Recent diary entries
I've made a video. Check it out ...but be warned. It's kind of dull, until 6 minutes 22 second in, when the mapping goes TURBO!
The sped up mapping effect is quite fun. I might have to try that again some time. But the idea of the video was to show how (or at least give a rough idea of how) we go about adding in the data, because we did the usual mapping party thing with beginners, of only demonstrating collection of data. ...I'm talking about at the Marble Arch Mapping Party two weeks ago.
Contributing to OpenStreetMap (usually/always for me) comes in two very different stages: Collecting data, and Inputting data. This is a challenge when it comes to teaching new people, because you don't get a good understanding of what data to collect until you've delved into the OpenStreetMap data using an editor. Unfortunately the cart comes before the horse at a typical mapping party. In fact worse, the cart is left without a horse. We don't usually have the benefit of a computer room or somewhere sensible to bring out laptops and have a serious editing session, and in the event of having lots of new folks to teach, we're left only teaching them half the process. Obviously if no new folks turn up then this doesn't matter. But this was not the case two weeks ago...
As I described previously, I was making a conscious effort to do something a bit different and a bit more welcoming to newcomers for this event. Meeting under the arch of Marble Arch, I was worried I'd be there on my own, but no! Thanks to all the usual OSM faces who decided to make the effort to be there in case they could help (or maybe in case something really exciting was going to happen). Some unusual faces too. Wasn't expecting to see RichardF. ...And some new faces! which was the point of the exercise. We did OK at attracting new folks. If we include the new folks who were jet-lagged and went straight to the pub (disgraceful!) and other new folks who only caught up at the end... we had quite a few!
Our concerted effort to attract new folks was inspired by an academic paper which analysed OpenStreetMap London meet-ups and concluded that we were bad at attracting (and retaining) them. This paper popped up out of nowhere, written by some people we had never met, as I described previously. Happily this situation resolved itself! It was fantastic to meet the authors, especially Desi Hristova who joined us on our walking talking tour of Marble Arch. It was fun to show them the truth behind the stats of an OpenStreetMap mapping party. They said they will be giving some presentations about their work, so I'm imagining a slide deck full of serious stats followed by sudden digression into pub photos!
So the people who wrote about the mapping parties, and about attendance of new people at mapping parties, then attended the mapping parties. Self-referential! particularly as they significantly boosted the number of new people who were at the mapping party. The question is... will we retain these new people? Also how many new people will come to our next event.
Well as a matter of fact, our next event is TONIGHT! Although in the new scheme of alternating event formats, tonight's one is not a "beginner friendly mapping party", but a "social pub meet-up". We do have the next mapping party scheduled in for Thursday 14th June and I'd like to try to promote this to new people again (any help appreciated). but I'm afraid I didn't pick a location yet. We will discuss this in the pub tonight! New folks are also welcome at the pub of course, and there are in fact a few of them expected! See you in the Blue Posts from 7pm. Details on the wiki
This will be a very beginner-friendly introduction to mapping. So if you're in London come along, and tell all your friends. Anyone who's never tried OpenStreetMapping, or curious about other ways of doing it. All you people lurking following @OSMLondon ... Come tomorrow evening!
OSMLondon events have always been wide-open for beginners to come along to, but I'm stressing this aspect in the tomorrows event. This is part of an exciting new event formula in which we alternate "pub meet-ups" and "mapping parties".
Last month we met at the Iron Duke pub with the intention of kicking off the mapping season. This is a good spot for some of the mapping priorities of central London, and I'm pleased that Derick managed to fill in some awesome details around St James Street. For my part I had signed up for a slice of cake to the South, but I spent so long in the office preparing a print-out of building outlines to check... that it go too late so I just went the pub (ohhh! That's cheating!) It was the daylight confusing me.
Outside the pub (in the new evening daylight) the hot topic of conversation was...
Amazing and slightly weird that these people have analysed this in such detail. It feels a bit like somebody wrote a ten page academic paper about my personal bad habits. It's weird that they do all of this without ever attending a mapping party themselves, but we figured it was a scientific analysis in which the scientist decides to avoid interfering with their subjects!
In judging the participation and retention rates of mapping parties, their data source was raw edit data in OpenStreetMap. Can't argue with that. But the analysis seemed to me to have a few holes. They've looked at the location and times of mapping parties over the years, based on the wiki records. They did observe that in the summer of 2008 we had a lot mapping parties. This was actually a mapping blitz. A "marathon" of epic proportions, with evening mapping happening in far-flung locations, and happening every week. We were filling in obvious gaps in the map at that time. I imagine this yields some clear and easy to correlate editing data. These days there's fewer obvious juicy gaps to tackle. As a result mapping parties have largely been in central London, a bit less frequent, and probably attracting a bit less participation in actual mapping. I like to think that these days London's map attracts more casual localised contributions from people dotted around London. The character of mapping party events has changed massively over the years. I'm not sure if this has been properly accounted for in their analysis.
In analysing edits they used a 48hour time window. Justifying it as follows: "We found that in 40% of mapping parties the peak of activity was on the day of the event, while in 89% of cases the peak activity was within 30 hours after the party. In 99% of cases, the peak activity was within 48 hours, after which the daily edits stabilise to the norm previously observed."
That's pretty interesting, but certainly doesn't fit with my own editing patterns these days. I quite often wait several days before inputting data. Also these days I often add more data (more objects) tracing building outlines in preparation for a mapping evening, with only minor tweaks and additions afterwards. Back in 2008 however, I was probably quite diligent about adding the data in (new streets!) soon after, particularly as we needed to be ready to do it all again the following week! Also back then I imagine that editing activity would have stood out a lot more from the lesser background editing.
So I had those quibbles, however reading on, there's definitely some very thorough and valid approaches. For example "OSM users greatly differ in terms of the amount of contributions they make, and over what timespan. In order to quantify the impact of mapping parties on different types of users, we have grouped them based on the number of contributions they made in the six months prior to each party". So I'll be in their "Group 4" heavy contributors category. They go on to say that these users don't actually get much mapping done at mapping parties "We cross checked the names of some of these contributors against what is publicly available in OSM wikis, and found that many of these users take on organisational roles, visiting an area prior to the party, creating ‘cake diagrams’, and identifying ‘problems’ they wish the party to fix." ...Got me down to a tee. And I think there's quite a few other OSMLondoners who probably managed to wheedle their way into the heavy contributors category despite never really bothering with any mapping at the London events.
These groupings also allowed them to scientifically conclude what we already know, that we suck at retaining newcomers. In this graph we score quite well on retaining experienced mappers (pink and blue), but pretty hopelessly with the newbies (red and grey)
CONCLUSION! "mapping parties do cause an increased editing activity during the events themselves; they also sustain engagement over time, though mostly for already active contributors; however, they largely fail on their third goal of engaging new-comers. After just a week following the party, these users stop contributing to OSM and do not come back to other mapping parties again"
In our huddle of "Group 4" retained mappers outside the Iron Duke pub, we had a good chat about all these things ...then we went for burritos
So it is with these things in mind, that I've shifted the way the events are organised just a little bit. Alternating "pub meet-ups" and "mapping parties" Not a major shift, but half the time we'll just call it what it is, a "pub meet-up" in which people who know each-other are meeting-up. We don't preclude newcomers of course, but we face the fact that it will most likely be the usual crowd (Nothing wrong with that. It's always fun!). But then for the "mapping parties" we try to angle it a little more towards newbies. I'm not under any illusions that this small change will make much difference, but it's probably taking things in the right direction. Also if I can chill out on the promotion effort and the cake diagram drawing effort for some of the events, that will be welcome, and maybe I'll have more time to do better promotion of these less frequent "mapping party" events.... was the plan... but I didn't get much time this time around. Promotion is a job for everyone though. Please reach out to people who may be interested, and pass on the link: http://bit.ly/londonosm9
...and I'll see you tomorrow evening for a beginner-friendly mapping session!
Boom! summer has arrived!
Well spring time at least. And most importantly, daylight hours after work. This means it's mapping time again. We'll be heading out mapping TONIGHT in the Mayfair area
I'm falling behind with... well everything in my life right now, because I've had holiday sandwiched between lots of work. But I've cranked out a mapcraft cake diagram. I've also laid down an invite for any new folks to get in touch with me if they'd like a demonstration of the kind of mapping data I tend to collect while roaming the streets of London. We'll be doing this mapping from around 6pm until 7pm when we'll be meeting at the pub. From them onwards it will be... a good old casual social meet-up!
