User:Ewmjc/OpenStreetMap Foundation/2014 Review

From OpenStreetMap Wiki
Jump to navigation Jump to search

OpenStreetMap Foundation: Where We Are

This is written in my role as Management Team chair against a background of stormy weather at the OSMF - partly real issues, partly personality conflicts. It is my strong contention that the biggest underlying cause for concern is too few people doing too many things, and that we, I included, have been remiss in not addressing this earlier. I am also a long-term "insider", board member 2007, and this is also to share with you things that I have learnt and not properly communicated.

You will see that I deliberately avoid anything about project Vision. Vision is important, but there are other members of our community that are more insightful and eloquent than I. This is an analysis only and looks at the creation and maintenance of an effectively functioning organisation to support and protect the OSM database under a simple doctrine of "More mappers, more users".

I think a road map for 2015 can be generated from this. But at the moment, it is not a plan but a review plus a set of ad hoc conclusions and suggestions. I hope the OSMF community and its board will incorporate them into their discussion and thinking. MikeCollinson (talk) 11:13, 14 November 2014 (UTC) + later modifications.

OpenStreetMap is a Massively Open Organisation

OpenStreetMap is what I now term a MOO or Massively Open Organisation.

By some measures, for example average daily contributors, we might plausibly claim to be the largest such in the world. But we are not a legal entity, we have no formal top-down organisation and it is difficult, in a good way, to draw a box around us and say "This is OpenStreetMap, this is not OpenStreetMap".

But is still, en masse, we are an organisation. We have figured out how to do things, and importantly, how to implement change and not become fossilised. Most of this does not involve the OpenStreetMap Foundation at all. The OpenStreetMap Foundation does not own OpenStreetMap.

At our very core, we are an individual mapper registering at our website and then mapping an obscure railway station in the middle of Siberia. We are an individual wanting to build a community within their country and asking for help setting up a mailing list. We are country organisations, sometimes informal, sometimes formal legal entities as in the Philippines, sometimes with substantial funding as in France. We are Humanitarian OpenStreetMap team, a "vertical" interest group involved in crisis mapping. We are a programmer seeing that a search facility on OpenStreetMap data would be neat, and saying "Hey ho", and writing one. We are commercial companies using OpenStreetMap data or providing services for those that do. Oh, and we are the OpenStreetMap Foundation, we run the servers, we stage an annual conference and work on licensing issues.

A MOO, however, can be characterised by four things: Contributions/contributors, software/programmers, governance, operations. I will use these to underpin the main analysis of where we are.


A big success story, OpenStreetMap will be hitting 2M sign-ups soon and we have several thousand individuals working over any given calendar day.

It is also easy for "verticals" to start up ... organisations that focus on mapping specific data for specific needs. It is also easy for organisations and informal groupings to start up and focus on a particular area of the world and particular language and culture needs. HOT and OpenStreetMap France are my poster-children here. Other organisations can spring up that do not consider themselves to be in the very core of OpenStreetMap but use our database as their GIS platform. Everyone benefits. Wheelmap is a remarkable example here.

The Foundation's role here is to keep the servers running and meeting demand, ([Technical] Operations and Engineering Working Groups) plus act as the place of last resort for integrity of the database, (Data Working Group). The Foundation does a pretty good job at this, but more help always appreciated.

Beyond that, the Foundation does not have to be further involved but co-operation and encouragement helps. With even more people, the sooner we can map the entire world everywhere with observations that are of interest to anyone regardless of nationality, creed, gender or personal/organisational interests. So we need to be diverse and welcoming. I welcome Kathleen Danielson to the board as I believe this is a topic that she will bring personal energy to.


MOOs simply cannot exist without software in order to take what they do globally. Another big success story. A major tenet of OpenStreetMap is "Just Do It", i.e. individual or small informal/groups take an initiative to do something and the rest of us try to get in the way as little as possible. This worked, continues to work, but perhaps we need to be a little more structure and outgoing in inviting such peripheral projects into our core international servers and systems.


Opinions will be more varied here but I personally consider this also to be a success story. Nothing particularly broken, but further evolution required. I have high-lighted in bold what I believe are the elements of the OpenStreetMap governance structure.

