SimonPoole has commented on the following diary entries

Post When Comment
Tagged and Untagged Nodes 3 months ago

The redaction process created a large number of orphan untagged nodes, typical example of that happening would be when a road was redacted away, but the nodes not (because they where created / moved by somebody that accepted the CTs). As a result the nodes may still have residual geometry information (by how they are arranged) and should only be removed when that aspect has been checked.

The other source of such nodes are naturally (broken) imports, unluckily there is no penalty for not cleaning up after you have messed up.

Using OSM to improve government data 4 months ago

Oldie from the LWG on the topic item 4

Nodes in Asian cities in the last 10 years 4 months ago

2011 is simple to explain: its the first year we had a good global imagery resource in the form of Bing. While we previously had access to the yahoo product, coverage and quality was rather limited.

Vespucci 0.9.8 Release 4 months ago

See The main problem is that I need to design, add and test UI elements for multiple dozens of possible OH elements and variants of them.

Let's Talk Local at the Global State of the Map 4 months ago

I nearly fully agree with Glassman: local chapters happen when there is a local community and maybe semi-formal local groups that have come to the conclusion that a formal organisation makes sense and could be sustainable.

There is no indication that doing things the other way around does anything for community development and I would oppose any attempts and any tendencies to require or even just to imply that higher degrees of formal organisation are necessary to participate in OSM. Besides driving away contributors that are specifically not looking for that in OSM, it just makes it too easy to install a Systemé Blatter.


A look into a sample of edits from MAPS.ME contributors from August 4 months ago

WRT the new contributors per month stats: it isn't so simple, given that there was both the launch of the editing capability and a larger HOT/MM event during the same period.

Down the road once things have stabilized a bit there will be some analysis, given that the peaks alone doesn't really say much.

@planemad well it is not different in that respect than numbers that MB published claiming everything is fine and dandy ......

"Welcome-to-new-mappers" program in the Netherlands comes to an end. 6 months ago

Marc I think it is a bit of a shame that you are stopping with sending the messages (as you know I've been running something similar since late 2014 for a similar number of new mappers per year in Switzerland).

While I'm the first one to agree that there is not much feedback or noticeable behavioural change, it does give the 10% or whatever that map a bit more a fighting chance of finding national/regional/local resources that otherwise would be very difficult to find (I have over the last two years put some effort in to improving the relevant parts of the general signup process, but as it is currently, it is difficult to provide region specific messages).

Off to message the six new mappers from today....


What's going in Egypt? 6 months ago

PS: naturally it could simply be that there is new imagery available, but the user in question would seem to be a very very prolific mapper in that case.

What's going in Egypt? 6 months ago

It is not unlikely that that is Facebook see if data is breaking it would be important to give both facebook and the DWG ( a heads up.

Experimental publishing of Sentinel 2 satellite data 6 months ago

@imagico iD has been using the "best" imagery for an area as default since early this year, and has been recommending it in the layer selection since it exists AFAIK.

Arme OSMF 7 months ago

@russdeffner of the good $1'000'000 HOT expenses last year roughly $750'000 was used for salaries and related things, not really different than any other organisation, commercial or not.

What is at best misleading is claiming that the projects are "OSM community building", they are simply things which of interest to HOT (which is naturally HOTs good right), but are neither coordinated with the wider OSM community, nor of interest too it.

Just as the techincal projects, with the one exception of the tasking manager which has seen some use outside of a narrow HOT focus, were neither speced with a wider OSM audience in mind nor are of general use. Some are obviously direct HOT marketing activies, for example "OSM analytics" which uses exessive lean-to doodling as a quality metric, which is really only useful for missing maps marketing.

Flood Lagoons? What Flood Lagoons? 7 months ago

See you might want to add a specific value for floods ....

Updated contributor stats 10 months ago

@PlaneMad except that in depth (numerical) analysis doesn't really show a clear seasonal pattern. Yes, sure, end of year tends to be slower, but thats about the only thing that can be said, it clearly doesn't explain the specific large drop (which btw your highlighting shows nicely).

CR+ID Crowd-mapping Workshop using OpenStreetMap - Prosperidad & Tandag, Caraga, Philippines 11 months ago

See various bits and peices are just relevant for yet unreleased 0.9.8, but 0.9.7 should be able to digest all of it.

CR+ID Crowd-mapping Workshop using OpenStreetMap - Prosperidad & Tandag, Caraga, Philippines 11 months ago

I wouldn't know what is fiddly in tag-only editing mode (you need to switch out of it to add new features but that is just a long click away). One of the major advantages of vespucci is that you can use customized presets which can make adding stuff real fast, and naturally, because you have the actual OSM data available, you know what is really already present in the data.