We haven't had one in a while, and yet I didn't get around to writing up my notes on things we chatted about at two previous pub meet-ups back in March:
Crosse Keyes pub near Bank 7th March
We talked about whether there would ever be an open version of google street view. There's openstreetview.org which isn't really it. Obviously stitched together photos of every street is a mammoth undertaking. With the prevalence of smartphones there is a increasing capacity for people to be able to swipe their phone across a panorama, and not only gather lots of pixels, but also use the phone's CPU power to do stitching/resizing and even image recognition e.g. gather the names of all the shops on this street in a single swipe. Boom. Mega-rapid mapping. That's still science fiction at the moment, but maybe things will go that way, and hopefully OpenStreetMap can steer this in an open data way.
That got us onto talking about UAVs and quadcopters. The technology's not there yet, but maybe one day we'll able to send out a mini quadcopter drone down every street at slightly above head-height to read all the shop signs. For now UAVs might start to be useful for getting aerial imagery, but even there, I haven't seen much success yet. We're interested in this for HOT disaster mapping. It seems like open hardware initiatives like http://www.openrelief.org might come up with something cool (and cheaper) but compare with the niftiness of something like sensefly (commercial).
We got talking about cycling. Having been cycling in London for over two years now I can rant and rave with the rest of them, although I don't follow things like the announcement from Boris which everyone was excited about.
As well as a cycling, Andy was talking rendering again, and handling of coastlines. Jochen's coastline downloads offer an "inverted" approach, letting you assume a land background and then paint on polygon areas of water.
We talked about bug tracking within OSM. We have a lot of very old "wont fix" bugs. Design ideas which are sort of left dangling for possible future reference. We talked about the Trac versus github situation. It's unfortunate we've ended up with bugs in two different places (for the main website codebase) Can't quite remember the reasons and solutions for that.
Grant wanted to announce that OpenStreetMap's SVN will be shutdown and discontinued at the end of May, because most developers (and more importantly the repo maintainers) are preferring git over SVN, and most sub-projects have been migrated onto git already. Don't know if Grant has announced that yet. Maybe I just did :-/
We got talking about OSM data archaeology. Grant had been delving around seeing who had copies of really old database backups, from the very early days of the project before the planet files and planet file archiving was regularised. The records are patchy back then.
Finally we were talking about recruitment into the OpenStreetMap Foundation working groups. We urgently need to do some of this (not least into CWG). We need some new folks to help us make things happen within the foundation. This means attracting the right kind of talent, but also as Andy was saying, we need to stop "fishing from a small pond" of people we know, who are already busy with OpenStreetMap stuff. I would say that this is one of the key challenges for the foundation right now.
And we ate "dragon" curries!
Monkey puzzle in Paddington 21st March
I brought along an old map poster which Matt had created back in 2009(?). This is fun because we can play "spot the rendering configuration mistake". It's difficult enough to spot, while at the same time quite glaringly obvious. It made it even more fun when Matt showed up at the pub. He's never going to hear the end of it.
Grant was stressed out after a couple of difficult sessions doing hardware upgrades/alterations on the main OpenStreetMap servers. Some issue almost meant he was not being able to the main database sever back up again. But it's OK. phew! We're going to be needing some new hardware again to keep things running smoothly.
We talked about road classification tagging. The old physical properties versus grid significance debate. We get this with African roads, but I always remember the road round Iceland as the best example. It's a dirt track in places, and yet it's the one and only road looping around the island, so in that sense it's a very significant road and deserves its primary/trunk tag. Paul was giving the example of Brick Lane as an unusual classification quirk here in London. By our rules we follow the government classification. It's a B road. It's a highway=secondary, and yet Brick Lane is actually a pedestrian clogged pokey little alleyway compared to what drivers might expect of a B road.
We talked about "facebook search", and some of the weird privacy issues hitting the tech news as result of that.
We talked about Nokia's funky 3D view... which isn't working for me just at the moment.
We talked about interactions in the community over IRC and wiki edits, sometimes somebody who is quite sensible/intelligent on IRC can be a nutter when editing wiki pages ...or vice versa. But often moving from slower channels onto IRC is a good way of turning a big flaming tit-for-tat debate into a reasonable discussion. (if you've never tried IRC, it's not really as geeky and mysterious as it sounds) ...of course, if you can, the best thing is a face-to-face pub meet-up!
And that's a wrap on the winter pub meet-ups of 2012-2013. Since then I've been mega-busy and on holiday in Brazil (did a little bit of mapping there). The weather is amazingly different on return to London. Before I was cycling to work in my wind-proof cycle jacket, with my face hurting in the cold. Now I can cycle in just a T-shirt and OSM hi-vis jacket. And the daylight... So much daylight after work.
Time to kick off the mapping! I didn't sketch my buildings yet but I plan to fill some annoying building gaps in this Mayfair area. In fact I'll be very disappointed if we can't nail that this summer. If you're in London, and you fancy a bit of that, or a bit of chit-chat on techy or less techy topics... come join in with the first London summer mapping (and pub) event 2013!
The next London pub meet-up is tonight! Join us at the Crosse Keys pub from 7pm. As usual we'd love to have some new folks come along.
We had a french lady called Victoire joined us last time at the Blue Posts. She was interested in getting started with OSM and learning more about how humanitarian organisations can work with it. So we were talking through all of that.
The advice I always stress while bombarding new folks with all sorts of OSM info, is that you have to try editing. Maybe you're interested in using OSM data, or getting involved in other ways, or spreading the word to others. Maybe you have ideas about contributing data sets, or engaging mappers to input particular types of map data, but step number one in understanding how to go about any of those things... have a go at mapping. Learn the basics of OpenStreetMap editing, and add something to the map yourself. Happily when I bumped into Victoire at an ICT4D event last week, she said she had taken this advice, working through the (the soon to be revamped) learnosm.org instructions and doing some edits with JOSM.
We also talked about IRC. If you think it's a mysterious super-geeky thing, you're not alone, but it's really not that hard. Go to irc.openstreetmap.org, pick a nickname (like your real name for example), Login, ignore the mysterious super-geeky mess that appears on the screen, and just type "hello" down the bottom there! IRC is a great place for friendly chit-chat or to fire a quick a question if you're stuck on something while editing perhaps.
We talked about the OSM leaflets. A new version (which we now need) can be quite similar design-wise. Obviously the license and the maps need a refresh, but we also need newer content for some of the instructions. How to map a street network from scratch using GPS is, in my opinion, not the most relevant mapping lesson convey to the majority of people who will receive this leaflet. Adding a restaurant POI might be a better lesson (because things have moved on since the leaflets were first created!)
We discussed the idea of an "OpenStreetMap App Of The Year" award, and how that might work. We have a bewildering choice of iOS and Android apps, and (like other bewildering choices OpenStreetMap offers) some way of highlighting the best would be good.
For conversations about London attractions I have never visited, or something more OpenStreetMap related... join us in the pub tonight! The Crosse Keyes pub has one massive room in an old bank building. We'll have to put a sign up to help people find the OpenStreetMap table. See you there!
First of all... It's the London OpenStreetMap pub meet-up TONIGHT. Come join us!
I've been working a bit more on the map embedding examples I have on my website. For a long time that URL would take you to an apache directory listing of a bunch of files, including various junk experiments. But I kept linking people to it in discussions, and it was getting a bit embarrassing. Time to tidy it up a bit.
More recently I've set this up to use URL rewriting, give nice ".view.html" URLs and simultaneously fix problems with relative file references. I've also set it up to pull out some explanatory text from the HTML file. I know there's a tonne of other approaches to this using jsfiddle, github gists or others, but I'm quite pleased with the result for my own website.
The examples themselves... need some work still. Some excuses/caveats:
- They started life as bunch of little junk files/experiments, and they kind of still are.
- They're generally intended to be very minimal, illustrating one particular trick while not doing other things (even things that arguably you should always do) although I haven't been consistent with this.
- Some of them are examples lifted directly from tutorials or example resources on other websites e.g. the OSM wiki, and I've deliberately not changed anything (even where I'd like to) so we can cross-link for consistency.
- Giving credit to OpenStreetMap is another area I may not have been entirely consistent with. In general I would say that examples like this should really follow the instructions on the copyrights page to the letter. In many examples I'm doing the minimal thing, and sticking with defaults. In particular note that the OpenLayers developers have not rolled out the default OSM credit text change to the hosted OpenLayers yet. Not sure when that's on the roadmap for. I don't think we need stress about it. The other thing is, I have been meaning to put together a set of examples/code/graphics for giving really awesome credit to OpenStreetMap. Alex Barth beat me to this idea recently. Basically the same idea, though I was imagining it slightly differently.
