User Tools

Site Tools


The role of a wiki in open hardware projects


This is a topic I have been asked to write about by Evi Swinnen who manages Timelab in Gent. It's worth noting that Timelab already has a very small wiki and short piece about the value of wikis, so I will take that as context and background.

Why wikis in the first place?

The short piece referenced above already points out the value of a wiki as a way of editing and maintaining web pages simply, quickly and (relatively) easily, through the interface of the web browser itself, rather than the old-fashioned approach of uploading web pages with FTP. This enables as many people as desired to edit the wiki without needing special assistance.

But people sometimes make mistakes, and the more people there are collaborating, the more individual mistakes are likely. So, a second key feature of all common wiki software is that they keep a log of changes to each page, and allow people to choose to be notified by e-mail of any changes, so that unwelcome changes can be quickly caught and reverted. This also allows the editors to be identified, which serves as a basis for accountability. The more people that have permission to edit, the quicker mistakes can be corrected.

Another key feature of all common wikis is that links between pages within the wiki are very easy to make and track. Commonly, you simply have to write the name of another page in (double) square brackets, and there is your link. Conventionally, if the page hasn't been created yet, the link will show in red, and following the red link will allow you to create the page. Furthermore, on each page, it is easy to call up a list of other pages that link to the page you are looking at. This is a great advantage over normal web pages, that have no such automatic ability.

This ease of linking between pages encourages an emergent property of well-structured wikis: each page can deal with a limited amount of information about the subject being written about, and rather than having to be copied, this page of information can simply be linked to. This, in turn, means that wikis are naturally well-adapted to hold learning resources that can be navigated through easily by people who want to learn about a topic of interest. Bite-sized chunks of learning materials are usually much easier to digest and learn from than textbooks, or textbook-sized web pages.

On the downside, the simplicity of editing wikis means that the displayed format is usually very standardised. For sure, you can add images and graphics to a page, but it's not easy to display the overall structure of pages in a wiki in a graphical way. Text is the basis of most wikis, rather than graphics, and different people get on better with text or graphics. So, it is hard to create a wiki site with its own special visual appeal, and it is hard to create graphical navigation tools.

A while back I made a list of other technical requirements for the kind of wiki that I see as most useful, but those topics are beyond the scope here.

Wikis as knowledge commons

I've written earlier (2023-09-09) about what I see as the essence of an effective knowledge commons, being the information aspect, or knowledge layer, of a community of practice. And I expect this to be a vital aspect of applying wikis to open hardware projects.

The common aspect of knowledge commons, in my definition, is that there is some kind of community of practice that uses, maintains and governs that knowledge resource. They don't sell the knowledge to each other, but rather maintain it between themselves for the collective good — and they are motivated to do this, because a combined knowledge resource is going to be more useful than several separate, siloed ones. You could say, it has greater utility. The challenge, as with many commons, is the free-rider problem, when people benefit from the knowledge resource without contributing towards it any any fair way. Hence the need for effective collective governance. All these points will be common to any knowledge commons, not just open hardware.

With any broad topic, again including open hardware, no one person or small group will have experience of the whole field. We expect different people to contribute to different parts of a wiki or other knowledge resource. It is my firm belief that the social structure of the community of practice using a wiki needs a lot more attention than it has usually received. If the knowledge resource is to be opened to a wider community than the members of the community of practice – a surrounding community of interest, or learning community – then it is useful to be able easily to identify the group or subgroup of people responsible for maintaining any particular part of the knowledge resource, for example, to request updates or additions.

So, open hardware?

First, I'll list what I would expect to be the content of an open hardware knowledge resource (recognising that I am no expert in that field).

