Open Hardware – Crowdfunding

Librem 5 [1] (Figure 1), Purism's free-licensed phone, was supposed to be a major milestone in open hardware. Other efforts at a free phone, ranging from Openmoko to the Ubuntu Touch, have failed in the past, but the Librem 5 looked like it might be different. Yet as the Librem 5 starts to ship, some are already taking its difficulties as proof that a free phone is unpractical. Very likely, the problems will eventually be fixed, but, for now, the Librem 5 is starting to look like a cautionary tale about the dangers of crowdfunding and of backing unproven open hardware.

Figure 1: The long-anticipated release of the Librem 5 phone has yet to live up to expectations.

In August 2017, the Librem 5 was announced with a crowdfunding campaign. Already a critical success with a line of free laptops, Purism created considerable buzz with a series of announcements about partnerships with Gnome, KDE, and other projects in the Librem 5's development. Cynics voiced doubts, but many in the media – including me – expressed enthusiastic and repeated support. After all, the Librem 5 was something that many advocates of open source had awaited for years. It helped, too, that Purism brushed aside the difficulties of developing a phone that depended on neither Android nor iOS. In a 2018 interview with Linux Pro Magazine, Todd Weaver, Purism's CEO, said simply, "this is a simple equation of momentum and time. We have hundreds of development kits shipping to active developers. Plus, we have emulators, so we will see more and more applications being written and ported to PureOS. It is only a matter of time."

However, as the shipping date grew near, problems started to arise. Originally scheduled for April 2019, shipping did not begin until September 2019 – and then only to Purism staff and selected developers. Such delays are not unusual in any product manufacturing, but rumors began to leak out from ex-employees that Purism faced financial problems and was internally in chaos. To make matters worse, Librem 1, a suite of services for the phone, proved mediocre and inconsistent in design, and reports of overheating circulated.

Finally, in late November, the first customers began to receive their phones. However, as Jim Salter of Ars Technica reported, relating the early experience of a backer called Azdle, what was shipped was more of a prototype than a functional phone [2]:

"First of all, it's not really a 'phone' yet. There's no audio when attempting to place a phone call. The cameras also don't appear to work yet. Azdle reports 'installing and opening up Cheese' – Cheese is a very basic Linux video application, installed by default in many distros – 'I just get a message saying "no device found".' There's also effectively no power management yet, so the Librem doesn't last long on battery. It takes a long time to charge as well.

"The software needs polish in lots of places: Azdle notes that few apps so far understand mobile screen layouts, and there's no obvious indications as to which apps have or have not been updated. The charging LED doesn't light up when the phone's off – although the phone is actually charging. And fine-tunables like kinetic scroll – the ability to flick a scroll-thumb down hard, and expect it to keep scrolling for pages and pages like a thrown rock – still need tuning."

In other words, what is shipping as I write is lacking two of the main functions expected of a modern phone and is rough around the edges generally. With any luck, software upgrades will gradually provide the expected functionality, but could anything be more disappointing after two years of anticipation?

As I write, any financial difficulties appear to have been staved off by a second round of investors, but supporters might be forgiven for having briefly worried whether they would ever receive a fully functioning product. At the very least, supporters might wonder what would happen if they backed another project that depended on sales alone to deliver the promised hardware.

Why Transparency Matters

Such worries are not unfounded. Crowdfunding campaigns that do not deliver what they promise have happened before. In 2016, Kickstarter suspended the Skarp Laser Razor campaign [3], apparently due to doubts that the product could be delivered. Similarly, in 2012, Eyez, a Google Glass-like product [4], never shipped. Admittedly, the Librem 5 is not yet in this infamous company, but the apprehensions that it might be should force the crowdfunding campaign backers to think twice about what they are doing. The caution is especially needed with open hardware, which is the center of so many hopes. Just to look through open hardware campaigns is to fall into a hardcore techie's version of SkyMall or the Lee Valley Tools catalog and to be lost in a dazzling array of gadgets. In the excitement, it can be hard to soberly assess a campaign or to estimate its chances of delivery.

In the case of the Librem 5, backing the project originally seemed like a safe gamble. Purism is incorporated as a social purpose company, which means it can make decisions based on social concerns rather than being focused entirely on profit [5]. Even more importantly, it has applied for Respects Your Freedom certification for its laptops, and speakers for the company have regularly voiced support for the privacy positions of the Electronic Frontier Foundation and the Free Software Foundation. To open source advocates like me, Purism appears to be one of us and trustworthy. And in fact, this same line of reasoning is the main reason to believe that Purism will eventually release a fully-functional Librem 5.