All said and done, there's a fair amount of wrongness in my examples, and I guess you can contact me if you'd like to suggest fixes. They're far from perfect, but taking on board everyone's feedback might count against simplicity. And besides we have the wiki for that. It's quite interesting to look at what's happened to the 'OpenLayers Simple Example' wiki page which I created way back in September 2007. At the time I predicted problems, and I wrote a section at the bottom begging people not add stuff to it which makes it more complicated. Since then various people have done just that, but there have been a few nice simplification edits too. I think a bold edit or two would whip it into shape, but maybe only in line with the way I see it, and only until somebody else edits it because they see it differently. Providing simple examples is not an exact science.
Discuss at the pub TONIGHT!? We're at the Blue Posts near Tottenham Court Road station from 7pm onwards (possibly hidden upstairs!)
We just had the London OpenStreetMap Hack weekend, and I think I'm right in saying this was the biggest OpenStreetMap hack weekend ever!
Big thanks to the Open Data Institute, who provided the venue, and also sponsored the event. The ODI are funded by the technology strategy board to promote and support open data initiatives, and there are surely non more deserving than our London-born global not-for-profit project. However the ODI are particularly keen on supporting innovation which drives UK economic growth. I'll be writing on the ODI blog soon, about ways in which OpenStreetMap does this (UPDATE: blog post on the ODI site now)
The ODI is also... my office. I work every day at a desk there doing placr.co.uk open transport data stuff. It was quite exciting holding an OpenStreetMap hack weekend at my place of work, but it did mean I had to get up early to be the first one there.
On the Saturday it felt like things got off to a slow/awkward start. I think this format could be improved by somehow unleashing people (those who need inspiration) on some simple hack ideas, but then maintaining lists of simple hack ideas presents a problem for developer engagement in general, which we didn't really solve yet.
I think it's true of any hack weekend, that you can get more out of it if you sneakily come along with some already-underway hack idea. Personally I had quite a few ideas, though non already underway as such. I started by working with Dan Stowell on merging in a pull request to XAPI URL Builder to add a 'Live Query' tab, which is pretty nice.
I then tried to order pizza online, and found myself having to hack the HTML form of the pizza website in order to let me submit without filling it in again. Pizza hack!
After pizzas we had "workshops and presentations". This worked pretty well because the event was well enough attended to make it worthwhile. Having said that, I was hoping to give the event broader reach, with more "OSM outsider" developers coming along to learn more. I gave my opentech presentation on the basics of the OpenStreetMap developer ecosystem with such an audience in mind, but to an audience who didn't really learn much from it. But hopefully at least if gave some ideas for how to explain OpenStreetMap.
Next Andy Allan gave a workshop on using TileMill. His enthusiasm for cartography rubbed off on a few people, and quite a lot of us installed TileMill and had a play over the course of the weekend. It's very easy to get started with shapefiles (e.g. osm shapefile downloads). I think Dan Stowell got as far as getting the openstreetmap-carto style running (this requires PostGIS and osm2pgsql)
Lots of other hacking went on besides. There's a list of stuff which got hacked on the wiki. With so many people attending, we found it was quite natural to gather people round for a presentations and show & tell session on the Sunday afternoon. As I've said before, the real benefit of these events is in sharing ideas, but in the past we haven't had sufficient people to create the audience-speaker dynamic. I think it's really great that we managed to do it this time. Thanks to the cyclestreets guys for kicking that off with an update on their work.
Having said that, the most exciting hacking was by TomH who was keeping himself to himself, while he progressed the 'Notes' branch. Having 'report a problem' on the OpenStreetMap homepage will be a signficant change when it happens.
We also awarded a bottle of Yorkshire beer to Ben Abelshausen, who was over from Belgium for the weekend. He gave a very interesting and impressive presentation on his system 'OsmSharp', a routing system for solving travelling salesman-like problems which he has been working for his masters thesis, but also as a project for a company who are now actively using the software to plan delivery routes. Ben is now hoping to do OpenStreetMap consulting work around these kinds of ideas.
Oh and we all went to the pub ...three times
The weekend kicked off on Friday with pub drinks. This was the convening spot for some of our international visitors, and we had a good chat including spending a long time puzzling over how to design a ReCAPTCHA-like system which gets people to solve a useful task for OpenStreetMap, while also (and this is the difficult bit) functioning as a bullet-proof captcha. I had a nice idea involving categorising aerial imagery tiles, but it didn't present enough answer permutations to be uncrackable through brute force automated requests. Other ideas were too difficult for human users to succeed at quickly. Tricky.
On the Saturday we had a bigger pub session. SteveC generously gave us £50 to get some beers in. It was such a big gathering that this stretched to just less than the first round of beers! The fox is nice, but a bit pricey. That certainly goes for the food too. Clearly I haven't done my duty and sussed out good cheap pubs around my office, because on the Sunday we resorted to the whetherspoons in Liverpool Street station (which I quite like)
After three nights of pubs and two days of hacking I've been feeling pretty exhausted, but it's such good fun, and triggers so many ideas, I hope we can repeat it again some time soon. Thanks again to the Open Data Institute for hosting and sponsoring.
No rest though. Thursday night is #geomob night at the BCS on the Strand, where there'll be some OpenStreetMap folks attending and some geotalks featuring OpenStreetMap (Dan Karran's bombsight.org for sure. I'd guess also Simon Willison's talk). For more OpenStreetMap events Follow @OSMLondon on twitter
Two big events coming up in London:
TODAY (Saturday) we're having a big lunchtime pub meet up. hbogner is visiting with a whole team of Croatian OpenStreetMappers. Nine of them I think! They've been doing some kind of OSM workshop in rural Wales, the bit which just got the heaviest snow, but I'm sure that wont stop them getting back to London this morning, and we can meet with them to drink beer in the traditional english environment of Ye Olde Cheshire Cheese pub.
I'm sorry I've arranged this at rather short notice. I was umming and aahing because it clashes with something else I have this evening. This is the only reason it's a lunchtime event actually, so if people want to carry on the merriment into the evening, or arrange a follow-on session at a different pub, I'm sure the Croatian folks won't complain! I will be aiming to get there before 1pm though. Should be fun. For a while there I wasn't sure if the Croatians had managed to find out about the event I'd set up. But it's OK. They're good to go. But I seem to be the only Londoner signed up on lanyrd so far! Looking forward to it anyway.
The other big event is the next London OpenStreetMap Hack Weekend. Feb 2nd 3rd. This looks set to be an international gathering too, with Emacsen flying in from the States, Ben from Belgium, and Katie from Germany. Again we need more Londoners signing up, but I know you all don't like to commit until the last minute. The Open Data Institute are hosting and sponsoring! This will be a great venue for this I'm sure. It has "Open Data" in its name, and it's a nice space. Check it out (we can get set up in the 'boardroom' there). Please consider coming along even if you're a little bit scared by the word "hacking". It'll be fun I promise!
So those are some upcoming events.
We were in the pub not so long ago for our first London meet-up of 2013 at the Monkey Puzzle. The pub was unusually crowded. Clearly after their spectacular switch2osm on their website, more people are managing to navigate to this pub, or perhaps just attracted to the pub due to the sheer classiness of its website map. Anyway it was busy, but everyone who got there before me did a fine job of grabbing a good table, this time in the room with ye olde map on the wall.
Ed Freyfogle from Nestoria was there, so we got talking about the property market, and the business of cleaning up data coming from estate agent listings, and the business of click-through money making with zoopla. I share an office with a startup called Locatable who do this too, and in a very map-based manner. Don't worry. They have promised me they will switch to OpenStreetMap :-)
Richard Fairhurst was in London for the day, so of course he should swing by the OSMLondon meet-up! He and Andy sounded like they were having an interesting chat about the future of OSM editor development. They worked together on some of the latest developments of Potlatch 2 in flex, and there's lots more that could be done there to bring some of the advanced features to their full potential, but it seems like iD is the future.
I was showing off my new christmas pressie, a Kindle Fire. It's a very capable tablet device, somewhat hobbled by amazon's custom android O.S. The homescreen interfaces are nice to use, but it very cleverly directs you buy stuff from them. Mostly annoyingly though, you can't get on "google play" to download whichever apps you like. The amazon appstore has a much reduced subset. Searching for "OpenStreetMap" finds a handful including Skobbler's Forevermaps, and MapsWithMe.
Now Derick has been recommending MapsWithMe for a while, but I only just got around to trying it at Christmas time, and I have to say... it's awesome! My new favourite app. I only have the free 'lite' version, but it's a really nice fluid offline vector map browsing interface. Grant was noticing that they're doing very well on download stats, including doing really well at getting people to pay £3.09 for the pro version. Making money from OpenStreetMap? Whatever next!