Let me give you an example of what I mean by governance:

Linux =
GNU General Public License
+ Linus Torvalds
+ a machine for getting patches/enhancements to Linus' attention
+ a series of organisations that come and go. They take core Linux and deliver it as a package. Redhat, Ubuntu, Mint ...

[I hope I am being roughly accurate here, do let me know if otherwise.] This describes how a line of code gets to us, the public, and who gets to make decisions along the way. Notice that I did not mention the Linux Foundation. Yet it exists, .

In our case, we do not deal with lines of code but collected latitudes and longitudes with a tag on them.

I don't know whether it meets the formal academic definition, but OpenStreetMap is an Anarchy that actually works. Probably because the basic goal, "More map data. More, more, more.", is simple, understandable and cuts across political and dogmatic barriers.

Right at the very beginning of the project, we enshrined the Just Do It principle. Anyone at all can register an account, no need to ask. If you want to put a new tag on something, just do it. No need to ask. So, there is a low barrier of entry to contribution ... I see this as the most important governance principle that we have and one that should be duplicated where we can ... in software for example. This can have down-side but we have developed, and continue to develop, methods to support peer pressure as the main quality check and balance. The OSMF has also developed a technical support group and last resort police-force, the Data Working Group. This works well but is stressful on its members.

Our founder, Steve Coast, then deliberately took a decision very early on to completely relinquish personal control and pass it to a group, out of which the OSMF was born. I emphasise this because it is an event that I think is not fully appreciated nor understood. It has allowed us to explore how a MOO works in ways that Linux and Wikipedia have not.

The OSMF then took on the mantra of "Supports but does not control the OpenStreetMap project". That is very clear, so it makes tactical planning easier. And it is relatively well communicated to the general OpenStreetMap community.

Then we put in place a coherent open data license. In retrospect, I think that this actually was not as important as putting in place coherent contributor terms. The contributor terms our most power governance document. Clause 3 says that whoever ultimately publishes the database absolutely must do so under a free and open license.

Lastly, we have very recently achieved the foundation for a formal web of OpenStreetMap organisations. This is the agreed master version of the Local Chapters Agreement. [I will pull out a link to this and a few hard facts shortly]. A team effort over many years but championed to actual conclusion by Simon Poole.

So what DO we need to look at to evolve further?

When I complete my first draft of this review, what screamed out at me was that we need to turn the vague phrase "formal web of OpenStreetMap organisations" above into, to use jargon, something holistic. I.e. the combined power of each organisation is far greater then the sum of what each organisation is trying to do separately. I have therefore slightly re-written the following paragraphs with this in mind ...

1) I strongly suggest that the board draw up and present for OSMF member discussion a simply worded resolution. This will formally adopt what is known as the The Principle of Subsidiarity ... which basically means that "The Big should not do what the Small could do". We can achieve much by adjusting our mentality to be a team player in a team that currently consists of HOT and a number of formally and informally organised country groups.

Work done on Local Chapter Agreements has finally borne fruit and provides the framework. Another under-emphasised event! But it is an empty meaningless framework unless if does something useful day to day.

So, OSMF should recognise itself as an international body, but not the only international body, and that national bodies can and are doing a lot of what OSMF was involved in 7 years ago. Christian Quest, for example, points out that it is logical for national bodies to talk to national bodies, for example mapping agencies. So if you look at OSMF, that really only leaves us the EU and UN as interlocutors! Christian, by the way, is certainly someone to bring into this kind of conversation. Much of my thinking here has evolved from conversations held with him.

As a practical example:

StateOfTheMap started as a single global conference that also, via sponsorship, provided the bulk of funds to run the servers for a year. Now there are country and regional conferences. All to the good, except that we now need to regard to co-scheduling and, for me more importantly, OSMF is losing its revenue stream. Companies have only so much sponsorship money. Where should it go? It is nobody's fault but as we started operating in our own silo, we still do. That needs adjustment.

So we need occasional summit meetings with the OSMF chair sitting equally with HOT chair and chairs and representatives of countries. Perhaps a mailing list or some other kind of electronic forum too. I suggest we call this "OpenStreetMap International". This should be a loose experimental forum to start but can be formalised over time.