Currently vespucci can't read .osc files (not that that would be a big thing to implement, simply nobody has turned up with a reasonable use case), it supports writing them and reading and writing JOSM format OSM XML files.

As already said OsmAnd is fine, it just doesn't seem to make a lot of sense to use it if you are specifically going out surveying.

CR+ID Crowd-mapping Workshop using OpenStreetMap - Prosperidad & Tandag, Caraga, Philippines 11 months ago

Any specific reason you used OsmAnd and not Vespucci for data collections? OsmAnds festures in that respect are at best rudimentary (ok for on the fly non-dedicated POI adding),

Removed 11 months ago

Any specific editing you are concerned about?

Welcome to the new Missing Maps 11 months ago

@pedrito1414 my concerns revolve around a couple of points (and none of them should imply that I'm or others are not supportive of MSF and other MM members)

  • siloization of contributors that have their first and most of the time only contact with OSM through MM and to some lesser degree through HOT activations. In general organized mapping events have very low conversion rates and we only get that one chance to introduce the breadth of OSM. to most of them, it is not clear how much exposure MM event attendees get to other ways of contributing to OSM (and even if they are told that it can be a useful undertaking too) and how many go away to never return with the impression that OSM is only about drawing buildings in underdeveloped countries. See to realize that this is a real issue.

  • contributing to OSM has this wonderful property that you can experience first hand that what you just added is useful. You can see the restaurant you just added on a map within minutes, or navigate to your friends address that you just added and so on. You don't need anybody to tell you that it was useful, you -know-. MM replaces this, admittedly slightly egoistic, intrinsic motivation, with a third party telling you something is good and useful and that the world will be a better place because of it (overstated: with motivation via guilty conscience). I don't believe that this is a recipe for long term success (neither for MM nor OSM). Frederik has already touched on the negative sides of gamifcation as motivation so I'll skip that.

  • overselling of the positive effects (this is less a MM issue, more a general one of humanitarian mapping in the OSM context). People are spending a lot time on mapping things with the impression that they are "saving lives" even though they might actually for example only be improving population statistics which could have been generated in many other ways, using a free and willing work force simply being more convenient and cheaper. In general it would become all to only activate and only ask for an activation when there is a compelling case that spending endless hours of volunteer time will actually make a difference. The overselling (I fully expect to see the next level of humanitarian hard-sell, aka crying babies, popping up on the relevant websites soon) tends to put off people that are willing to help in real emergencies but are not quite so gullible.

  • drowning out of all other aspects of OSM in the media and public perception. Admittedly this is mainly the issue that the non-humanitarian parts of OSM are too quiet, but the net effect is that looking back at 2015 there is no way you can't get the impression that humanitarian mapping is all OSM is about. This clearly impacts our ability to recruit new mappers for "normal" OSM given that the one thing we know is that media reports will drive new signups.

A final note: historically humanitarian mapping in the OSM context was something done mainly by at least moderately experienced mappers that had already mapped at least a couple of things that they could directly relate to (I'm not claiming that net quality was particularly good even then or that there were not other problems with the concept). MM has turned that upside down with the consequence that you need ever more coordination, validation and so on to handle the model of churning through lots of newbies to get comparatively small tasks done. Maybe some rethinking of how to get the tasks done most efficiently or at least trying something else for one target area would be a good idea.

Welcome to the new Missing Maps 11 months ago

@mikelmaron woodpecks transparency concerns are IMHO quite well founded particularly considering the direction MM is going in.

At least some of the funds seem to be funelled through the ARC. Given that the ~$3 billion ARC budget is quite opaque at the level that would interest OSM, it doesn't seem to be unfair to ask for:

  • pro-forma accounts (as I understand MM is not formerly incorporated and as a consequence is essentially not subject to any formal controls financial and other)
  • a list of any donations and associated donors that were earmarked for MM and either given directly or via member organisations.

A further note. MM and the ARC seem to assume that repeatedly spending funds on and talking to the same companies inside the beltway amounts to engaging the OSM community, it doesn't.

Addressing scheme (why not?) 11 months ago

While most OSM contributors will agree with you that the two cases you list turn up often enough that they need a solution, the solutions are already here and have been in use for a long time:

  • multiple addresses in one building: multiple address nodes, or if known addresses on the entrances nodes

  • large site with multiple buildings: site relations.

The former doesn't need any special support, the later is currently not rendered in the standard style (there are some technical hurdles before that happens). But as always there in OSM there is always an interaction between use of a specific mapping scheme and its support in the tools, simply use it more and it will get more support.