We talked about the evils of imports, the France Cadastre debate again, and closer to home, the practice of copying in street names from Ordnance Survey open data. I'm a little suspicious of the progress we're seeing on ITO's street names comparison map now. Everything's turning blue, which is good news, but are people going out and surveying these streets? If we just copy in names from O.S. open data "we're in danger of doing ourselves down" was how RichardF puts it. We should aim to create a richer more exciting dataset with a community of local OpenStreetMap enthusiasts maintaining it.
I really like Pascal Neis' Overview of OpenStreetMap Contributors map, which adds the little user icons (they look like babies) based on the centre of their active areas. I was saying in the pub, that in a way this is more like what the users home location map should present. A view of who is nearby, and genuinely active there. The home locations map has a weird kind of clutter problem which means it doesn't really help connect new people to nearby mappers.
We talked about landuse mapping. The confusion created by TimSC's crude mapping approach of blatting in areas of landuse based on looking at aerial imagery (and often getting classifications completely wrong). Also we discussed the general problems of landuse classification. Commercial vs Retail vs Industrial. Areas tend to be a bit of both. Edge cases are more common than not.
Speaking of landuse mapping "Rendering of Farmland not 'Light' enough?" Yes! Longtime bugbear of mine. We should make the big brown stains fade to the background a lot more in my opinion (on the 'standard' style)
Speaking of the 'standard' style, we had Steve Chilton in the pub with us. Lots of conversations about rendering including Andy Allan chatting to him about his carto port of the OSM stylesheets, and how Tilemill designing would work if we were to adopt this on the main tile servers.
Speaking of the main tile servers, we talked about the "operations" side of things, and the vision for scaling onto new servers for load balancing and/or failover. We've got quite a few tile caches now, and these are starting take the vast majority of the load, but they're pointing to one "parent" server doing actual rendering. Yevaud is churning out an average of 6 metatiles per second. OSM U.S. may be providing some servers which could be arranged to split the rendering load, but then the caching approach becomes more complex if we try to avoid two servers rendering the same meta-tile.
As well as my kindle, I also got a new bike for christmas, so I was showing that off too, but we didn't hang around too long outside the pub before heading home, because it's mighty cold this time of year.
Today we're having a spot of snow, but don't let that put you off. The Ye Olde Cheshire Cheese is a warm cosy pub. Just the place for an OpenStreetMap lunch! All the details on the wiki
We had an awesome hack weekend in December, but last time we had a regular pub session was way back in November. We went to the Parcel Yard pub in the new bit of Kings Cross station.
In the pub we talked about ...random topics which I wrote down at the time. Does anyone remember what we said about "floods in the tube", and "airoplane radars"? And at the end of the list I have written down "weight loss solutions". I don't remember that converation at all. Maybe we came up with some brilliant medical breakthrough, but I'm afraid I didn't write down any more detail. The cure for obesity is lost forever.
We talked about OpenMeatMap.org, a map locating fine meat produce from butchers, farmers markets, and other meat outlets. ...which sadly doesn't exist. But it should!
We talked about MapRoulette.org and fixing up TIGER mess. (Oh that's cool. I've just noticed my little tutorial video is linked from there directly now)
Someone (don't remember who) was browsing the OpenStreetMap.org front page map (newly converted to leaflet.js) using his android phone and we observed a bizarre squidgy dragging effect which has never been seen before or since. You have to see this little video to see the squidgyness. It was a pretty cool bug in a "you couldn't do that if you tried" kind of way.
We also had a fairly extensive discussion covering some interesting more serious topics:
Back in November the French cadastre import of building data (Import semi-automatique des bâtiments) was causing some agro because the Data Working Group had blocked some french users who were bulk importing without following the guidelines. Many French users are taking part in the 'semi-automatic' merging in of building data. Some users took this too far, importing too many objects too quickly. In these cases it was a bulk import, but without following bulk import rules (in particular not creating a separate user account for this). Blocking users is always liable to trigger some agro, but somehow this escalated into a French community versus DWG battle on the mailing lists. Matt (of the DWG) was giving his perspective on this in the pub. Here's a thing I noticed from a wiki perspective (I always like the wiki perspective) Looking into the debate after it flared up, I was surprised to find that the french language wiki page about this import carried no cautionary information, about it not being intended as a bulk import, and about the problems of bulk imports. The warning on there now was only added when I pointed this out. Some users will get the wrong idea if you don't give guidelines and just say "hey everyone join in with uploading this data".
We moved on to talking about the OpenStreetMap foundation, and particularly the Management Team. The MT group lost steam a bit. We discussed reasons for that. MT meetings have always been voice calls, rather than IRC meetings, which some people prefer, but voice calls can feel like very hard work sometimes. More fundamentally it was felt MT doesn't have enough authority (or doesn't feel itself to have enough authority) to make decisions, and isn't properly looped into a the chain of command. Personally I think the management team would work better along with many other working groups, if we could generally make these roles and responsibilities more desirable coveted positions somehow. We, at the OSMF, need to figure out better ways of recruiting people from the wider project. We discussed this because Matt (OSMF board) was organising a re-boot of the management team.
I realised we were giving Matt a bit of a grilling in his capacity as board and DWG member, during these conversations which were unusually serious for the pub. But we made it up to him by going for his favourite... burritos!.
See the majestic timetable of Kings Cross station in the background there. No? Well ok. But you have to admit the roof is quite nice here:
The Parcel Yard pub is kind of hidden at the back of this new bit of King Cross station, and has none of the skankiness you often get in station pubs, perhaps because the beers are a bit pricey. Quite nice pub though. And good proximity to burritos. Perhaps we'll go again.
But not tonight. Tonight we're going to Paddington, and no the station this time. We'll be in the good old Monkey Puzzle. All the details on the wiki. If all of this serious and not so serious OpenStreetMaptastic conversation sounds like your kind of thing, or you're just curious about the basic idea of OpenStreetMap... if you are one of the 1 million, or even if you're not... come join us in the pub!
1 million users! Woosh! Welcome to 2013. This is going to be a good year.
1 million is a big number. My uncle is a retired school teacher, and he has this great story. He told a classroom full of school kids "1 million is a big number. Between all of us, d'you think we could draw one million dots?". He showed them how to draw a block of 100 dots on a piece of paper in a neat 10x10 grid, then they raced to see who could draw 1000 dots as a row a 10 of these blocks. Some kids did it very quickly. Some kids took a bit longer and did them in colourful stripes. "Not bad" he said "If we add them all up, working together, you've drawn a lot of dots, but if you lot can draw 1 million dots I will eat this banana whole, including the skin!". Excited by this prospect, the kids all feverishly carried on drawing dots, but at the end of the lesson they added them up and found they had only done about 200,000 dots. Good progress, but they had a long way to go to reach a million. Lesson learned. 1 million is a very big number.
...but a couple of girls went home that night and carried on drawing dots. Two months later they presented my uncle with a box piled high with dots on paper all carefully arranged in colourful checkered grids. My uncle (now retired) showed me this box which he had kept ever since. It's a true story.
...and yes, he had to eat a banana whole, including the skin.
I'm not sure what the lesson is from this. Maybe "never underestimate the power of bored kids", or "if you ever thought editing OpenStreetMap was a pointless waste of time, people do more pointless things", or how about this lesson "A massive task might just be possible if you have lots of people to helping out and a few very dedicated people getting really into it".
Anyway 1 million is big. The above graph is from osmstats.altogetherlost.com. It shows us blasting over the 1 million user mark some time in the small hours of yesterday morning. This morning's raw stats report clearly show us over the line. The graph shows a downturn at the beginning of the week, which isn't usually possible in the count of registered users. This is caused by the sysadmins doing a clear out of some spam user registrations. They've done this before in the past, and they wanted to do it again in advance of the 1 million users announcement.
Whenever we do announcements about the number of users we have, people are quick to complain that this is misleading. This number includes people who signed up, but didn't make a single edit to the map data. That is no small component of the total, in fact it's a very large component. About 2/3rds of signed up users never get as far as clicking the 'save' button on a map edit. 2/3rds! This is based on Pascal Neis' analysis of this from a couple of years back, but I'm assuming this ratio hasn't changed much.
So we might say a less misleading figure to look at, is the number of users who made a least one change, but this quickly leads to the next question "Should we really count users of OpenStreetMap who have only made a single edit before disappearing?", and then we're into different ways of measuring activity. People have done this analysis. Andy Robinson has been re-plotting this graph since the early days, showing a "rolling week" measure of node editing activity, currently fluctuating around the 8000 mark. We have to measure the size of a community somehow. There's no doubt this 8000 figure is a more down-to-earth realistic figure than the over-blown 1 million figure we're singing and shouting about. But why take a 1 week rolling count? Why look at node editing? There's lots of different ways of aggregating these stats, all with different quirks. The fact is, the least arbitrary measure we have available to us is the cold hard simple count of user sign-ups, and today this is showing a big fat 1 million! Yes we're singing and shouting about it. Guess what? We're trying to promote OpenStreetMap. Nothing wrong with that, so please join in!
