Catarina Mota – Open Hardware

Mota, a distinguished resesearcher and founder of the Open Hardware Association  opened Day 3 with her keynote on open hardware. She explained how open hardware was lowering the barrier to entry to a number of fields such as DIY 3D printing, DIY biomedicine and many others – by making it easier to get involved.

She told the story of Sparkfun, a company which was started through the need to source parts for open hardware. Parts were shipped from someone’s basement – and they’ve now grown to have a turnover of over $20 million annually.

She went on to demonstrate how the growth of Arduino has been exponential and how ‘backyard brains’ are now changing the world through open hardware, producing projects like the global village construction set – which is essentially lifesize lego. 3D printing is also starting to reach maturity, with technology such as Repraps and Makerbots having more of a presence in the community.

This growth has led to the birth of the Open Hardware Association as a body to help support the community as it grows up.

Mota drew a number of differentiations between open hardware and open software. Asking the question

What does open mean for hardware?

she explained that the assets used in open hardware were things like bills of materials, schematics and CAD designs, and that standard formats for the production of these were again lowering the barrier to entry of the community. A challenge remains in terms of an equivalent repository to something like GitHub or SourceForge for open hardware.

Some of the other challenges facing the open hardware community were things like sourcing local parts and materials. For example, the specialist plastics used by Reprap and Makerbot are difficult to source and expensive if you live in the developing world.

Mota also explained how open hardware works through iteration, in contrast to open software which works through a process of collaboration. Each hardware design is an iterative and possibly derivative design – rather than collaborative.

Small and medium scale component manufacturing also remains a challenge – for instance breadboard designs are not efficient to produce on a small scale. However, there is now a lot of momentum to aggregate these – say 4 or 5 developers having their boards printed at once by a manufacturer to reduce costs – effectively leading to the crowdsourcing of manufacturing.

Mota illustrated how open hardware is leading to a rebirth in tinkering and repair – arts that have been lost within a consumerist society. Technology like Ardunio is being used in education to pique the interest of students in technology, and the business model that is currently used –

give away the bits, sell the kit

is still apparently viable.

Mota concluded that 2013 is the year of open hardware. The community is on the cusp of some major developments – such as breaking through into markets such as consumer electronics – and there are exciting if uncertain times ahead.

Mota’s presentation was one of my favourites so far of LinuxCon – not just because it was from a female presenter who is so well respected within the industry, but also because her presentation style and general demeanour were so respectful and humble. She was one of only a handful of presenters who thanked the Linux Foundation for the opportunity, and her whole attitude of wanting to learn and grow from the open software community was lovely.

Linux Torvalds – Where Linux is going

The room was jam packed to hear the Linux superstar being interviewed on stage by Dirk Hoendl. His first question what what he was most proud of – which he answered with what can be achieved by the community as a large body. In fact, it is the community, and what they do with Linux that keeps Linus so motivated. Technologies led by the community but founded on Linux – such as the one laptop per child project are an example of what the community can do together. He also noted that he finds some of the smaller open source projects to be the most interesting.

When asked by Dirk which award he was most proud of, he stated that he didn’t really care about awards, and again posited the community as his key motivating factor. His key quote here was definitely worthwhile;

There’s no Nobel Prize for computer science

When asked what concerns him about the future of opensource and Linux in general, Torvalds stated that a number of opensource projects that he’s seen have less direction than they could have, and that they would benefit from taking direction from outside. He was also a little concerned over personality clashes within the community – stating that while disagreement is a catalyst for growth, sometimes it can be quite personal.

He also mentioned that the patent system was fundamentally broken- an unsurprising claim from an open source advocate. Similarly, he talked about how embedded Linux remains a challenge – particularly for things like mobile phones – here referencing the divergence between the Android kernel and the main Linux kernel (although now they are set to merge again). He explained how supporting embedded systems can be a lot of work for kernel developers.

When asked who is the next Linux Torvalds, the next innovator, he sidestepped the question by stating that

sometimes the old ways are the correct ways

particularly with respect to operating systems.

Torvalds also mentioned the lack of diversity in the community when asked about he age of kernel developers steadily increasing – citing the fact that many are now employed by tech firms once they become developers, in contrast to students and younger developers several years ago. He broached the gender question himself, stating that 99% of Linux developers were male, and that the community needed more female developers, but that he wasn’t sure how to fix this.

I thought this showed a lack of insight into the work of groups like the Ada Initiative (inter alia), who have done research into open culture participation by women. This research has led to concrete, tangible recommendations for steps that can be done to increase participation by women.

Stephen Hemminger – Taking the fear out of contributing

Vyatta’s Hemminger set the scene by stating that it is hard to start contributing to open source generally – and to the kernel in particular – when there isn’t a lot of support – you can be easily scared off.

He covered the different reasons why people contribute to open source, such as wanting to learn, being intellectually curious, needing a challenge and to a certain extent being altruistic. He also mentioned that people wouldn’t do it if it wasn’t – at least in some way – fun.

Hemminger gave an overview of what the kernel contribution process is supposed to look like – and contrasted this with what it *is* actually like for developers – where there is a lot of flaming and trolling in the mix instead. In his experience, he has found encouragement and feedback really important – particularly when navigating a community where fiefdoms and political battles exist.

He covered three main cases of submission failure;

• Big changes – Massive changes are less likely to be accepted because they are harder to review and assure
• Arrogance – an unwarranted sense of pride in ability can detract from how the utility of a patch is perceived by the community
• Divisive changes – Changes that are incompatible, are proprietary, are reinventions or just plain ugly are likely to fail.

So, how can this be addressed? Hemminger drew a number of analogies with his involvement in Toastmasters, where the process for evaluating public speeches is highly formalised. The techniques used by Toastmasters, such as being prepared, active listening and sandwich techniques for giving feedback, focusing on the presentation and not the person – were all worthwhile techniques to improve kernel submission evaluation.

Hemminger also touched on the role of mentoring within the community, and explained how local mentoring is generally better – as face to face can be very productive. Having the role of mentor clearly documented helps  – with functions such as being a local dvocate, and a cheerleader for the person.

This presentation was really about culture change within the Linux community – and while the assertions were made I remain skeptical that they will be adopted widely.