Back to the practical example: For SOTM 2016+, an outcome like this might work: Countries continue to volunteer hosting, but at a regional rather than international level: Europe Middle East Africa (EMEA), Asia-Pacific and Americas conference. OSMF global SOTM then rotates between them as a combined conference. ... or it might not work, but that is what OpenStreetMap International would look at.

2) In support of (1) we should tighten our mission statement:

"The OpenStreetMap Foundation is an internationally-focused non-profit making organisation. It supports but does not control the OpenStreetMap project, an open participation project that creates and distributes free geographic data for the world for use in creative, productive, or unexpected ways. Specifically, the Foundation is guardian of the OpenStreetMap geospatial database as a common asset of mankind. It protects and supports the database, encourages public contributions, encourages use of the database and works to maximise public availability of the data, both physically and legally."

To me, the ramifications of this are:

  • OSMF is the database, or to be exact, is the master copy of the database. Its core general operations are:
    • Running the servers to maintain access in and access out. [Operations Working Group]
    • Keeping the database free and open. [Democratically elected board + currently the not very efficient License WG, which I suggest it is time to transfer this to a paid Legal Counsel]
    • ... And raising money to pay for this! This is one area of my review that I think is deficient.
  • The OSMF can take on other things depending on need and the wishes of its members provided the following tests are met:
    • The initiative must directly or indirectly either "encourages public contributions" or "encourages use"
    • We've seriously considered our capacity to do it. In simple terms: Can we get the volunteers and/or money?
    • We've seriously considered whether some other new or existing group can't just do it better and we will not get in the way of they want to Just Do It.

3) House keeping. Internal governance of the OSMF is controlled by its Articles of Association. Great strides have been made over the last two years in making these synchronise with what we want to do and how we want to do it. There is still some tweaking to do, but I believe that this is no longer strategic in impact and only include because as of writing board member term limits is a hot topic and for the first time we'll have formal but non-annual General Meetings. I simply urge OSMF members to consider and only consider the long-term interest of the Foundation. Articles of Incorporation should never, ever by tinkered with for short-term or political reasons!

General Operations

So now onto the area where I think things are broken. But, if we focus, we can turn around in 6 months.

By general operations, I meaning doing stuff. And I mean doing stuff other than mapping and other than personally writing a cool software program. Technical operations, administrative, legal, PR, liaison, communication, help ... stuff.

We - OSMF members, the Management Team (which I chair) and the board - have collectively dropped the ball on this one. It has been an issue that we have slowly allowed to get out of hand. Too few people doing too many things. Wonderful vision ideas languishing because there are urgent mundane things to be dealt right now, and, oh, now I need to go to my real money-earning job or studies. Vision discussions turn into frustration and misery because of lack of follow-through.

Let me emphasise that at each SOTM we put out a clear call for folks to join working groups. And some folks have ... but not enough.

Let me describe our organisation. I am going to take a negative view, but start developing what I feel is the solution at the same time.

First, we tend to equate general operations with the OSMF. So, a question, why? As a thought experiment: Why shouldn't we raise money for OSM Philippine Inc. to run our servers and let folks who have less chance to get hand-on on experience of computer operations than those in London? ... Well, you may have a good answer to that one, but then there a lot more questions like that to ask!

But for now questions per se is not a concrete way forward, so I will focus exclusively on the OSMF.

OSMF Members and Associate Members

So. At the top of the organisation is the OSMF membership. And I mean top. OSMF members could do far, far more I believe. This is not happening because OSMF "insiders", such as myself, are not asking for help often enough and are not communicating what goes on day to day at an informal level.

Membership is growing slowly and has nearly reached 500. Some folks, hallo Randy Meech!, think that is too small. 500 versus 1.8M registered users. That is a mighty small ratio. A larger, more diverse membership base would hopefully make the organisation more stable, less prone to being taken over on a whim. I hope Randy and others will continue to work on that in a concrete fashion. I won't because I do not consider it strategic as regards operational effectiveness. Two reasons:

1) It is more important to grow national and linguistic-based organisations. That is where our representative numbers will be. The OSMF operates, has to operate, in English. That disenfranchises an awful lot of people.