But the revelation that 2/3rds of sign ups do not edit, is a frustration for people who work on improving OpenStreetMap. I like it when the discussion comes around, because it leads to inward-looking self-criticism. What are we doing wrong? I would urge people to think of the zero-editors as part of the long-tail distribution of our community. I'll explain what I mean some other time, but basically we shouldn't beat ourselves up about it too much. But self-criticism is good. It leads to conclusions like:
- OpenStreetMap editing software needs to be easier for beginners,
- Documentation needs to be better for beginners
- Editing on mobile needs to be easier
- The website needs "social" features for welcoming and helping new users
- We need simple bug submitting and workflows around fixing them.
- We need to make contributing fun, like a game, with microtasks.
The good news is that we're tackling things on all these fronts. We saw some great progress in 2012, and we're set to see exciting announcements in 2013. In particular...
- We may soon go live with a "Report a problem" link on the OpenStreetMap homepage. 'notes' branch
- In 2012 we've seen the PushPin OSM app brings easy OSM editing to iPhone.
- In 2012 we've seen MapRoulette challenges breaking TIGER fixup in to interesting little microtasks. Now we're seeing the 'Kort' presenting KeepRight bugs as a fun little game.
- switch2osm.org launched at the beginning of 2012, providing targeted documentation on using OpenStreetMap for web developers. RichardF was the main man behind this, and he's itching to do something similar again. Meanwhile learnosm.org is set to receive a facelift very soon.
It's things like this which will improve our "conversion rate", moving people from signing up to making edits, and allow us to really tap into our 1 million strong army of registered OpenStreetMappers. That's why 2013 is going to be a good year.
As usual the run up to Christmas seems to be building up to a crescendo of chaos for me. But that's the way we like it. Some of the Hack Weekend promotion was a little hurried/non-existant, but even so we did alright.
We had a lovely venue courtesy of 10gen the MongoDB company where Derick works. Our venue requirements are pretty simple, but this funky warehouse office in Shoreditch surpassed them. Thanks to 10gen for that, and to Derick for helping me run the event.
And we had pizza!
I arranged for pizzas and other snacks courtesy of placr.co.uk, the little transport open data company where I work. I wrote a post about this on the placr blog, and about some ideas for railway data hacking. One of my hacks for the weekend was to tweak the replication diff updating scripts to try to make this run robustly on the placr server. I'm not as expert at running download servers as the geofabrik guys. Hope it doesn't become too much of a pain in the ass, and I hope that I and other developers will be able to chain some other railway ideas onto it.
Speaking of OSM railways data, we got chatting about TrainLord, which is pretty awesome use of this data. It's a game. It's not that easy to get started, but I will admit I found the idea of lording it over the East Coast mainline to be quite appealing, and so I've spent rather longer playing it that I should have done (which is probably what makes a good game) and unfortunately discussion lead to playing TrainLord instead of hacking at least some of the time!
But apart from that, there was some hacking going on (This list is also on the wiki page):
- TomH worked on Notes (formerly known as OpenStreetBugs) and converting it OpenLayers -> Leaflet. See activity on github
- Robert Scott showed us a tool he's been building which process GPS traces and compares routes with those from OSRM to discover where streets seem missing or oneways are wrong (assuming it's trace taken while driving) [UPDATE. It's described on the wiki now]
- Derick worked on his "year of edits" video animation, and made a MongoDB powered XAPI-like system to help power Tom Morris' thing
- Tom Morris worked on osmcheckin a foursquare-like checkin system for OSM POIS
- Andy Allan worked on openstreetmap-carto a port of the OpenStreetMap 'standard' stylesheet into the more compactly elegant 'carto' stylesheet format.
- Adrian nw3biz learned some OpenLayers markers tricks following my marker popups example
- I made some improvements to XAPI URL Builder to make it output osmosis commands for data filtering, to offer an alternative to hitting XAPI.
I did that hack quite quickly, but spent the rest of the time being distracted by TrainLord, taking pictures of myself eating pizza, and watching nice previews of hacks other people were working on. A hack weekend really needs to involve people feeding ideas off eachother, and there was plenty of that going on. All in all, a pretty fun way to spend a weekend!
I hope to do it again soon. In fact we'll probably get something together for January or February. If you know of a good venue, and/or a company who would like to sponsor the pizzas for this kind of event (or a different kind of event!) let me know.
We've got quite a range of London events upcoming in various stages of planning:
Thu 22nd - geomob - They've just announced the line-up. Six different geo-presentations followed by pub. This is always a fun geo crowd, and usually a fair few OpenStreetMappers in attendance.
Sat 24th / Sun 25th - London branch of Operation Cowboy - This one is being organised by Tom Morris. He's yet to confirm the venue, but has a plan. I hope this can be a good "training" event, at which people can go and learn or share tips about OpenStreetMap editing (That should be the focus I think. Not so sure about the "cowboy" ideas) But it's not my event! Somebody else is organising something. Hurray! For that reason alone, I'd encourage everyone to support it. It will need experienced mappers and of course lots of inexperienced mappers to sign up to it. Tom has been promoting it to the wikipedians, but there's probably quite a few other people who have been waiting for a non-pub related non-hacking mapping related OpenStreetMap event. How to get the word out to those people? I've pondered training kind of event formats for a long time. Hope it works well.
Sat 1st & Sun 2nd December - Hack Weekend. The dates are not necessarily set yet, but this seems like the most likely weekend. It means it'll clash with Sanitation Hackathon (a humanitarian tech event), but it's the best weekend for OpenStreetMap hacker availability. If you're within range of London and interested in the technical side of OpenStreetMap ("internal" development or just using OSM data) then keep this weekend free. More details to come very soon hopefully.
And finally the Christmas Party 2012. We've got a wiki page so far at least. We should probably pick a date.
Phew! So thats the upcoming stuff.
The last London winter pub meet-up was at the Blue Posts. We had a small gathering but with a couple of new folks.
Aimee was along to learn about OSM and get ideas for possible projects. She's looking to do something map related for her MSC. We discussed "circular routing", choosing a route to go on a circular cycle/walk taking in some sights before looping back. The cyclestreets guys have done some work along these lines.
We also had Jack along. He was investigating the various ins and outs of running your own tile server (The kind of things introduced here on switch2osm). Andy was able to give him plenty of pointers. In particular he was asking about memory and other requirements when renting virtual machines, which is something Andy knows all about from running OpenCycleMap and transport map.
Also here's Andy with a banana:
We went for burritos later, and I forgot my bank card behind the bar, so we just had to go back and have more beers after that. No choice!
So tonight we'll be at the pub again. All the details on the wiki. We'll be in the Parcel Yard pub in the new bit of Kings Cross station. Something a bit different. Easy access for people entering/leaving London on the train! Come join us at this and/or any of the other events mentioned above.
The first London winter pub meet-up is TONIGHT. We've rolled over into "winter pub meet-up" mode, since the clocks have changed and the evenings are properly dim and dismal now. So join us tonight for... no mapping just beers!
Three weeks ago we had a "Summer" mapping evening. I got off a crowded tube train at Old Street and found I'd lost my android phone. Either dropped or pick-pocketed. So as I went outside and tried to gather some map data in the gathering winter darkness and pouring rain, around the stark tower blocks streets of St Luke's area, I gently sobbed to myself (*)
Imagine how my spirits were lifted as I arrived at the Wenlock Arms pub, to be greeted by a table full of OpenStreetMap friends, a wonderfully cosy atmosphere, a fine array of real ales, and a phone call to say "I've found your phone on the tube. D'you want to collect it tomorrow night?"
More OpenStreetMap friends arrived. We only had one little table but maybe 10-15 people! The Wenlock Arms is a small pub, but we were able to spread onto quite a few seats, plus a lot standing at bars etc, but still feeling like a nice cosy OpenStreetMap gathering. We also didn't really fit into one photo. This was our best attempt:
It was great to have Ed Freyfogle along, chief of Nestoria, chief of #geomob, and newly crowned foursquare mayor of the pub (he's a regular) And Matt was along too. Here you see his hair has grown back after the stress of the license change, but the french community have piled on some more DWG stress recently, so he's probably lost it all again :-( But this was 3 weeks ago, so Matt and Ed are laughing jovially ...and there's me:
I didn't take notes of our discussion topics this time (or if I did, I lost them) but I'm pretty sure we discussed the OS Locater musical chairs tool since ris was there. And with Ed we discussed Nestoria's uses of OpenStreetMap, and problems with tube station rendering (they're not shown prominently enough by MapQuest or osm 'standard' style) . Shaun was down from Ipswitch, and we even had Edward Betts back in London after several years away. In fact Edward was saying he attended the Map Limehouse event. 2005!