  • Knowledge about what open hardware exists, where it is, who made it, when it was made, etc. Software isn't located in space, while hardware is.
  • Open hardware needs physical tools for it to be made real, and each instance of recreating the hardware will involve physical processes. The creation process could be more dependent on local conditions, and so more valuable to record in a knowledge base.
  • Plans, designs and methods, along with notes on implementing them, would seem to me to be a natural part of any open hardware wiki, and this is similar to other knowledge bases.
  • A last point that occurs to me is that it would be good to include details, including contact details, of the communities of practice, their groups, subgroups and members, and especially who takes “ownership” of the different parts of the wiki. This is not only so that learners and other interested people can point out when something needs attention, but also so that human contact can be there to supplement what is written in the wiki.

The open hardware community appears to be much smaller than the open software community, and for software I get the impression that much of the knowledge commoning is done by word of mouth, in a shared lab, or at meetings such as hackathons. Where people are more spread out, there is more need to share knowledge online, and that is where a wiki can be a great help.

Commercial or non-commercial?

In my earlier piece of 2023-09-09 I drew out the difference between commercially sensitive and non-commercial knowledge. Now, personally, I don't know enough about open hardware to know which areas are commercially sensitive and which not, so my reflections here are no more than common sense.

Simply put, any open hardware project – indeed, any open or commons project – will need to consider what the commercial interests are in that project. Where there is no commercial interest, there are no concerns with creating a wiki that is open to browsing, and that ideally creates opportunities for communities of interest and learning communities to gather round. Many wikis (obviously including Wikipedia) are completely open to browsing by the general public. But what I have not personally seen in these cases is a way of registering personal interest in order to find and be found by others with a similar interest. In my view, doing this well will make a vast difference to the impact of an open project, and well worth considering from the outset.

Where a project is potentially a competitor to an established commercial interest, it is not advised simply to make a wiki open to public browsing. The danger with fully open-source licencing is that there is no restriction on commercial companies simply taking the knowledge produced and exploiting it for their own profit, with no return of value or benefit to the open source community. If the commercial business races to put something similar on the market, that might detract heavily from the impact of an open alternative.

I'm not saying that there are simple answers here. I'm just saying, don't be naive. Recognise the essentially predatory nature of capitalist corporations, and take proper defensive precautions, as you would against actual dinosaurs if they still existed.

The common ontology, and how it connects things up

Let's say, then, that you want some shared knowledge resource, quite possibly a wiki, whether it is open from the start or whether it starts out as private to avoid predation. My question now is, how can that be organised so that the knowledge is accessible and comprehensible to the people who want to use it?

With any complex field of interest, and I imagine this includes open hardware, people specialising in different areas will have different conceptual models of their area, and perhaps different concepts of the area as a whole. But if the knowledge resource is to be a true commons, it needs to be at least understandable by all contributors. If a wiki is designed and structured by one person or one group alone, the structure is likely to reflect the way that person or group thinks about the field. This will not suit other people. Thus, I advocate an approach I call ontological commoning, which in my way of thinking means helping or guiding different people and groups to some common underlying structuring of a field that makes enough sense to all of them. The “common ontology” that results from this process will then serve to structure the wiki.

This common ontology can include, typically:

  • terms and definitions across the field
  • how different sub-areas relate together
  • the ways in which different topics (and thus, pages) are seen as potentially relating, such as
    • being more or less advanced
    • preceding or following on from
    • supporting or putting a different view
  • the usual metadata fields: titles, authors, dates, etc., and in particular:
  • a common categorisation scheme making wiki pages or other parts easier to find, including
    • some kind of ‘level’ of learning, to help learning community people find materials they can relate to.

This, in effect, then covers what has been called a “skills framework” or “competency framework”. Such a framework then provide a reference point to describe what learning someone has done, and what level particular courses are aimed at. It also would provide a useful basis for describing the content of qualifications (or badges etc.) issued outside the mainstream educational system. This in turn could greatly help with recruiting people for projects, or even employment, in this case in the field of open hardware — though this can naturally be extended to any field.

In these ways, a wiki can serve a community of practice; a surrounding community of interest; those who want to learn, either by themselves of with others; those who provide courses or learning opportunities; projects or companies or coops who want to find suitable people. It's a broad and worthy set of aims.


d/2023-10-28.txt · Last modified: 2023-11-06 18:33 by simongrant