2) This is more philosophical but, I consider OSMF to be a sort of walk-in parliament where you elect yourself. Having a parliament of 500 is manageable. Having a parliament of 10,000? Hmm, that gets difficult. I take what is almost certainly a minority position, but that I think OSMF members should be members because they really want to be members ... and if that means us being small, that is not the end of the world.

Anyway, back to the main track ...

I see OSMF membership dividing into 3 types of people, I'd like feedback on how true members feel this is:

  • You see OSMF as basically a force for good and you given them some money each year to keep the servers going.
  • You are passionate about OpenStreetMap. You'd like to have a democratic say in the future of OpenStreetMap as a whole.
  • You want to get more involved in doing stuff.

Some of you will be just one of these. Some of you will be a mix of more than one.

At the beginning of OpenStreetMap, basically everything that happened went through the talk list and was pulled apart in great detail. You did not need to be a member of the OSMF, and indeed, there wasn't one. In other words, everything got "reported" informally as buzz, chatter and the back-and-forth of dealing with practical issues. We are now too big for that. But there are only a few hundred on the osmf-talk list. But we don't do it there either. In retrospect, a mistake.

If you ignore everything else in this long, long narrative, here is the key thing I want to put across. I would love your feedback:

There is a group of perhaps 100-200 OSMF members who would like to help out but don't because:

  • You do not know what needs doing because folks like me have not been informally communicative enough and we simply have not asked your help.
  • The current "recruitment" mechanism of joining a Working Group puts you off. It involves a commitment that sounds as if it will take a lot of your time and be long term.
  • You think working groups are cliquey. This is absolutely not so! But if it is a perception, we need to do more.

OSMF Working Groups

Some of this won't make good reading, but again I am deliberately being more than a little cruel and certainly do not criticise any individual.

25 people doing 50 people's work. And that is not counting new stuff we could be doing.

Let's start with the good news.

Take away the OSMF board, take away the Management Team, even take away the entire OSMF membership. Leave the working groups. Much of what OSMF does actually deliver would just continue as if nothing had happened.

Operations Working Group and Engineering Working Group collectively run and expand our core hardware/software system. They have run it well from the beginning and continue to do so. They are a close-knit group, which can be both very good and a little bad. Slowly internationalising without any impact on efficiency.

Data Working Group. Another highly effective group. It is their job to remove and repair data due to copyright violation and vandalism. They also get involved in settling disputes between individual or groups of mappers, so are often between a rock and a hard place and get criticised no matter what. When I look at some of these criticisms I've never found any lack of due diligence, probity or wisdom but some more help with folks with people rather than technical skills might help take some of the stress off.

StateoftheMap Organizing Committee. The Dutch team that ran our Amsterdam conference took our conference planning and implementation to a new level and that machine and process has been running conferences well since by just focusing on the conference and getting on and delivering it. But the time of the standalone conference is passing and there needs to be more general communication and board involvement to fit in with a suite of other related conferences to the best interest of the community and also in fund raising.

Now some slightly bad news.

Communications from working groups are improving but still could be a lot better for the board to function and indeed for the OSMF membership to act as a source of discussion and help.

We have made some progress in that all but two working groups now publish fairly up-to-date and regular minutes to their wiki pages at . One of the other groups also reported consistently useful summaries which were published in the Management Team minutes.

Report from some working groups often a little cryptic. I.e. the report is made but it is difficult to understand without asking questions. Neither I nor the board have a good feel for what is going on. This is not necessarily the fault of the individual groups. The Management Team should also be pro-actively reading minutes, asking questions etc. This has not happened.

And the bad news.

The following groups are effectively moribund, i.e. not delivering what they need to. The are various factors, but the main one is under-manning and folks wearing more than one hat.

  • Licensing Working Group
  • Communication Working Group
  • Local Chapters Working Group (although it could be argued their work is done with a final Local Chapters Agreement).
  • Strategic Working Group

OSMF Management Team

I chair this, mostly because no one else would. Like some of the working groups, we are effectively moribund.

The basic idea was that the board restricts itself to being strategic, see below, and so the Management Team:

  • provides the board with the support and information needed to do its business.
  • looks after matters that involve discussion and co-ordination between working groups.
  • kicks everything else to the working groups.
  • collates information on what the working groups and presents it to the board in nice easy to read fashion ... no dancing over 20 different email lists and wiki pages looking.