But what if Purism, or some other company in the same position, collapses financially and is unable to deliver on its promises, even with the best of intentions? How can backers foresee that event? Certainty does not exist, and backing any crowdfunding campaign means the risk of losing your money. This is a seemingly obvious fact, although no warnings of it appear to be on any of the crowdfunding sites.

No simple answer exists. However, in looking at how Purism has handled the crisis of its releases, what strikes me is the lack of transparency in its public statements, especially when compared to other crowdfunded open hardware projects. Purism has explained next to nothing of its circumstances, stating only how it plans to manage the release – and the resulting vacuum has filled quickly with rumors.

By contrast, the EOMA68 recyclable computer (Figure 2), which I wrote about several years ago, has yet to ship at all. My impression is that project head Luke Leighton severely underestimated the amount of work required to bring a product to market. Feature creep may have also occurred. Yet the project struggles on. Leighton has repeatedly apologized for the delay and continues to write updates every couple of months, explaining what has caused delays and what his current efforts are. In his November 2019 update [6], Leighton writes:

Figure 2: The EOMA68 recyclable computer has yet to ship, but backers are sustained by regular progress updates.

"I'm grateful to those people who get it and say 'thank you' for the level of detail and the description of the experiences that we've had, because for their project, they then know what to expect, and enjoy being educated on the kinds of pitfalls that might occur, and how to get out of them. I learned from the OpenPandora and the Openmoko and many other projects, and am repaying that in kind."

Leighton's transparency helps to sustain his backers' faith in him. He may be learning as he goes, but at least some still believe that he is trying to deliver on his promises. Just as importantly from his point of view, because he keeps his backers informed, there is less room for rumors. By chronicling his mistakes, he may help future open hardware projects to achieve their goals. In that way, even if he fails, he is doing a service to the community.

Much the same is true of Keyboardio, which manufactures the Model 01 keyboard. Keyboardio took nearly three years to deliver its keyboard after its successful crowdfunding campaign, yet while backers chafed, they mostly remained patient. Throughout that time, Keyboardio's Jesse Vincent and Kaia Dekker demonstrated their prototypes extensively and wrote detailed blogs [7] about the challenges of development and manufacturing and of doing business in China (Figure 3). As with Leighton, no one could doubt their intentions. Keyboardio delivered their product two years ago; they are now developing new products and have transformed Keyboardio into a small business.

Figure 3: Transparency and public appearances kept Keyboardio backers patient despite manufacturing delays and setbacks.

The lesson of EOMA68 and Keyboardio is that manufacturers seem more reliable when they communicate freely about what they are doing. Transparency is a mark of sincerity and helps backers to trust the crowdfunding project. Purism's misstep seems to be that they have failed to keep their backers informed. Obviously, it would be embarrassing to admit to financial difficulties, if any existed, yet failing to communicate only makes their backers fear the worst. Of course, transparency is no guarantee of success, but it seems likely that any open hardware manufacturer who goes to the trouble of being transparent is also likely to be conscientious enough to deliver what they promised.

Playing the Odds

Transparency is not the only way to decide whether to back a project. Increasingly, I find myself making decisions about whether to back a project depending on the site that hosts the campaign. My personal preference is for Crowd Supply. Compared to Kickstarter and Indiegogo, Crowd Supply is a small site, but it works closely with campaigns to refine their approach and also sells products once they are released. By my count, its campaigns have an 85 percent success rate, 12 percent higher than its rivals. Because Crowd Supply is vested in the success of the projects it hosts, its projects seem a safer bet.

Similarly, I am steering clear of overly ambitious campaigns. For instance, I would welcome a Linux tablet, but having reported on Aaron Seigo's failed Vivaldi project, I would hesitate to back a similar effort, because I have seen the challenges. I would be more likely to invest in campaigns that involved writing software and using preexisting parts as much as possible.

I hope that Purism prevails through its present difficulties. The hopes for the Librem 5 were so high in the community that its failure would be devastating. Still, as the community waits for events to unfold, the situation can serve as a warning to be cautious about supporting open hardware projects. The excitement of innovation should not make backers forget that they are taking a risk, much like a venture capitalist. At the very least, some due diligence seems necessary.