AT&T announced a new “sponsored data” scheme that lets content creators buy their way around the company’s data caps. It’s bad news for everyone—but AT&T.

Yesterday AT&T announced a new “sponsored data” scheme, offering content creators a way to buy their way around the data caps that AT&T imposes on its subscribers.  Although AT&T touted it as a “win-win for customers and businesses”, it is actually just a win for AT&T.  This plan is a tremendous loss for everyone else.

Going to the Heart of Net Neutrality

While people sometimes get lost in the details, at its core net neutrality is a pretty straightforward concept: it is the principle that the company that connects you to the internet does not get to control what you do on the internet.With its sponsored data scheme, AT&T is proposing to do just that.  AT&T’s relationship with a website, app, or service will control the way that AT&T’s subscribers interact with the internet.

AT&T is imposing its own tax on anyone who wants to connect to its millions of subscribers.  Of course, this tax is only attractive to content creators if AT&T’s normal service is too shabby to deliver their content without it.  That gives AT&T a big incentive to keep data caps low and overage fees high.

Who gets to innovate in a world where you need to pay AT&T to compete?  The answer? Established services that can afford to pass the fee onto customers.  That’s why net neutrality isn’t really about Netflix or Facebook.  They may not like it, but when push comes to shove they will probably pay AT&T’s tax and pass it along to their customers.  But the next Netflix or the next Facebook won’t be able to afford to do that at the start.  Startups will abandon any potentially “data-intensive” innovation to big players—a recipe for stagnation.  As more established services move to AT&T’s special lane, caps covering everything else may go down bringing the definition of data-intensive down with it.

AT&T is Creating Artificial Scarcity Just to Monetize It

AT&T has proudly moved past the days when the iPhone crashed its network for millions of excited subscribers.  In May of last year CEO Randall Stephenson told investors that AT&T anticipated reducing expenditures on its network and that data caps were really about charging content providers He repeated his confidence in AT&T’s network in December.

The sponsored data plan itself further highlights AT&T’s confidence in its network: if the network truly was fragile AT&T probably would not be inviting creators to dump a lot more content onto it.  Any problems in the network that exist going forward should be traced back to the fact that AT&T is investing in its special paid access lanes instead of the parts of the network available to everyone else. 

“The FCC’s strategy of closing its eyes, putting its fingers in its ears, humming, and pretending that data caps don’t exist needs to stop.”

Furthermore, even if AT&T is painting an overly rosy picture to investors and deluding itself about its network capacity, monthly data caps are an incredibly inefficient way to deal with momentary network congestion.

But they are a great way to gouge content creators.

Special Problems for Rural and Low-Income Communities

For many people wireless data caps are a frustration, but one they can work around.  Wired internet connections at home and at work give them plenty of uncapped internet access when they need it.  But for many others, wireless is their only way to connect to the internet.

In rural areas, 4G wireless connections are pitched as replacements for decommissioned wired internet connections, or as ways to affordably bring high-speed internet to hard-to-reach areas.  For these areas, capped wireless internet becomes their only option.

Similarly, many households in low-income communities do not have the luxury of purchasing two internet connections.  As a result, many of them are mobile-only internet users.

Both of these communities are under threat of being trapped in a second tier internet, pushed only to sites and services that can afford to pay their way around AT&T’s paywall.  For them, the “real” internet will start to leave them behind.

Where is the FCC?

In its Open Internet Order, the FCC detailed the problems that the sort of two-sided market proposed by AT&T creates such as inefficiently high fees, provider arms races, and overall reduced innovation.  Unfortunately, in the past two years the FCC has done nothing to understand the role that data caps can play in creating just such a market.  Even the FCC’s own Open Internet Advisory Committee admitted that it didn’t have enough information to begin explaining how data caps fit into the Order.

The FCC’s strategy of closing its eyes, putting its fingers in its ears, humming, and pretending that data caps don’t exist needs to stop.  AT&T, along with other ISPs, will continue to test the boundaries of the FCC’s stomach to protect an open internet.  If the FCC doesn’t push back then AT&T, like boundary pushers everywhere, will have no reason to stop.

And that’s a loss for everyone besides AT&T.

