Request for Open Source Mapping help
Anyone care to review the following snippet, about open source GIS projects, for accuracy, completeness, fairness, etc.? It’s part of a much larger piece on the GIS industry. I’m on short deadline here…
<Draft>
One of the leaders at this point is the University of Minnesota’s MapServer. It was initially developed as part of the ForNet forestry management project, funded by the state of Minnesota and NASA. While the MapServer lets an application display a browsable map, the site notes it, “is not a full-featured GIS system, nor does it aspire to be.” The US Geological Survey announced last month that it will use the MapServer technology to help build The National Map, an open source map server with access to 20 terabytes of data from TopoZone.com (a site created by Maps a la carte, Inc.). TopoZone’s map data comes, in turn, from the USGS as well as other sources.
There are other open source GIS web servers and applications. For example, GeoServer [led by whom?] implements the OpenGIS Consortium’s Web Feature Server with the noble aim of making the citizenry better informed about matters geographic. PostGIS adds support for geographic objects to the open source Postgres SQL database. GRASS (Geographic Resources Analysis Support System), an international effort hosted in Italy, Germany and at Baylor University in the US, is strong in producing map graphics.
Of course, the open source projects generally don’t provide all the functionality that the commercial services do. For example, if you give MapPoint or ESRI’s map server a list of points you want to visit, you will get back an optimized routing map. MapServer and its like don’t offer that functionality. But, because the open source map servers are non-proprietary, the community can add the features it needs as it needs them. For example, the Rosa Java Applet is one among several tools developed by the DM Solutions Group that add functionality to the MapServer, enabling users to interact with an image of a map by clicking, pressing buttons and dragging and dropping. Open source lags commercial development in this space, but it is good enough for many applications…not to mention it has the advantage of being open source.
</Draft>
Thanks!
Categories: Uncategorized dw
You may want to state who the users of these projects are- the people managing these projects can certainly tell you that.
There’s also the OpenMap toolset (Java) from BBN… .
GeoServer development is led by the Open Planning Project (openplans.org). Their web feature server is the reference implementation for OpenGIS compliance (more info on that here: http://cite.occamlab.com/reference/). I agree with UMN Mapserver as the most significant OS project. I think the next most important are GDAL and PROJ (http://www.remotesensing.org/projects/hosted.php), GIS file format processing and projection libraries (at the heart of Mapserver and others), and Java Topology suite (http://www.vividsolutions.com/jts/JTSHome.htm), a Java library for spatial processing (at the heart of Geoserver, GeoTools, and others).
The Open Planning Project is the lead for GeoServer. The webpage is horribly broken and outdated, as we’ve been too focused on development, but my task for next week is to fix it up. TOPP is a non-profit ‘dedicated to enhancing the ability of all citizens to engage in meaningful dialouge with their environment’. TOPP also supports the virtual terrain project: http://vterrain.org – another open source gis related project, focused on 3d modeling.
Chris, how should I refer to you in the article? What’s your attribution? (Feel free to respond via email.)
Two things –
First, it’s worth mentioning the link between free software and free geodata; and you should definitely link to FreeGIS – they have the most up-to-date and comprehensive list of both. You already mention the link to open standards. The three together are interesting.
Second, something unusual is happening with Free Software and geospace. Usually Free Software lags proprietary tech – but in this area, we’re leapfrogging them. Openness affords interoperability, which when you’re dealing with Geodata (or any kind of content, really; this is like the story of the Web redux) provides a multiplying effect (or a network externality, whatever you want to call it). Also, traditional GIS looks only at the high-end. whereas open source is looking much more at the multitudes. One example is (cough) the project I’m involved with. 3map (GNU GPL and LGPL) provides a Virtual Reality front-end to online geodata. The idea is to help people interact with the emerging geospatial web… This is something that the proprietary GIS people aren’t doing.
Viveka, Thanks for the info. I already had included FreeGIS in the reference list, but I had missed your site. Sorry! (I spoke with Keyhole.com yesterday, a commercial site similar to yours. How does your data set compare with theirs?)
I’m not convinced that the open source projects are leapfrogging the proprietors, although I’m certainly ready to be convinced. E.g., my understanding of open source map servers such as MapServer is that they don’t offer the same level of features/services as you can get commercially. E.g., optimized routing. So, while there may be some areas in which the open source community is ahead of the proprietors, when it comes to the functionality that commercial clients are looking for, is open source really leapfrogging? Or is it only leapfrogging in areas that are fringier to commercial customers? Please straighten out my thinking!
Thanks.
Hi Dave,
I think you’re right – the open source map engines are exploring areas that the mainstream proprietary vendors consider fringe. Those vendors have a traditional large enterprise customer base – corporate/government entities using geodata internally. They’re very comfortable doing what they’ve always done, servicing their existing customers. I don’t see the open source tools taking over in that sphere. Probably GRASS is the closest.
I do think though that the current fringe – mass-market non-expert users – is potentially huge. And that’s where a lot of the open source work is focused. Grassroots mapping, PPGIS (Public Participatory GIS). Also a lot of the open source work is exploratory; looking at the intersections of geospace and social networks, that kind of thing. Not in direct competition with the proprietary tools, but working at a tangent, innovating in unusual ways. This seems unusual for open source software, which has a reputation for following closely behind proprietary work (e.g. openoffice, gnome/kde).
On your question about how our dataset compares to Keyhole’s – it doesn’t. While there’s a superficial resemblance in our two projects (we both present a 3D globe to the end-user), they’re really quite different. From what I understand, Keyhole started out with a proprietary dataset, and developed their globe as a way to present it. They’re interested in building up their proprietary data holdings and selling access to them through their digital globe. They’ve done a marvellous job with their software and have a very impressive dataset; all credit to them.
3map doesn’t have a dataset – we’re not interested in selling our own data. We’re interested in helping people who do have geodata, and who want to publish it.
We have a base map built up from free sources – mostly NASA. There’s a trend towards openness, particularly in government-created geodata, which we are riding and will hopefully facilitate (we have two trial programs going with government agencies & universities in Australia, where we’re based; they seem surprisingly eager to publish their geodata in open and standard formats). We think that maps are a fundamental method that humans use to organise spatial information, and that good maps are easy to use. Current GIS systems are not really maps, they’re tools for experts to create maps with. OK, I’m devolving into rant now, I should stop.
No no, don’t stop! Great stuff. Thank you.
I don’t know how off-topic this is, but the open-source Internet geolocation projects are now better than the commercial equivalents. TJ Mather (with GeoIP) and I (with IP::Country, javainetlocator and ColdFusion GeoLocator) have basically destroyed much of the commercial market for this kind of product.
Nigel, can you tell me more? What exactly does your geolocator do? Can you quantify, one way or another, how much better it is than the commercial products?
Thanks.
Don’t forget about JUMP by Vivid Solutions. It’s an awesome and well documented desktop GIS written in Java.