Baker St Mapping evening + Artillery Arms Tonight

Posted by Harry Wood on 29 May 2014 in English (English)

We’ve got a London OpenStreetMap pub meet-up tonight! Join us in the Artillery Arms from 7pm. It’s just a social pub meet-up. We’re alternating, so last time was a “mapping evening”.

A couple of weeks back we did a bit of mapping around Baker Street. When I say “we” I mean me and Marco. He turned up at the designated meeting place, which was a surprise. But then we failed to find someone else who was supposed to be joining us for a walking talking mapping demonstration. Turns out she was actually there, but somehow didn’t find us. I was loitering outside a bank, wearing an OpenStreetMap Polo shirt and clutching a leaflet, but next time I need to make sure everyone has my mobile number.

But luckily Marco was there, so we were “go” for a mapping session. A couple of weeks back it felt like the height of summer, and so it was very pleasant to be wandering some quiet backstreets in the evening. I demonstrated taking lots of photos of shops and buildings, and then I took some photos of Marco taking photos of shops and buildings.


We took a snaking route which took in a few roads full of shops, and lots of higgledy piggledy buildings with funny shapes backing onto quiet cobbled backstreets.


Then we went to the pub! The metropolitan bar. It was quite a quiet meet-up. Just five or six of us. Actually there was one more new person who tried to find us at this stage, and again failed to find us for whatever reason. She was someone who works for what3words, so it was probably a good job she wasn’t there, because we weren’t being all that complimentary about it.

We were discussing it because it was presented at geomob the night before. what3words is yet another idea for a scheme to locate things, kind of as an alternative to the postcode. There are many such ideas to choose from, so whenever anyone comes with a new one, put on your skeptical hat. It had better be good, otherwise why not use one of the many other schemes? So the nice unique selling point of what3words is the readability and memorability of describing a location with three words. Good idea, but sadly what3words scores some pretty big fails. There’s some subtle points, like maybe it should be hierarchical with similar word combinations representing nearby locations. Some not so subtle points, like it requires access to a database of word location mappings. They could have designed it to be algorithmically reproducible for compact offline use, but the word location mappings are arranged within a database instead. This makes for some advantages, e.g. city locations use easier shorter words, while longer words are used for offshore locations. But of course the main advantage is for them. The old API lock-in chestnut. And no, the database is not open data. I’m afraid that’s a massive FAIL. There’s no way I would ever recommend using a system like that over and above a free and open scheme (of which there are many to choose from). Sadly they go beyond this fail, and also have patented the what3words idea, so now nobody else can do it better which is a shame.

Better alternative? OpenStreetMap shortlinks is one which has been around for a while. e.g. A simple geohashing algorithm resulting in nice simple short links. It’s pretty awesome, and not really celebrated enough. Matt was at the pub to join in this discussion. He was keen to stress that he designed shortlinks to solve a different problem (making map links short!) but OpenStreetMap shortlinks are an awesome way of sharing a location in lots of ways. Chief among the advantages: it requires an algorithm not a database. I often joke (half seriously) that we should push for adoption of shortlinks to replace postcodes. We got onto discussing how Ireland doesn’t have postcodes (??!). Probably the main reason for this is that there are too many proposed schemes to choose from. OpenStreetMap shortlinks are the best though. I think I’m going to start sending all mail with shortlinks written under the address :-)

We talked about 3D buildings complications. How to map a thing on the 2nd floor of a building spanning over the top of a row of shops? That sort of thing. Personally my answer is to not really worry about it too much. If it’s getting too complicated, just map things as nodes within a building, with a level tag. But maybe that’s just me being sloppy.

We talked about searching for “Kings X”

We talked about android routing.

We talked about and we classified the pub we were in, but this immediately threw up the usual problem I have, of deciding just how wheelchair accessible a place is or isn’t. But anyway we all agreed that wheelmap is awesome, and is perhaps the best example of an topic-focused data collection application using OpenStreetMap as the back-end. We need more of these!

We talked about the curious situation which is developing, as a number of “sister” organisations (H.O.T. and OSM France, and maybe OSM U.S.) are starting to move money around on a much larger scale than the core OpenStreetMap Foundation. This is weird and maybe a bit annoying for the people who are worrying about the core OpenStreetMap servers. We have a lean small organisation which raises a bit of money in donations, and makes income from running an annual conference, and then spends a bit of money on server hardware. Mostly though, in the grand scheme things, OSMF doesn’t move much money around. [Edit. I originally used the word “satellite” organisation, but that has a confusing double-meaning in OSM context]

I’m not sure of the details of OpenStreetMap France, but I know they attracted some large number of euros of funding, and one result we see is spending on OpenStreetMap France servers which run various things, particularly various services set up by Christian Quest. Not sure if Christian draws any salary on that. It would be interesting to know more about their financial situation. I do know more about H.O.T.’s finances. There’s a full time employee Kate Chapman, and a few others in Indonesia. So that immediately makes it seem as though H.O.T. is flush with cash, but actually the majority of HOT’s income is from projects, and the majority of employees time is spent working on these projects. A lot of income is “grant” funding. I guess the OSM France money was the same. Important point about this is that it’s money with lots of strings attached. Money for doing various specific projects which have gone through detailed proposal phases. This kind of thing tends to require paid employees (a mysterious concept!) That’s partly because somebody has to put quite a lot of effort in to drawing up proposals, also partly because in the proposal itself you’ll tend to be committing man-hours to work on a thing (difficult to make such commitments if you’re just hoping to find a volunteer to do it), but perhaps mainly because sadly people simply don’t give out grant money to organisations unless they can see a traditional organisational structure. With H.O.T. we also get charitable donations. People like the fact that we’re a 501c registered charity, for tax reasons, but again charity law forces us towards a very traditional organisation structure.