Afinia pushes back in a lawsuit that could impact the entire desktop 3D printing industry.


In November, the 3D printing company Stratasys suedanother 3D printing company Microboards Technology, LLC (which distributes the popular Afinia desktop 3D printer).  Beyond being the second example of one of the older commercial/industrial-oriented 3D printing companies suing the new breed of desktop competitors, this lawsuit was interesting because held out the potential that just about every desktop 3D printer infringed on Stratasys’ patents.  If this turned out to be true, it would be a big deal. 

When the original complaint was filed we summarized it here.  Today, this post will summarize Afinia’s response to that complaint.  

As with Stratasys’ complaint, it is important to include a quick caveat when thinking about Afinia’s response.  Afinia’s response is written by Afinia’s attorneys and is designed to present all of the facts in a light as beneficial to Afinia as possible.  That doesn’t mean that there is anything untrue in the response, but it does mean that any claims made in the response should be taken with a grain of salt.  That being said, as with the earlier analysis of the Stratasys complaint, in an attempt to explain what is going on this post will more or less take Afinia’s claims at face value.

Afinia’s Response

The short version is that Afinia challenges the validity of all four of the patents that Stratasys accuses them of infringing.  This goes beyond simply asking the court to decide that Afinia does not infringe on any of the Stratasys patents (although Afinia does do that).  Afinia is also asking the court to find Stratasys’ patents totally invalid, which would prevent Stratasys from asserting them against anyone else.  In light of Afinia’s belief that Stratasys’ patents are invalid, Afinia also accuses Stratasys of using this lawsuit and its patents to try and monopolize the 3D printing market.

    The ‘925 Patent

For a quick overview of the ‘925 patent and the rest of the patents at issue, take a look at thesummary of Stratasys’ complaint.  The ‘925 patent allegedly covers controlling the infill of an object being printed.  To understand how Afinia is challenging the patent, it is important to understand a bit of how the patent application process works.

In many ways, a patent application is more like the opening position in a negotiation.  The applicant submits an application.  The patent examiner at the Patent and Trademark Office examines it and searches for prior art, which would show that the proposed invention is not actually new (and therefore not patentable).  If the examiner finds prior art, the applicant often revises the application to distinguish it from the prior art and/or explains to the examiner why the existing application is different from the prior art.  Some of this back-and-forth (although not enough of it) is done in writing.

Afinia looked into this written record to find evidence of ways that Stratasys narrowed the scope of its patent in order to get it accepted.  Afinia asserts that this record is evidence that Stratasys narrowed the ‘925 patent to the point where it no longer includes what Afinia is doing.  This narrowing was especially important to distinguish the application from existing patents5,490,962 and 5,518,680.  Afinia especially highlights the ‘680  patent.  Additionally, Afinia points to patent 5,121,329 as being prior art including what Stratasys is claiming in its ‘925 patent.

Afinia ends its response to the ‘925 patent infringement claim by accusing Stratasys of asserting this patent in an attempt to monopolize the 3D printing market.  This final claim is included in Afinia’s response to all four patents.

    The ‘058 Patent

The ‘058 patent allegedly covers the heated build platform that is used by most desktop 3D printers.  In its response, Afinia includes the high degree of detail included in Stratasys’ patent.  As with the ‘925 patent, much of this detail is presumably included in order to distinguish this patent application from prior applications.  Every detail narrows the scope of what is protected by the patent.

Afinia then explains that elements in the ‘058 patent, such as a referenced “control system,” do not exist in its printer.  Additionally, older patents such as the ‘680 patent referenced above already included a discussion of heated build platforms.  Essentially, Afinia is asserting that all of the elements of the heated build platform that it uses predate the ‘058 patent.  The also accuse Stratasys of knowing about patent 5,501,824 which includes references to heated build platforms and not disclosing it to the patent examiner.  This is generally frowned upon.

    The ‘124 Patent

This patent allegedly covers Afinia’s extruder.  Here, Afinia looks at the patent and determines that it requires a tube with a wall between 0.005 and 0.015 inches thick that passes through a heating element.  Since Afinia uses a tube that is four times thicker than that requirement, Afinia asserts that the patent simply does not apply to them.  

