gregory at arenius.com
Mon Aug 16 05:55:50 BST 2010
Since I would like to hear more about what goes on at some of the
conferences I can't make I thought I would post some of what stuck out the
most for me at this one. This is just what stuck out to me. If I miss
something or am wrong about something I apologize in advance.
Nama Budhathoki gave a good presentation (over Skype!) on who the
contributors to OSM are and why they do what they do. It had breakdowns of
how much people contributed verse what their reasons for contributing were.
It also had a bit on the backgrounds of OSMers such as age, gender, and
traditional GIS experience.
Randal Hale and Leah Keith gave a talk about using OSM as a teaching tool
with high school students. Her students seemed to really take to it. It
was also very good because it doesn't cost theschool any money if they
already have computers. The FREE component was really important. They can
just make accounts and get started. They used Mapzen because they found it
to be the most user friendly. Even after the class project some of the
students have continued to contribute useful data to the map.
Jon Nystrom gave a talk about ArcGIS being able to work directly with OSM
files. Many attendees were excited about this because many people in
attendance came from a GIS background and 'grew up on' ArcGIS. People like
to use the tools they know. It will probably help more professionals
contribute to OSM because they won't have to learn a new tool set.
David Cole gave a talk about Mapquest starting to use and examine Mapquest
data. The next day David Nesbitt gave a talk on how Mapquest routing can
work with OSM data in their testbed. Basically Mapquest is looking at using
OSM data instead of proprietary data sources. They plan on contributing
back to OSM in kind and with financial support. The routing data talked
some about shortcomings in the OSM data set especially in the US. Some
problems were missing turn restrictions, bad topology (missing connections
or connections that don't actually exist), handling of roads to ferry
terminals, and driveways tagged as residential roads. Oh, and
addressability. One mentioned strength was good road classifications as
their routing algorithm relies on that pretty heavily. They're using a
mostly open stack except for their routing algorithm. They've released
they're stylesheets under an open licence but they're still a work in
progress. They have a big tile server that is for open use that can handle
pretty much anything we can throw at it; if I recall correctly something
like 4000 requests a second. You can check out their work at
http://open.mapquest.co.uk. Really awesome stuff. I'm excited by the
possibility that the maps I help make could touch that many people. Wicked
Learon Dalby gave a talk about getting government data into OSM from the
government side. He is part of (head of???) the Arkansas (AR!) GIS team.
They've collected a lot of good data and have released it for free for
anyone to use and he would really like to see it in OSM. The main problem
is how to get it into OSM. There was a general consensus (don't quote me on
that!) that there isn't really a set of well defined best practices or a
good tool chain to make this happen and go smoothly at that large of a
scale. Also, OSMers usually only work on areas that interest them and there
aren't many OSMers in AR. Another problem was how to flag changes we make
to the data set and send those flags back upstream. They wouldn't be able
to take our edits directly but just knowing where changes needed to be made
would be a huge help to them. I think it rocks that the whole open data
movement has made it to the point where there are people in government who
are not merely willing to make data available but that actually want us to
use it and are willing to expend time and effort to make that happen.
Carl Anderson had a similar talk the next day about using government data.
He suggested that using GIS conflation and road matching tools might help
ease imports some even if we have to translate to a GIS format and back.
OpenJump in particular was mentioned as being a good open source tool for
that purpose. He also mentioned how checking the merged data with different
renderers and stylesheets was helpful because they all have different
strengths and weaknesses.
Ian Dees talked about using shp-2-osm to import data into OSM.
We had the OSM-US annual meeting. OSM-US is incorporated, is trying to
become a certified non-profit, and has approximately $250 in its vast
coffers. Voting for the new board kicked off and will be open to OSM-US
members for the next two weeks. Voting will be run on a survey monkey
platform by outside observers from two different open source projects. You
can vote even if you haven't joined yet. You just have to join before you
Thea Clay talked about building community and running mapping parties, mappy
hours and mapathons. Steve mentioned that all of the successful European
mapping communities have monthly mappy hour like events.
Again, this is just what stuck out the most to me and how I remember things
going (I didn't take notes.) If I got something wrong, misspelled a name or
missed talking about your talk I apologize.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Talk-us