The Wenlock Arms pub is great, not least because they use OpenStreetMap on their website (a little bit), plus the fore-mentioned cosiness and ale selection, but one fatal problem is the lack of food options. We held out for quite a while, but eventually burritos called, although we left a small OSM team behind to do the pub quiz!
We probably would have had a pub meet-up a week ago, but that would have clashed with w3g conference. There were some OpenStreetMappers mingling in the geo-crowds there. And some OpenStreetMappy discussions. Personally I went a bit overboard on the pre-conference geobeers, so was hungover for the whole thing, but it was fun. Maybe got some people interested in coming along tonight?
Join us tonight at the Blue Posts on Newman Street from 7pm! Details here. I've given up trying to predict whether we'll get a big turn out. There's no logic to it. Could go either way. But I've always found small pub meet-ups are just as a fun as big ones. In theory we could look at how may people signed up on lanyrd but people don't all do this (c'mon! at least "track" the event!) But hey, four people so far. That aint bad.
(*) This is just dramatic storytelling. I didn't actually sob.
I was so hyper-enthused with a glowing sense of geotastic-all-conquering pride after last week's London pub meet-up, that I went ahead and set up the next pub meet-up. Yes. We've had the details on the wiki page two weeks in advance. How very well organised! Do come along and join us on Thursday 11th in a pub over in Hoxton.
But what was I so pleased about at last week's meet-up?
The Monkey Puzzle did a switch2osm! It was a sort of live-action-laptop-over-the-bar switch2osm.
I've chatted to the landlord, Gary, a few times before. He knows about those crazy guys who come to merrily drink with an "OpenStreetMap" sign on the table. And when I said they should switch from google maps to OpenStreetMap on the Monkey Puzzle website, he said "Errr. Yes... yes we should", and after some confused behind-the-bar discussions about who knew how to update the website and what that would involve, he went and fetched his laptop.
This was highly amusing to me at the time, but then I had had several pints. Presented with a very slowly loading wordpress admin interface I didn't try to do anything too fancy, but swapped out their google maps iframe and put in an OpenStreetMap one. Boom! switch2osm live and direct!
Check out the new themonkeypuzzlepub.co.uk contact page! The Monkey Puzzle was our favourite pub before, but this makes it the greatest OpenStreetMappiest pub in the whole of London for sure.
What else happened? As well as that, we had some good beer-fuelled map-chat in a tight circle, with not many people showing up (6 people). One of these was a new chap. Someone called Vin, who I had come across via n0tice. He's big into n0tice, and is using it to devise a concept he calls "pubmapping" involving stream-of-conciousness photoing and 'reporting' from a pub... or something. We tried it. Here is a our set of iPhone photos and other nonsense which I have "reposted" onto the "OpenStreetMap noticeboard". See how that works? ...no. Me neither. I did show him how we do pub mapping, using JOSM to highlight the pubs data gathered in the Paddington area. As an aside, on IRC yesterday ris pointed out this awesome changeset by a brand new user. Now that's pub mapping!
Vin explained that he was interested in capturing how a pub makes you feel, and this got us onto a conversation about "subjective" mapping. We try to avoid putting subjective data, such as ratings of pubs, into OpenStreetMap. Other similar ideas pop up all the time with greater or lesser merit. The H.O.T. list has debated whether measures of the risk of a disaster can be captured on a the map somehow, e.g. where a local resident anecdotally reports that an area is at risk of flood or landslide. All worthwhile data to capture somehow, but maybe not using the OpenStreetMap database directly. We could do with more tools like ushahidi which let people layer data on top of OpenStreetMap. Ushahidi handles data on temporary situational reports, which mostly wouldn't belong in OSM directly. Subjective data could be captured in similar (but not the same) tools.
We also got talking about radiation leaks and aeroplane crashes :-O
Oh and Grant was showing off his new head mounted camera
By the way I did also do some mapping beforehand, though I think I was the only one, and I was running late so just very quickly swung by on my bike and got a bunch of very blurry photos in the gathering darkness (it's getting to that time year again). But this was enough to check a few things I wasn't sure about while armchair mapping the building outlines. For example if you look at the bing imagery around this building it's not obvious whether it's attached to other buildings nearby or not. A very quick check on-the-ground makes for better data.
So we're going to Hoxton on Thursday 11th, all the details here. We'll be at the Wenlock Arms. Why did I choose this pub? The Wenlock Arms has been threatened with closure, with various outrageous plans to turn it into ...something other than a pub. Always sad. But actually we're continuing a switch2osm theme! wenlock-arms.co.uk has been linking to OpenStreetMap for a while now, so it's time we go reward them for this, and show our support by drinking a few beers there (That's the deal by the way, if there's any pub landlords reading this... switch2osm, and we'll come drink there!)
Also the pub is reasonably well placed for a mapping excursion. The area around Hoxton Square, further to the East is probably higher priority actually. (We really need to finish off the Shoreditch "tech city" area with building outlines. Individual mapping missions encouraged!) Around the Wenlock Arms the buildings have been filled in, but maybe not with a survey. not sure. Some more footpath details could be good, plus I see some buildings missing, so I've laid the cake slices out accordingly.
Join in with that if you fancy it, or just join us for a social meet-up, come along to meet and learn more about OpenStreetMap!
As I get psyched up ready for a London mapping session this evening (Listening to "Eye of the tiger". That sort of thing) I realise it's been an interesting few weeks...
Last week I went to Amsterdam and presented OpenStreetMap and H.O.T. at the PICNIC festival. Here's a video of the session. There's another interview video still to come I think. I'll blog some photos and more details about that. UPDATE: blog now here
Last Thursday I had been sprinkling a few comments around the web, on press coverage of Apples debut of their iOS6 maps. iPhone users the world over, were loudly complaining about apple's switch from google maps. This really shows how strong the big G's dominance of web & mobile mapping is. Phone users and also tech-savvy app developers are insistant upon google maps. It's hardwired deep into their psychology now, such that forcibly presenting a different map causes great upset. This is what OpenStreetMap is up against, but last week apple swung a battering ram at google's mappy castle. They may be annoyed but users and developers being liberated. They're experiencing a realisation that there is a choice of map providers. It's only a short leap of logic from there to go seek the open alternative. That's us! Over here!
So after spending the whole day pointing out what an evil grip google holds on people's map preferences (slagging off google essentially) on various forums, I half expected to be refused entry the google campus building for Thursdays #geomob. Maybe I thwarted them by turning up late. I caught the end of Lawrence's talk showing various old maps, and concepts for maps, particularly looking at "linear maps". Anyone done any more OSM experiments on Linear Maps to list on this wiki page?
Then "recce" guys presented their map. It's stunning, but be warned, it's not OpenStreetMap (boo). Actually if you want an OpenStreetMap-based similar thing, UK travel options has a similar 3D zoomy twisty finger swiping experience (and UK wide), but admittedly not as gorgeous as Recce's semi-realistic cartoony look.
In the pub afterwards I had a good chat with Bart Denny of recce, formally of google maps. He said he liked the idea of OpenStreetMap, but what's a good mobile app for contributing easily? I had no convincing answer to this question, and he was quite forthright in his criticism of OpenStreetMap for failing to get this sorted out. Now admittedly I'm not an expert on all the apps available (who is?) but I had to agree with him, that we really badly need better mobile apps for contributing to OSM.
He had a theory (which he also stated in relation to Recce, and his previous work at google) that map data gathering is about to become very smartphone-based and that the problem of building a complete and accurate representation of the world will largely disappear. Phones are already powerful enough and widespread enough that map providers could get their users to capture data with a simple panoramic sweep of their camera (and some very clever optical recognition) Who owns that data is still a question. If OpenStreetMap can stay in the game it can eventually win out by offering it as open data. But we have a little way to go before that!
We have an opportunity to grab some attention of iPhone users who are looking around for better map apps, but it's an opportunity we're missing a little bit because our list of iOS apps is a bit confusing. We're listing some quite old apps, some which don't work any more, some which have nothing to do with OpenStreetMap except that they can be used for recording tracks. It's quite difficult to find the gems in there. Those doing interesting things with offline data or vector data (the OpenStreetMap advantage) or those with editing capabilities. A lot of the wiki pages need a bit of a rewrite, sometimes with obvious formatting / info from the appstore (see Wiki Maintainance Tasks ), ideally by actually installing the app to test what it actually does. Who wants to help?