It is worth noting that the 0.005-0.015 sizing requirement appears to be derived from interpreting a combination of elements in the patent, and does not explicitly appear in the patent.  This does not necessarily mean that a court would or would not agree that the patent includes the requirement, but the ambiguity is worth flagging.

    The ‘239 Patent

Finally, the ‘239 patent allegedly covers a process for concealing the seams that can appear because of the layering technique used in 3D printing.  

Afinia again looks to Stratasys’ attempts to distinguish its patent from prior art.  Armed with this record, Afinia asserts that the technique used by its printer and software lacks some of the elements required in the ‘239 patent.  Specifically, the lack of start and stop points allegedly keeps it outside of the scope of the patent.

Afinia’s Counterclaims

In addition to responding to Stratasys’s claims, Afinia makes some claims of its own.  The first is to request that the court declare that Afinia is not infringing on any of Stratasys’ patents.  Afinia requests that this declaration – called a “declaratory judgment” – apply both to Stratasys’ claims of direct patent infringement and the allegations of secondary infringement.

The second request is likely more interesting to the broader 3D printing industry and community.  Afinia asks the court to issue a declaratory judgment that the patents themselves are invalid.  This would go beyond merely finding that Afinia does not infringe on these patents.  Instead, it would mean that the patents themselves are invalid.  If granted, invalidating the patents would prevent Stratasys from accusing anyone of infringing upon them.

What Do We Know Now?

The most important thing that we know now is that Afinia is taking steps to really fight this lawsuit.  As mentioned in the summary of Stratasys’ complaint, this lawsuit is important well beyond the fate of Afinia.  If Stratasys is correct that it has patents that cover most desktop 3D printers, they will have significant control over the fate of the industry.  Since Stratasys purchased desktop 3D printer manufacturer Makerbot last year, this type of control could make it much harder to launch new printers.

Afinia hired a law firm to respond to this complaint, and clearly bankrolled a reasonably expansive search for prior art.  This is a strong alternative to immediately settling with Stratasys.  It is the kind of thing that a company does when it decides to fight a lawsuit instead of settle, and is encouraging.

We also now know where Afinia is looking for prior art.  This should help the larger 3D printing industry and community think about how to address these patents.  The more eyes on the patents put forward by Afinia, the better the chances are of finding additional prior art.  If you are familiar with the operations of 3D printers, take a look at the patents linked to above.  And also look at Afinia’s response, especially paragraphs 82-84, 87, 90, 95, 97, 109, and 124.

What Do We Not Know Now?

Unfortunately, we have no idea how long Afinia will be willing or able to fight this lawsuit.  Patent infringement lawsuits are very expensive propositions, and the option to settle is always there.  There is nothing to suggest that Afinia has any plans not to see this thing through to the end, but the truth is that you never know.

Perhaps more importantly, we also do not know how strong Afinia’s legal arguments are.  When we only had the Stratasys complaint, it looked like Stratasys had four patents that covered the Afinia printer.  Now we have Afinia’s response that makes it look like none of Stratasys’ patents cover the Afinia printer (or are even valid).  Going forward, the arguments of both sides will be tested.  Hopefully that process will give everyone a better sense of the patent landscape.

Finally, we still don’t know what impact this lawsuit will have on the larger 3D printing industry and community.  We (or at least I) don’t know if Stratasys is reaching out to other companies in an attempt to get them to license this patent portfolio, or is Stratasys would even be interesting in licensing this portfolio to anyone else if they win in court.  If other companies are getting those inquiries, we don’t know how those other companies are responding.  The mere existence of this lawsuit may be enough to chill development, but development may simply be churning on in spite of it.  As frustrating as it is, it will simply take time to figure it all out.  

Of course, I’m headed to CES next week where it looks like there will be lots of 3D printers on display.  Kickstarter is chock-a-block full of new printers.  And amazing work is going on in the open source development community every day.  So the good news is that this lawsuit hasn’t stopped everything.  At least not yet.

Image: Flickr user jabella.

3D printing was left out of the Undetectable Firearms Act, but the discussion about 3D printed guns raises broader concerns about how well lawmakers understand making things at home.