But it has not worked like that. Meetings were sparsely attended, then I stopped calling them (no excuse but lack of motivation), then I called one and no one attended.

Reasons? Probably a mix of all of these: Board not being systematic in devolving interesting projects to the team, not an effective chairman - there needs to be a high amount of diligent task following, lack of interest by working group heads who could not see the point, all members doing at least one other major job within OSMF.

But the best comment I got was from Richard Weait. Paraphrasing: "I thought it was doomed to failure. It took the load off board members by taking the same people and placing them on another committee. Yeah, right".

The OSMF Board

A basic idea was developed about two years ago that the board should be strategic and avoid day to day operations. I.e. think and make decisions on the big picture, discuss and approve policies, embark in bold new directions, yada, yada.

I was an enthusiastic promoter of that concept. In the time I served, 2007-2010, we spent an annual face-to-face day enthusiastically discussing grand strategy. Then each month for the next year we would meet and never get around to following up because there was always a fire to fight. In other words we became a reactive not a proactive board.

So, free the board from day-to-day operational management, et voilà, a proactive board implementing vision generated by the community or by itself. Nope. Ain't happened. Progress, absolutely. But keeping up with the growing power of OpenStreetMap now that we have an increasingly viable map of huge areas of the world, no.

To conclude this "as I see it" part of this report, let me highlight Richard's words above. Here is a list of the hats folks on the last board wore, and apologies if I've missed any hats. Some not all concurrent, but you get my point:

  • Simon Poole - Articles of Incorporation WG, Local Chapters WG, License WG, Chairman
  • Frederik Ramm - Data WG, Secretary
  • Oliver Kühn - License WG, Treasurer and busy, busy businessman.
  • Henk Hoff - StateOfTheMap WG, Management Team, Membership Sec
  • Matt Amos - Engineering WG, Operations WG, Data WG, Management Team
  • Dermot McNally - Strategic WG, Management Team
  • Kate Chapman - Communications WG, Executive Director at the Humanitarian OpenStreetMap Team.

What Next?

My review ends above. But the obvious question is, "Well what should we start doing now?". Here very briefly are my personal conclusions. I will either work further on this section, or better, I would like to see a collaborative document starting.

  • We need more help for OSMF members, which I am speculating we can easily get by asking folks to do small and/or one-time, occassionally re-occuring things instead of asking you to join a working group. See the Do It team below.
  • We can also get help by working more closely with other OpenStreetMap organisations. A simple experimental start is to form "OpenStreetMap International", a loose forum consisting of the the heads of OSMF, HOT, OSM France, Japan and US plus any other countries that would like to join. They then pick one experimental topic and see what happens.
  • We need paid staff. I have deliberately picked these positions because I believe they can help immediately and, importantly, not mean any sudden change in direction or character of the OSMF:
    • LWG should transform into a paid part-time General Counsel plus helpers.
    • If we are to continue doing what we try to do already, (no reason why not, but we need to review), then we most definitely need a paid part-time coordinator.
    • We should also expand our current book-keeper's role to include preparing detailed un-audited financial transactions, (basically the journals and ledgers), into a format that can easily be published on our website.

The Do It team

We need to increase the number of people doing some sort of work, no matter how small, to 75 by the end of 2015.

No need to be OSMF members ... indeed it may be a path to membership.

Ideally a large proportion should come from outside Western Europe, USA, Canada.

They can be individuals or friends working together.

No need to join a working group.

No need to make any regular commitment. The only condition is that if you say you are going to do something, then you Do It or try to Do It.

We do it in two ways:

  • Identify small, and I emphasise small, tasks and put out help wanted messages.
  • Create an environment where folks who want to Just Do something but need some sort of OSMF blessing or access to OSMF resource, can do so.

I have already run a small pilot with the help of Dermot McNally in expanding his 2014 AGM election team. At least 17 people volunteered. My thanks to them. To my knowledge, (sorry if I am wrong!), Ilya Zverev and Severin Menard (a very active HOTtie!) have not been particularly involved in specifically OSMF activities before.