In recent times I’ve become aware of more opportunities for grant money here in the U.K. both for UK-based humanitarian projects and also tech funding of UK-based open data projects. In both cases I’m hobbled in my ability to chase down the opportunities, because I don’t have much time to spend on speccing out a project proposal, and when it comes to proposing who will be doing the work, I definitely do not have the time to be putting myself forward for it. With this in mind, it feels like we need a HOT UK, and an OSM UK organisation. I think it would be good to do. If I was at a loose end, I’d totally do it.

But of course we’d be moving into a situation where, yes lots of money is moving, but mostly with strings attached. OSMUK for example, would be another sister organisation not really helping with core OpenStreetMap funding. But maybe it could help after growing a bit. In fact I will stick my neck out and make a controversial suggestion here. I think we need to keep an eye on sister organisations, and their finances, and eventually move to a situation where (if it seems reasonable/sustainable) they would contribute a bit towards the core OSMF. I believe wikimedia chapters are organised this way. Most local chapters have fund-raising as key activity, and some proportion funds go towards the global foundation. I think a “reasonable/sustainable” test would be quite tricky. For example I know H.O.T. has good quarters and bad quarters with the eb and flow of project money, but we’re unlikely to be swimming in “spare” cash any time soon. Maybe the bigger problem is, on whose authority we are demanding this funding arrangement? I suspect OpenStreetMap France would kick up an almighty “zut alors!” at the very suggestion of it. After all I have been told in no uncertain terms that OpenStreetMap France does not consider themselves a local chapter of the foundation. Bit of a tricky discussion to be had there methinks :-/

Matt and I were chatting about these things, but no solution was forthcoming. More beers needed perhaps!

In fact having written all that, it seems to be beer o’clock RIGHT NOW.

In London? Come join us at the Artillery Arms

Location: Marylebone, Westminster, London, Greater London, England, W1G 6JN, United Kingdom

Comment from cquest on 30 May 2014 at 10:19

Hi Harry !

For your info, I’m not getting money for what’ I’m setting up with other guys at OpenStreetMap France. I’m not alone behind all the stuff we’re running, coding, hosting on OSM-FR donated servers.

We’re a small team of a dozen of volunteers taking care of a dozen of servers all donated to us (some old ones, some brand new). Hosting is also donated in different places (universities, companies, etc).

Regarding the money we’ve been able to collect, a large part of it has been donated to us for humanitarian and social development goals and comes from the “Fondation de France”.

We also collected some money this year thanks to our 2nd State Of The Map (in april) which will be used to run our organization.

OSM-FR has no paid employee.

Helping the core OSMF funding/hosting is something we would really like to happen, working closer together. We’re quite new organisation (a little more than 2 years) and still in the process of organizing ourself better… trying to cope with fast OSM adoption that is happening here thanks to so many government and local authority contacts we’ve established in the past years.

Our last move creating “BANO” (nationwide open address database) should even accelerate all this !

Comment from Harry Wood on 30 May 2014 at 12:07

Thanks for that Christian. Good information. Of course I should’ve just asked you about it really before posting guesswork on my diary entry, but as you may have noticed I tend to use my diary as place for blurting out what’s on my mind :-) (plus I needed to get the post out ahead of the pub event last night)

So that’s interesting. I was guessing you were drawing some salary/payment for all the good work you’re doing. Seems like you’re managing to get a lot done! Also lots of donations of servers and hosting. That’s also interesting, and makes me think more about the idea that we would be able to attract similar donations e.g. from universities in the UK if we got organised (I mean for some sort of separate OSMUK thing above and beyond the support we already get for OSMF servers)

Good to know you wouldn’t be opposed to the notion of OSM France helping the core OSMF funding. I had another big chat about this with Matt in the pub last night. He’s not really in favour of structuring things like that, or didn’t think would work. So yeah… just a vague idea from me really. Matt also reminded me the “Local Chapters Working Group” was supposed to be coming up with a plan for such things. Not sure how that’s progressing though.

Comment from baditaflorin on 3 June 2014 at 06:49

Did you use mapillary at the mapping party ?

There is a plugin to export and use in JOSM

Comment from Harry Wood on 15 November 2014 at 14:28

Chris Sheldrick from what3words actually read this diary entry and then chatted with me in the pub about it at the following geomob. We sort of agreed to disagree. He understands the desire for openness, but obviously he’s trying to make a business.

On an open versus closed debate, I’m fully able to comprehend the closed money-making side of the argument, but the trouble is with something like this, it is “standards defining”. It’s not just a business idea for a new social network or something. The name of the game is to persuade many many people adopt the scheme as a way of locating stuff. So my point is… I would persuade people to adopt an open scheme. One which is not going to go pair-shaped (along with its closed copyright-protected database) when the company goes pair-shaped.

I chatted about that longevity aspect of things in the pub, and he acknowledged this concern. He said he could look into writing a kind of get-out clause which opens everything if the what3words company goes under. I’d certainly welcome that, but I supposed another problem is that “goes pair-shaped” does not necessarily many “goes under”. What3words might go into a sort semi-liquidation state, where the opportunity is no longer exciting, so they sell out and start to introduce new charges / advertising to bleed it dry. Not a hypothetical scenario. This kind of thing happens a lot.

One thing I guess I should concede. I confidently said there are many addressing schemes to choose from, including many open ones… I’m not sure where there’s a good list of such ideas. It’s a shame nobody seems to have prior art on the idea the idea of words mapping to a geohash (would love to hear if somebody does). The idea does seem pretty obvious to me, but maybe only after it has been stated :-)

Comment from SimonPoole on 16 November 2014 at 16:30

BTW see for mapping POIs indoors ….

Login to leave a comment