Yesterday the Senate passed an extension of theUndetectable Firearms Act.  While this act is mostly about what its name suggests – undetectable firearms – discussion about the bill has managed to bring in 3D printing.  As an organization, Public Knowledge takes no positions on gun policy.  Therefore we would not normally have anything to say about a gun-related bill.  But Public Knowledge is involved in 3D printing policy.  With the passage of this extension in both the Senate and last week in the House, now is a good time to explain what has lead us to this point, what is happening now, and how it all impacts 3D printing.

What is this Undetectable Firearms Act?

The Undetectable Firearms Act is nothing new.  It was first enacted in the late 1980s and has been reauthorized at fairly regular intervals ever since (the Act has a sunset clause built into it that schedules the Act to expire after a fixed amount of time if Congress does not reauthorize it).  It was apparently originally motivated by concerns about all-plastic weapons that could pass through a metal detector (like at an airport) undetected by security.

The most recent version of the Act just so happened to be scheduled to expire yesterday.

Enter 3D Printing Guns

When stories about 3D printed guns first started popping up, many lawmakers felt compelled to respond.  Those responses generally took two forms.  One form was a concern that 3D printing could allow people to make guns at home.  Once it became clear that people have been making guns at home for decades, and that existing gun regulation already accommodated this fact, this concern started to die down.

The other form was a concern about the undetectability of 3D printed firearms. While there aremetal 3D printed guns, many of the initial stories were about guns that were made with plastic.  These 3D printed plastic guns, like any kind of plastic guns, were potentially undetectable by traditional metal detectors.  Thus, they potentially raised the same security concerns that all plastic guns raise.

Representative Steve Israel was one of the first lawmakers to respond.  His initial bill, which was designed to extend both the duration and scope of the Undetectable Firearms Act, focused heavily on 3D printing.  Because of what we viewed as an unnecessary focus on 3D printing, Public Knowledge expressed concern about the bill.

Fortunately, unlike other lawmakers who seem dedicated to a legislate first, think second approach, Rep. Israel was encouragingly receptive to our concerns.  

Our primary request was that Rep. Israel focus the bill on the undetectability of firearms.  Press coverage aside, the method of manufacture of an undetectable firearm should not particularly matter for undetectable firearm regulation.  If a gun gets smuggled through airport security, it does not really matter how it was made.  What matters is that it was undetectable, not that it was manufactured on a 3D printer, or a CNC machine, or with an injection molder.   Public Knowledge takes no position on firearms regulation, so our only request is that any potential firearms regulation not unnecessarily undermine 3D printing.

Encouragingly, the revised version of Rep. Israel’s bill spends a lot less time talking about 3D printing and a lot more time talking about undetectable firearms.  In addition to extending the prohibition on undetectable firearms, it focused primarily on including a requirement that any firearm feature at least one functional metal part (as opposed to simply include some detectable metal somewhere in the design).  In the Senate, Senator Schumer has also been responsive to our concerns about the 3D printing element of this legislation.

The Most Recent Bills

With the expiration looming, last week the House moved to extend the term of the Undetectable Firearms Act.   Instead of voting on Rep. Israel’s bill, it focused on a less contentious “clean” bill that simply extended the current law without substantively changing it.  In good news for 3D printers, the current law does not include any references to 3D printing.

Yesterday, the Senate passed an identical “clean” extension.  With passage in the House and Senate, President Obama signed the bill last night and extended the duration of the current, 3D printing free, version of the Undetectable Firearms Act.

Going Forward

With the extensions passed, the deadline pressure to pass a bill that expands the scope of the Act in addition to extending it has lightened somewhat.  As such, it is unclear what will happen to the Israel bill in the House or the Schumer bill in the Senate.  That being said, Public Knowledge does look forward to continuing to work with both Rep. Israel and Sen. Schumer to avoid unnecessarily drawing 3D printing into any gun legislation.

Perhaps more importantly, this entire episode highlights just how disconnected we as a society are from manufacturing and making things.  3D printing is an exciting technology that empowers people to make all sorts of things (not necessarily the “everything” that some lawmakers imagine, but plenty of things nonetheless).  But it is not the only way that people can make things.  Although it can often feel this way, not every object or device comes fully formed from a factory.  People have had tools to make things on their own since the dawn of mankind, and many still have those tools at home today.  

