs the
> mainframe). Most important, they said they didn't want a kit and they
> didn't get off on building tools. They wanted something they could
> just put to work. That's the business we got involved in. These were
> not the new enthusiasts/adopters for digital fabrication that we were
> expecting, but they wanted tools and did hold our feet to the ground
> with respect to tools that were ready to go, that had solid
> performance, that were easy to use and maintain, and that were
> oriented to individuals.
Besides the issues I've brought up above, would ShopBot be interested
in expanding past that? :-)
> However, I take from this thread that there is a disappointment with
> the absence of a more dynamic and supportive culture for the open
> manufacturing that digital technology makes possible. I couldn't agree
> more, though for me the more specific question is why distributed
> manufacturing using digital fabrication is not more widespread.
> ShopBot has had some success in the woodworking niche, but the tools
> we sell into this market are primarily using CNC to do better or
> faster what was already being done in these shops. Our tools are very
> general purpose and will cut and machine a wide range of materials,
> but they are rarely used for new production purposes.
What again do you mean by distributed though? Different parts made at
different physical locations? Simply a proliferation of manufacturing
facilities that do stuff in-house? Job allocation algorithms exist
that transform a manufacturing house job into separate parts for work
at different areas- much like compiling software to run on a cluster-
so maybe it's something else, more culture-oriented as the reason that
it hasn't exploded yet.
> There does seem to be relatively widespread recognition that
> eventually there will be affordable 3D printers that can make big
> things out of real materials -- before or after the nanoparticles are
> actively participating -- those StarTrek replicators. But in real
> contrast, there seems little recognition of just how much can be done
> with current, affordable, 'subtractive' 3D printers as represented by
> various CNC milling, laser, vinyl, and plasma cutters.
Yes. Many people are sitting around on top of Drexler's books waiting
for molecular nanotechnology to make all of their dreams come true.
The capabilities that we already have are mostly hidden and obscure,
proprietary information, catalogs are poor and rarely in databases (as
far as I can tell); this situation needs to change.
> One of the basic goals of the Fab Labs is to simply make this point.
> To show how anybody can make almost anything with today's tools for
> digital fabrication. In fact, they evolved out of Gershenfeld teaching
I know how different TechShop and fablabs are, but I still see a
common theme that the majority of their parts and tools are not made
by them. In particular, I got an email from Jim Newton a few weeks ago
when he was asking for another round of funding, something like $600k
plus $2 mil. I was confused by this, and so I called Jim up and asked
him how is it exactly that TechShop is a place where you can make
nearly anything, except apparently its own tools? Now that I've met up
with Les Filip and some of the others from what was going to be the
Austin TechShop, I'm starting to get a feel for what was going on.
> a class on this topic to MIT students. We really encourage them in
> this mission because communicating the possibilities of digital fab is
> part of the important consciousness raising that needs to occur in the
> transition to new production methods. I might naturally personally
> prefer the Fab Labs put more emphasis on their ShopBots which can do
> many of the things the 4-5 other types of CNC tools are doing -- but
> it's hard to beat a quick laser cut-out and not having to worry about
> cutter-compensation and bit selection. Because Fab Labs have taken on
> a relatively ambitious communication and education role, I don't
> imagine the official MIT version will be the right 'fab center' for
> every neighborhood but I do see them seeding a lot of interest. As
> just one example of recent developments, the TechShop franchise,
> should it catch on, will have a good complement of digital fab tools.
Originally I was easily confused into believing that a fablab was a
shop where they were trying to compress as much fabricational capacity
as technically possible into. But the more I look at it, the more it
seems like an educational initiative, and that's how they are getting
grants and funding (which isn't a terrible idea)- more like community
outreach and funding.
> [Just an aside with respect to the earlier question of why Fab Labs do
> not self-replicate their own tools: On more than one occasion
> Gershenfeld has made the comment to me that he hopes to put us out of
> business in 5 years. He has actually shown me both motor drivers and
> stepper motors that were fabbed by a Fab Lab (these are core
> components for any robotic tool) -- so that's one they are working on,
> though I figure we have closer to 10-15 years. At the same time, the
> Fab Lab group has done pretty extensive work standardizing the design
> and tool-control process as well as developing systems for sharing
> methods and designs. They are hoping to make the design and production
> process the same for all the types of tools they are using. I'm not
> entirely comfortable with the approach, but their investment in the
> work seems consistent with the theme of the thread and perhaps has not
> been publicized enough.]
Huh, yeah, I haven't heard anything about this. I've been working with
fenn and a few others on a system for the interoperability,
compatibility and automation generation of assembly instructions (or
building instructions) for tools, so that we can determine what
exactly needs to be made first, and what tool depends on what other
tool, etc. etc. And if fablab is actually going forward with this,
hopefully we can offer something of value.
> My own guess about what is needed to spark the next level of digital
> fab is that it is content, in particular content that derives from
> design software that makes it possible to produce a design that can be
> easily parametrically sized or modified or customized by almost anyone
> and then straightforwardly output to a subtractive 3D printer. We also
> need a process that encourages these digital designs and makes them
> generally available, as well as a system for linking to networks of
> 'printers'. To be compelling, the designs need to take advantage of
> the capabilities of the digital tools, using them for new approaches
> to assembly and manufacturing. There are some niche models, but I'm
> not sure anyone has hit on this yet …
Yes, exactly. The open source hardware packaging ideas that I've been
spewing about is exactly on that line of thought. I don't know if
you're familiar with debian dot deb packaging formats, but that's how
software does it in the linux world. I'll have to assemble some links
and materials once I get my server back up (it's down at the moment,
sorry Paul- I know Paul noticed this earlier today).