I started doing this and happened upon DirectU (now with some wiki description). It's a free app which gives me the whole of Europe (or other regions) for offline viewing. Rather like MapDroyd but working on ipod touch. Nice!
We're also not doing too well at documenting and promoting the "upgrade path" for iOS developers to use various libraries for embedding OpenStreetMap. There's a list of options here, but maybe we need to compose some new pages for switch2osm.org on this topic. Lots to do!
But for now... I need to go to the pub! No doubt we'll chat about some of these issues while eating the nice steak pie they do at the Monkey Puzzle. If you're in London, come join us (TONIGHT!)
What's been happening lately? Oh yeah...
We've finally changed the license! Hurray!
On the foundation blog I've been posting quite a lot about the license change process, as we in Communication Working Group have been working to ensure that this channel is a useful place to get "official" (carefully worded) updates on this sort of thing. Obviously in this capacity it's important to be positive about it, but I'd just like to say from a personal perspective...
We've finally changed the license! Hurray!
The best thing about this is... we don't ever have to talk abut the license change ever again! No doubt there'll be lots of talk about the new license, but the fact is... it's done. The license change sucked so much energy from the community and particularly from people who are core to making things happen in OpenStreetMap. So many ideas have been on hold while we battled with this thing. Thank goodness it's over. I really need to go to the pub to celebrate (more on that later)
The Society of Cartographers conference also happened. This was maptastic, and quite OpenStreetMappy. In fact the very first talk was from David Earl (Freelance) - Project Drake: new maps for The University of Cambridge using OpenStreetMap. Steven Feldman talked about OSM and a bit about OSMGB. I blogged here about my little workshop on using OpenStreetMap data While I was doing that Andy Allan was doing a workshop on TileMill, which I wanted to go to. And on the third day as part of the ICA neocartography conference we had Richard Fairhurst presenting the Unstoppable Advance of OpenStreetMap including the awesome neo Steve Chilton slide:
Lots more map stuff besides, also chocolate muffins, and a tour on an open top bus. It was a pretty great conference. Not as great as State Of The Map obviously. Would've loved to be in Tokyo. In October I wont be going all the way to SOTM U.S. either, but SOTM Scotland could be a goer!
The vote for the board happened. I didn't actually vote in the end. Couldn't make up my mind. Some U.S. representation on the board would've been good. Kate would've been great on the board, though she's already very busy being great for HOT. Alex could've brought some of that shiny MapBox magic to the core of the project. But congratulations to Frederik and Simon. They're both very well known as massive contributors to the project, and people with strong opinions and powerful arguments about how the project should be run. They'll do great things in the driving seat I'm sure, but also this will maybe help make more German and Swiss OSMers feel involved in the OSMF a little more (e.g. joining the working groups) which could be a great boost.
I'm in Amsterdam next week for the PICNIC conference where I'll be presenting OpenStreetMap and HOT.
Also next week on Thursday it's #geomob, this time hosted at the google campus. Dunno if there's any google related geo presentations lined up. Time for some heckling? I'm sure it will be a geo-treat as ever, and will involve a pub.
All of this means that, despite my desire to have a good OpenStreetMap pub celebration, I'm not very free until the week after. If somebody wants to organise a OSMLondon meet-up before that, next Tuesday could be a good day, but I'm coming back from Amsterdam late that day. Whatever happens the OSMLondon event details are on the wiki page
Thanks to those who came along and helped me celebrate my birthday with a big BBQ again. Thanks especially to Alex who crafted a custom rendered map birthday card for me:
That was a couple of weeks ago. Time flies, and I've been a bit slow to organise stuff, but THIS SATURDAY it's already time to for a birthday celebration for OpenStreetMap!:
So far we're partying in Toronto Canada, Cleveland Ohio U.S., Lisbon Portugal, and of course London! I've just picked a pub, trying to think of one with some scope for enjoying sunshine and/or views of the river. The good old Doggets on the South Bank. As we're partying on a Saturday, hopefully lots of people can make it, even if you're travelling into London from afar. Do it! You know you want to! And get in touch if you want some cheap couch surfing accommodation. Add yourself on lanyrd to make the event look more popular. As ever I'm hoping we'll get some interesting new folks, including maybe people who are just a little curious about OpenStreetMap. Who should we be inviting along?
And, as I often seem to accidentally prove, you can organise an OpenStreetMap party at quite short notice, so for those elsewhere in the world, now is the time to add your city to the list and help make this a worldwide party! Which german city will be the first to set something up??
A few weeks back (Thu July 19th) we had a little London Summer evening mapping session around Chancery Lane and a pub meet-up at the Penderel's Oak. I probably should've written about some of the conversations closer to the time because the timing was interesting, but things have moved on. We were right in the middle of the redaction bot run at that point. Up until the the day before, I had been assuming that the redactions wouldn't have reached London by this date. In my slice of the cake I had already deleted bad building outlines data and remapped them from bing, just in time, it turns out, to give the bot a little less redacting to do. By the time we were out mapping that evening, the bot had aleady delt with the whole of the UK, most of europe and the U.S. (it's now complete of course)
This was the top topic of conversation in the pub. There was some relief and satifaction that things were going smoothly and the bot was running quite quickly. It had been widely predicted that speed was going to be more of an issue. We might have been watching my progress map for months and months, but by this evening it had become clear that the we'd be finished in a matter of days not months. It had also become clear that my progress map was going to be showing rather too many vector objects for most browsers to cope with. I had spotted this flaw from the beginning, but was expecting to have a week or so to figure out a solution!
While the bot was running there was a danger that it would whip up a frenzy of pointlessly alarmist complaints about things like which order it was processing the world in, so it was a relief that these were rendered mute by the speed of the bot's progress. Haiti was granted a special delay for a few days because people asked nicely, but even there, the editing community probably came to a realisation that actually remapping gets a whole lot easier when it's just a matter of filling in gaps. At this stage Poland was also being delayed because of a glitch with the whitelisting and blacklisting there.
In the pub (The Penderel's Oak big wetherspoons in Holborn) we had quite a big crowd, so lots of other topics of conversation. Thanks to Alex for getting some photos
Steve Chilton was there. He's organising the Society Of Cartographers annual conference in London in September. I went to one of these down in Plymouth. Very friendly and enjoyable conference for anyone interested in map making. I recommend it! And it's coming to London (Hosted at UCL)
Martin Cyclestreets was there (not his real name) He was chatting to me about the latest cyclestreets developments. Some frustrations with OpenLayers and things which seem to be a whole lot easier in leaflet (popups were mentioned in particular) He's hoping to drum up some more support for the England Cycling Data Project and he's made a nice screencast of the fat lines merging feature. Maybe we should all chip in and work on improving OSMs cycling data in the UK, because the big G have very recently announced new support for cycle routing in the UK using data given to them by sustrans. On the other hand their routing is a bit lame. What a shame they don't use OpenStreetMap data. What a shame the very large and well funded UK cycling charity sustrans, pay no attention to OpenStreetMap and OpenCycleMap whatsoever.
We had a new chap along in the pub whose name I forget, but he was interested in tube transport data, something I do a bit of work with myself. He was asking, could OpenStreetMap (or something) be used to build up a 3D picture of underground tunnels, including pedestrian access tunnels? I did show him this nifty 3D underground thingy, which is not OpenStreetMap powered. Not open data at all in fact. boo! But it is really neat. Could be done with OSM, but putting this much underground detail (and levels/depth info) into our map data would make for some confusing clutter.
He was also talking about measuring crowdedness on tube trains by firing lasers through the carriages. He was imagining a project TfL could do, but I was already plotting a more grassroots "unnofficial" laser data gathering project :-)
So this was a little catch up on a previous London event. There is another one I'm still to catch up on, and there'll be more news of...
The OpenStreetMap Anniversary Party plans for London (just as soon as I've thought what those plans might be) . This is the BIG pub meet-up of the year. Keep the date free (Sat 18th) and make plans to travel to London if you're within range. If you're not within range, head to the Toronto one, or maybe you need to throw your own OpenStreetMap Anniversary Party in your corner of the globe!
Last week we had a London Summer OpenStreetMap event near smithfields. Although close to my office, it was a great urban explorational experience for me again. I was mapping St Bart's Hospital (here), which I hadn't particularly ever noticed before. A mix of old buildings around a nice square in the middle, a new bit with a sneakily hidden cafe, and lots of big new construction.
Actually I wasn't mapping it, I was remapping it. It's easy to forget that though, if you delete the red data in separate operation beforehand (which feels quite destructive), then later come back and imagine it was never there. You're left with a nice new patch, and it feels like new mapping.