Many lawmakers responded to 3D printing as if it was the first time they had considered that it is possible for people to make anything on their own at home.  Much of the discussion around 3D printed guns is a symptom of that shortsightedness.  Unfortunately, it is unlikely to be the only one.

It is true that 3D printing will have new legal and policy implications that need to be addressed by lawmakers.  But concern about simply “making things at home” cannot become a justification to cripple 3D printing – or any other maker technology – with regulation.

image: Thingiverse user HaveBlue.

Last week at our annual IP3 Awards, we debuted a new addition to the Public Knowledge family: the GIFerator.  Basically, it was a photobooth that let attendees make their own animated GIFs and publish them to the internet.  We designed the GIFerator with openness in mind and on top of open technologies, so this blog post is intended to share our process and document it well enough for you to set up your own. 


At its heart, the GIFerator is fairly straightforward.  We needed a way to capture pictures, stich them together into animated GIFs, and then publish them on the internet.  We also needed an interface that was easy to use, explained the process to people, and was robust enough to keep working unattended through an event with an open bar.   This allowed people to walk up to the GIFerator, mash a few buttons, and publish their GIF to the internet.

Bill of Materials

1 computer with a webcam and internet connection (it does not need to be a new one – we used the oldest iMac in our office and it worked great)

1 USB keyboard

1 Box (here is our design)

4 Buttons (we used these concave arcade-style buttons from Sparkfun because they were fun to mash and only cost $1.95 each, but pretty much anything should work)

Wire and Solder

Firefox plugin

The Software

The genesis of this project can be found at Eyebeam.  We were up in NYC to meet with Eyebeam in preparation for our artist in residency program.  Eyebeam happened to be running a show called Open(Art) they had put together with Mozilla.  One of the projects on display was a demonstration of Meemoo, an application built by Forrest Oliphant that created stop motion animation gifs.  With the IP3 awards coming up, we thought that we might be able to integrate it into the event.

Meemo is an HTML5 data programming environment that does not require programming expertise.  Instead, it is a graphical environment where each action is represented by a box.  These boxes are connected by “wires” that let information flow between them.

Building off the existing webcam to animated GIF demo we started to put together our project.  I encourage you to play around with it on your own, but you can find our final version here.

One of the many advantages of Meemoo was that it allowed us to trigger events with the press of a key on a keyboard.  Once we had mapped keys to specific events (such as take picture and make GIF), it was time to build the controller.

The Controller (electronics)

The demo at Eyebeam was controlled by a MaKey MaKey board.  MaKey MaKeys essentially allow you to turn anything (like, say, a bunch of vegetables) into an input device.  We were about to get our own MaKey MaKey but realized that – as cool as it was – it was a bit too cool for what we wanted to do.  All we really needed to do was type four letters.

That’s where the USB keyboard idea came in.  It turns out that USB keyboards are quite hackable.  Plus we had another one hanging around the office.  Once you take them apart you are left with some simple elements – a membrane under the keys and a board that connects the membrane to the USB cable.  

For these purposes all we needed was that board.  Keyboards are basically matrices – there are two banks of connectors and pressing a key connects one pin from one matrix to one pin from the other matrix.  The combination of the two pins is mapped to an input like the letter “g” or the right arrow.

So the first step was to take apart the USB keyboard and pull out the controller board.  The second step was to try and map the matrix.  This sounds a lot sexier than it was.  In order to map the matrix we plugged the board into a computer and brought up a text editor.  We then just took a wire and connected one pin from one matrix to one pin to another, working our way through until we had identified at least four combinations that would reliably give us letters:

With the pins identified, we soldered wires onto the pins and connected them to the pushbuttons.  Pushing the buttons now caused the computer to recognize the key that corresponded to the pin combination.

With the wiring done, we now had a neat way to trigger all of the events to make the animated GIF.  But we couldn’t just leave the buttons lying around in a jumble.  They needed a home.

