A guest post – Solving a problem with Open Source Hardware by David Carrier (Parallax)…
I recently listened to a Freakonomics podcast that talked about using contests to promote innovation, in this case the efforts of the X Prize foundation. If you listen to the podcast, at 18:15 they mention the significant effort that people put into the contests. Some lost their fortunes, their homes, and their personal relationships in an attempt to win. I see it as somewhat of a moral dilemma, because despite the costs, the benefits of the contest on the community far outweigh the accumulative costs to all the participants. One downside of many contests though, is that often only the work of the winner benefits the community. Everyone else’s work is just a less successful attempt to do the same thing.
It is natural for businesses to make the most of their resources, so some have found a solution to this. Instead of a specific goal, pit everyone against each other in making the best design based off of your hardware. The downside though, is that to justify the cost of the contest, they claim intellectual ownership to everything entered. This does solve the problem of everyone working apart to produce the same thing, but their claim of ownership reduces the benefit to the community. It is only worthwhile for the company to host the contest if it directly benefits them, so without the claim on intellectual property, the contest may not be feasible. There was a recent Hack a Day post about a contest like this, and many comments are not in favor. In fact one of the comments suggest applying for the contest to get the promotional hardware, then not following through with the entry to get back at the company for their “scummy idea”. How scummy their behavior is, is another debate in and of itself, but it definitely isn’t promoting a sense of charity within the community.
At Parallax I have been working on an embedded web server, the Spinneret Web Server, and we originally planned to outsource the firmware. It is an open hardware design, but we consider it our due diligence to have the basic firmware written. We learned from the PropScope that if we have a 3rd party work on the firmware, and we want the entire product to be open source, we have to agree on it from the start. We initially found someone to create the firmware, but it didn’t work out due to availability. We had already been planning on hosting a design contest, and requiring every entry to be released as an open hardware design, but we didn’t have any official firmware for contestants to base their projects on. This ended up putting us in a situation that required an innovative solution. Everyone entering the contest needs to write code to perform the tasks specific to their project, but there is also significant functionality that is common to most entries. We could have written this firmware ourselves, or outsourced it as we had planned, but that does increase the cost of the product. We could also let all of the contestants create their own versions of the core functionality, but we would be back in the conundrum I started off with, with only the best version being of benefit to the community. Why not incorporate the creation of an open-source base firmware into the design contest itself?
We are basing 30% of the contestants score on how well they interact with and contribute to the community, and 20% on how professionally they present their end result. The other 50% is divided evenly among the more individualized aspects of the contest, how capable their project is and how useful it is. Traditionally, it has been in the contestants best interest to keep everything secret until the contest is closed; now half of their score is based on how much and how well they share. This will benefit the community, because the firmware will have significant design time behind it and be open-source and available to everyone, the contestants will be working with each other instead of against each other, and the costs of internal firmware development will not increase the price of the Spinneret Web Server. (It is only $50 it would have been $70 or more otherwise.) Had we written the firmware ourselves, there would also be little incentive to create open-source additions, stifling future creativity.
At the Open Hardware Summit, one common question that was difficult to answer on the spot was: How does open-source hardware benefit your business? (Or in more terse terms: Why should I give away my designs?) This is an excellent example; it significantly benefits us too. Since we are selling the Spinneret Web Server for only $50, it will apply to a much larger market. We know from experience that the open-source community created firmware will be very capable and high quality, which will also boost the marketability of the hardware even more. Yes someone may copy our design, but we make the microcontroller in it too, and WIZnet who, is cosponsoring the contest, made the Ethernet controller in it, so we both benefit either way.
If we had taken the more traveled path, and required all contestants to give us their designs, there would be less interest, everyone would be working against each, other and the quality of the work would be much less in the end. Because of this, we would still need to develop our own firmware, and there wouldn’t be an open-source community to embrace results of the contest and continue development after the contest has ended. I love it when everyone wins.