Of course there'll be plenty more "nice new patches" pretty soon, when the license change redaction bot does it's thing. With Andy Allan and Matt Amos both in attendance at the pub, redaction bot development was a top topic. It's become a bit meaningless to say "it's nearly there", partly because we've always been wrong, but partly because we can break down the development and testing of this thing into many stages. But Andy's been making some great progress see rebuild list posts and we're expecting to be running the bot on some real data imminently. In the pub we were discussing this, but Andy was also cautioning not to get too excited. There's plenty of things which could go wrong requiring some re-coding.
We had along Mike (user ViewFromTheBoundary) who is interested in using OpenStreetMap for a website about opera houses. He had done some rendering investigations, but was confused by the project documentation and the confusing array of alternative approaches. Coming along to an OSM meet-up is a great way to cut to the chase! He had been experimenting with a thing called "Generic Mapping Tools", a set of tools with a bad name. Looked like he'd had some success with basic zoomed out road maps. We were chatting about various map display options. The basic markers/mash-up approach versus raster tile rendering versus static map images. Andy Allan was offering advice. Since he's been providing the transport map and setting up thunderforest.com, he's now happy to be known as "the guy that does worldwide custom OSM tile rendering", rather than just "the guy that does OpenCycleMap".
At the data end of this problem Mike was encountering familiar problems of tagging inconsistencies. I think arts venue tagging isn't all that well decided. Heck there isn't even a properly documented tag for an art gallery at the moment (due to too many tag ideas rather than too few). Mike was saying he has some datasets which could be imported. I suggested the "community import" approach, presenting the data and letting mappers to decide how to merge it in on a case by case basis. Shaun used to have a nice example of this with "Bike Shop Locator". Seems to be offline at the moment though.
Speaking of Shaun (the excessively bearded man in the above picture), he's taking a new job at itoworld and it sounds like he's going to be working on "the fun stuff", OpenStreetMap and their work with map renderings and data QA tools. Awesome! But it means he's leaving London and heading to that mysterious place at the start of the 2008 year of edits video. Shaun seems to cycle the length and breadth of the country twice a week anyway, so I'm sure we'll be seeing him.
Paul the archivist is also leaving London. He's done some great detailed mapping in his time down here, but his archiving contract came to an end so he's heading back home to Derby and searching for archiving pastures new. If you have an archiving opportunity, Paul the archivist is your man! Maybe he'll find something in London again, or hopefully somewhere which needs lots of mapping at least!
It's been rainy as hell in London lately, but we were fortunate to hit upon a warm (humid) evening last Wednesday. On Saturday 4th August I'll be having a BBQ round my house. By then we'll surely be having some summer weather right? So keep that date free, and I'll give you some more details. But let's squeeze in another pub meet-up / mapping evening between now and then. How about next Thursday 19th? Suggested locations anyone? Details will be announced on the wiki page.
London's map is still looking like a bloodbath when you look at the OSM Inspector License Change View and there's not much time left to sort this out before the redaction bots do their thing and sort it out in one big chop. However the bloodbath view is a little misleading. Zoom in and you see patches of building outlines are the main cause of red blotches in central London for example. Easy enough to sort those out. Building outlines would have been mostly put in purely via armchair mapping with bing imagery.
Speaking of which, a few weeks back bing removed their super-hi-res imagery over London, leaving us only with the kind-of-hi-res imagery which was newer but more shadowy, and generally much harder to pick out details in. I suspected this would be a temporary thing, and today I was pleased to discover that the super-hi-res is back, but newly updated. They've now eliminated the kind-of-hi-res imagery set, and made the lower zooms more consistent, with no jarring changes or offset problems while zooming. Here's what Fenchurch Street station (here) looked like last week versus now.
But I don't have an image of how it looked three weeks ago. Back then we had similar super-hi-res, but older.
It's tempting to just sit back and see where the building outlines disappear when the license tainted data is removed, and then fill in some gaps using bing when the time comes.
However it is better to do remapping because we can do a better job of it (survey!) where it was only sketched from bing before, and in some places we need to try to avoid losing too much of the tagging data which has been added to the building outlines over time, after being initially sketched by a non-agreer. That's true of this area around St Paul's where we're going mapping this evening.
I feel bad that I haven't been sorting out this area on my lunchbreaks. I guess I need a pub prize at the end of it all to motivate me, and pub a prize we shall have in form of way 146986827 The Butchers Hook and Cleaver. Clever name you see, because it's right by the Smithfields meat market. See what they've done there? Come join us for beers and map chat. Join us for (re)mapping beforehand if you fancy it. Also I can meet people at 6:30 and give a mapping demonstration if this is requested (so far nobody this time) ...But if it all sounds too confusing you're very welcome to just see us in the pub. All the details. See you TONIGHT??!
On Thursday we had one of our regular London OpenStreetMap events. Standard stuff, but this one turned out a little different. We had quite a low turn out of people who normally regularly turn out, but then three new folks coming along, all of whom opted to join me for a data collection demonstration.
Happily it stopped raining and cleared up nicely in time for mapping. My cake slice slice number 6505 of cake number 116 involved some building mapping and remapping, for a bunch of big office buildings including the thing I always call "the Natwest tower", more properly known as "Tower 42". Tallest building in the city until the new Heron tower was built. When I say the city I mean "The City" financial area of London. Canary Wharf is taller, and is also in this city known as London, but not in "The City" if you know what I mean. Anyway I'd never looked around the base of this building. It's quite interesting. Various complicated 3D arrangements of buildings which will be nightmare to input in all their detail. Also multi-level walkways, which I found quite interesting to explore. They are looking horribly run-down due to being made in concrete which looks messily stained thirty years on, and with drainage problems. There was also signs on the walkways saying "no public right or way" and "no through route". I don't know if they are looking so shabby because they are disused, or we're disallowed from using them because they are shabby. Seems a shame. If only there were some wealthy financial institutions nearby who might spend a bit of money tidying it all up. Anyway, with signs dutifully ignored, we had fun exploring (me and the three people I was demonstrating to)
In the Crosse Keys pub we sat at a tall the table in the corner and admired the cavernous vastness of it. Many hundreds of cubic metres of glorious wetherspoonsness:
I had the feeling the three people I'd been demonstrating to all knew a thing or two about mapping already. Thomas from MapAction is obviously a GISer. I really hope we can make him into more of an OSMer, along with other MapAction folks. Florian was asking lots of questions about different mobile apps to track with, and had obviously tried a bit of mapping before. And then there was Tom Morris who quickly whipped out his laptop and used JOSM on the wetherspoons wifi to input some of the data we'd gathered!
Lots of interesting conversations with Tom actually. We got talking about his local mapping of countryside footpaths around Tunbridge Wells, and also...
Tom Morris is very involved in the wikipedia community. Long time contributor to the Wikipedia Signpost community newsletter. In his admining roles he gets embroiled in endless debates on many and varied topics which come up when you're trying to collaborate on writing the entire canon of human knowledge.
It's pretty interesting to look at some of the policies and power structures they have in place. These are big and complex, and it feels like wikipedia is building a government, which on the one hand is pretty impressive (I find it amazing), but on the other hand it's widely criticised as heavyweight and bureaucratic. We were particularly chatting about the notability criteria, which seems to create a lot of debate, and whether there would ever be a parallel for that in OpenStreetMap. Given that we map the real world as it is, there's less to debate, but I wonder whether deciding on a maximum level of detail is a bit like deciding on notability.
If you're interested in polices and power structures of wikipedia and how they compare to OSM (also in relation to funding and staffing choices) have a read of Frederik's 'learning from wikipedia' essay.
Incidentally discussion came up in this thread a couple of days ago, of whether or not OSM's messaging system could follow wikipedia's public "user talk page" approach. I think we should go this way, and I'd like to chat about it some more. Is this one for the 'design' mailing list?
And speaking of mailing list discussion, we also chatted about how OSM is perceived within the wikipedia community. Tom Morris was saying there'd been a discussion recently prompted by the TomTom thing. found it here. Some of the usual criticisms of OSM there, but on the whole it seems like wikipedia folks are getting in on the OpenStreetMap fun a bit more these days. Their new virtualised clusters of development tool servers known as 'wikimedia labs' already include some OpenStreetMap rendering experiments (see the OpenStreetMap mention here), and Tom hinted there may be more of that to come. Collaboration of the collaborations!
We'll announce the next London OpenStreetMap summer event on the wiki page and on @OSMLondon. As ever I'm pondering whether an event of a different format should happen, and who could help organise it, and what venues could we use, but in couple of weeks time I'm sure we'll just go to a pub again. And that will be whole lot of fun anyway :-) (other suggestions welcome!)