The Controller  (enclosure)

We could have pretty much used anything to hold the buttons.  A shoebox with holes drilled in the top would have worked fine. But we decided to go for it with a custom designed laser cut box.  

In addition to looking cool (a not insignificant factor), the laser cut box would allow us to include instructions right on the box.  This was important because we wanted people to be able to use the GIFerator without a PK staffer having to man it all night.

We used this site to create the box and then PK’s own Clarissa Ramon added the holes for the buttons, the instructions, a PK logo, and an Open Source Hardware logo in illustrator (inkscapewould have worked too).  

With our design mostly done, all we needed to find was a laser cutter.

Fortunately, the fantastic Fab Lab DC has a laser cutter.  Fab Lab impresariess Phyllis Klein helped us refine the design so it would cut well.

And then it was time to cut.

And put it together

Once Phyllis’ husband Alex helped us join the box pieces together, the last step was to screw in the buttons and connect everything together.

Controller Plus Software

The next step was to make sure that the keys on the box matched the keys Meemoo was looking for.  That was easy - when setting the input for a specific action we just pushed the right button on the controller.  But then things got a bit more complicated.   

While Meemoo made it incredibly easy to create the GIF, it is not yet easy to save the GIF.  Without that, every GIF would disappear every time we reset the page.  

Fortunately, PK’s Charles Duan tapped the same programming skills he used to explain to the Supreme Court why seemingly complicated patents are really quite simple to write a custom Firefox plugin that saves the GIF locally.

At that point, we had a controller that did four things – started a new GIF, added frames to the GIF, animated the GIF, and sent the GIF to the internet.

But wait.  Charles’ plugin just saves the GIF locally.  How does it get onto the internet?

Getting the GIF Online

This walkthough from IDEO helped us figure out the final step.  We used an IFTTT recipe to monitor the local save folder (synced with a public dropbox folder) and upload any GIFs it found to our tumblr page.

That’s It

The GIFerator actually worked at the IP3s and was a big success.  We’re planning on setting it up at the entrance to PK’s office, and maybe take it on the road a bit more.  As you can see, once we figured it out it was pretty easy to put together.  So come on by and try it out.  And go build your own.  Improve upon our design.  If you do, send us a picture of it set up and working!


This one is pretty straightforward.  I wanted to make a big red button that would play a sound of my choosing when I pushed it.  I thought about trying to reprogram one of the Staples easy buttons or even trying to design my own custom board.  But both of those options turned out to be much more complicated than what I ended up with.

Turns out you only really need 2 things:

1 Big Dome Pushbutton (I got the economy one from Sparkfun): $5.95

1 Recordable Sound Module from Invite By Voice: $6.99

You will also need some sort of box.  While I have dreams of making a laser cut box for this thing, as you can see a Sparkfun box works pretty well for  now.

The sound module is incredibly straightforward to use.  Basically you download a bit of software from their site, connect the dongle, and upload it to the module.  It can actually support quite a long sound clip (at highest quality about 100 seconds) which is totally overkill for my purposes.

Once you have uploaded it, pressing the button will trigger the sound.  But the stock button is a tiny little thing.  That’s where the big red button comes in.

The button is just a big way to trigger the switch.  To introduce it into the sound module, just cut off the stock button and solder in the switch that comes with the button:


Note that you only have two wires but the switch has three terminals.  One of the wires attaches to the terminal at the bottom (labeled “COM”).  Choosing between the other two terminals defines the  nature of the circuit.  The top terminal would make the circuit default on and cause the button push to break the circuit.  The bottom terminal would make the circuit default off and cause the button push to connect the circuit.

Since we want the button to turn the circuit on, I used the bottom terminal. 

Now connect the switch with the rest of the button.  At this point you are pretty much done - pressing the button should play the sound.  But just having a button isn’t quite enough.  It really needs a place to live.  Enter the box.

First I cut (actually drilled) a hole into the top and installed the button:


The cardboard muffled the sound a bit so I cut a second hole and attached the speaker to it:image

Note that I also cut the speaker off from the control module itself.  The wires are still attached, but cutting the paper let me position the speaker a bit easier.


And that’s really it.  It took me a long time to figure this out but once I found the sound module it was super easy.