[HOT] OpenStreetMap data, good and bad

john whelan jwhelan0112 at gmail.com
Tue Nov 15 20:33:45 UTC 2016

One final comment that has been mentioned before is that many mappers are
mapping on laptops without a mouse so their control isn't that good.  Add
in iD without a building plug in like JOSM and you get the very odd shaped
buildings we see on the map.  They can be squared but that is another

JOSM with the building tool plug in is seen as too complicated by many
mapathon organisers.

Cheerio John

On 15 November 2016 at 14:13, john whelan <jwhelan0112 at gmail.com> wrote:

> Looking at the other half there are only 3,000 untagged ways.  In general
> it doesn't look too bad but yes there are some areas that might not meet a
> desired standard.
> I think the problem is three fold, first experienced mappers who map as
> they would in Europe without regard to the http://wiki.openstreetmap.org/
> wiki/Highway_Tag_Africa or the convention that huts should be mapped as
> an area rather than a node.  These aren't a major problem as the quality is
> otherwise fairly good.
> Second mapathon or inexperienced mappers not knowing what they are doing.
> Validation helps here if you give them feedback you get more out of them
> and its better quality but you need a validator who knows what they are
> doing and don't even try and validate in iD.  One major difficulty is that
> many new mappers will not finish a tile which makes their work more
> difficult to spot when validating. If you can ask them nicely to map from
> the top then it gives someone a chance to go over their work without having
> to scan the entire project.
> However we don't have enough validators, especially ones who can give
> feedback without turning the mappers off.  To be honest when faced with
> another seventeen new mappers sixteen of which you know will probably only
> map once validating their work isn't that rewarding and its faster to map
> it yourself than to do a full validation feedback.
> http://tasks.hotosm.org/project/1396# is an example of a validated
> project.  We still get new mappers and mapathons on there but the quality
> isn't too bad.  We also have a pair of very nice mappers who cleans up
> tiles that have been marked invalid because they're incomplete.
> http://tasks.hotosm.org/project/1087 is another project that gets a lot
> of validation, its big but at the moment it seems to have attracted the
> attention of a local mapper and developing local expertise is always a good
> thing.
> I think the third cause in Africa is the variety of local languages used,
> they aren't all supported in the training guides and wiki.  Local mappers
> are not supported as well as they could be.
> Training helps, but not every mapathon has it nor do all the organisers
> know much about OSM.  Projects that are limited ie buildings or highways
> and settlements mean there is less information to absorb by new mappers.
> Cheerio John
> On 15 November 2016 at 08:26, john whelan <jwhelan0112 at gmail.com> wrote:
>> Looking in JOSM at half Tanzania there are some 4,000 untagged ways,
>> 26,000 duplicated nodes etc.
>> I'll take a look over the next few days.
>> Cheerio John
>> On 15 November 2016 at 01:12, Douglas Ssebaggala <
>> douglas.ssebaggala at hotosm.org> wrote:
>>> Dear community members,
>>> While at the spatial data workshop
>>> <http://i2ifacility.org/Spatial_Data/> a few days back in Tanzania,
>>> there was unique feedback from some of the people and organisations who
>>> have used OSM data: they greatly appreciated the quality of the data in
>>> OSM: buildings, roads, and business information, e.t.c thanks to the Ramani
>>> Huria team, and community members who are doing the mapping.
>>> This data is being used <http://www.worldpop.org.uk/> for analysis,
>>> insight generation and maps!! However, the need for more quality data was
>>> also echoed as a very important aspect needed in all these efforts.
>>> A while back there was a discussion on new mappers, and generally making
>>> wrong edits to OSM, especially during need for maps and data in critical
>>> situations: considering that *mapping for flood resilience* scaled to
>>> involve *new* community members/*mappers* whose data is now proving
>>> *usefu*l, and good quality, i think we should always guide, understand,
>>> and help out any new mappers who might make mistakes, as with time their
>>> contributions will be helpful.
>>> --
>>> *Humanitarian OpenStreetMap Team
>>> *Mapping Supervisor "Financial Services", Uganda
>>> Email: douglas.ssebaggala at hotosm.org
>>> Skype: douglo.m
>>> Twitter: @douglaseru <https://twitter.com/douglaseru>
>>> UG Mobile: +256 772 422 524
>>> *Using OpenStreetMap for Humanitarian Response & Economic Development*
>>> web <http://hotosm.org/> | twitter <https://twitter.com/hotosm> | facebook <https://www.facebook.com/hotosm> | donate <http://hotosm.org/donate>
>>> _______________________________________________
>>> HOT mailing list
>>> HOT at openstreetmap.org
>>> https://lists.openstreetmap.org/listinfo/hot
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/hot/attachments/20161115/75c991bb/attachment.html>

More information about the HOT mailing list