After spending a full day in a hotel conference room in Chicago, the twenty five attendees at the “MapServer Foundation” meeting and many peers participating via Internet Relay Chat (IRC, a way of sharing typed comments via a group) and listening in via phone, offer up good news: there is enough common ground to set up a foundation and quite a lot of agreement on how to move forward. Adena Schutzberg summarizes what’s been decided and what’s still up in the air.
After spending a full day in a hotel conference room in Chicago, the twenty five attendees at the "MapServer Foundation" meeting and many peers participating via Internet Relay Chat (IRC, a way of sharing typed comments via a group) and listening in via phone, offer up good news: there is enough common ground to set up a foundation and quite a lot of agreement on how to move forward. That's not to say every possible issue is ironed out, but a solid vision and structure was created in what's being called an "exciting and historic day for open source GIS."
The first piece of business was finding a name. Polls and informal discussion beforehand on the foundation e-mail list apparently made that task rather straight forward. Participants selected the Open Source Geospatial Foundation (OSGF or OsGeo, soon at this URL) and I have yet to read a comment from anyone with a significant gripe. Personally, I find it a bit long, but there is no allusion to any existing organization/project/product and it clearly states the group's focus.
The formation of a Board of Directors was addressed early on. Five individuals were selected from a slate of seven offered (all were in Chicago) as an interim board. Four more will be added later. Frankly, any of the attendees and probably nearly any of the virtual participants would have been fine members, but those selected offer a nice mix of geography (North America/Europe) and technology (C+, Java, etc.):
Arnulf Christl - Mapbender /ccgis.de, Germany
Chris Holmes - GeoServer, Open Plans, U.S.
Gary Lang - MapGuide, Autodesk, U.S.
Markus Neteler - GRASS, Istituto Trentino Di Cultura, Italy
Frank Warmerdam - GDAL, OGR, etc., Canada
The first set of board members may serve for a year, or perhaps until the Free and Open Source Software for Geoinformatics to be held this September in Switzerland. What will come next for the board? That's not entirely clear: perhaps longer staggered terms for board members.
OSGF membership will be on an individual, not organization, basis. For now, the five board members, and 16 other individuals in Chicago are members. More are to be added as OSGF matures to a goal of 45. So what are member requirements/responsibilities? Allan Doyle described members as "those willing to do some measure of work" and who are nominated and accepted by the organization. The latter idea parallels with how the Apache Foundation works.
Licenses and Legalities
Few legal issues seem to be nailed down, though a break out group did address the topic. One key topic will be which open source licenses the Foundation will support/accept. For now, reports Doyle, it's likely any Open Source Initiative (OSI) license will be acceptable. The long term goal is to encourage licenses that allow the different projects to work better together. That may mean that in time, some projects seek to change licensing, for example, between GPL and LGPL (GNU General Public License and Lesser/Library GNU General Public License). Such changes, Doyle points out are challenging, since they require all contributors to "ok" the change. Another important process that must be in place involves any sort of potential use of patented code. Should a project be accused of using proprietary code, the organization needs to have a way to immediately halt distribution of it, while the matter is explored. Again, this concept is borrowed from Apache.
The requirements for a project to be part of OSGF have not been fully decided. Moreover, several projects, some suggest, want their communities to help decided if joining is the correct move. After the initial announcement of the MapServer Foundation, I think many are working very hard to be open and inclusive in going forward. It "looks good" that these will be early "founding" OSGF projects:
Estimates suggest that OSGF will require somewhere between a low of $40,000 to a high of $250,000 per year to run, depending on how much is undertaken. Autodesk has provided funding for the first year. The consensus on funding the organization in the long term revolved around sponsorship, but with an interesting twist. Sponsors will gain publicity, good will, and perhaps other perks and be able to allocate 2/3 of their contribution to a specific project. The other 1/3 goes to running the foundation, that is, "overhead." The funds allocated to a project are directed not by the sponsor, but by the Project Steering Committee.
Each OSGF project will have a steering committee which will likely be a mix of OSGF members and non-members). Each will select a representative to report to the OSGF board. It's most likely that person would be a member. The logic behind having the Steering Committees determine how to allocate funds is that too many things are, as Matt Perry puts it "not sexy enough" to be called out to be addressed, like debugging and cleaning code, and the foundation wants to be sure those tasks are done, too. Anyone of course, can hire any programmer to add a feature or fix a bug outside the auspices of the Foundation. The project steering committees follow Apache, reports Doyle, expect that Apache started with a single project and OSGF is starting with multiple projects.
One other point Doyle shared was that not everything within OSGF need be a software project. There was significant interest in a committee to explore open data. He noted that his organization, EOGEO, the Free Earth Foundation, as well as Autodesk, among others expressed interest in such a committee.
Live and Virtual Meeting
Attendees shared their face time with time on IRC sharing what was going on with those around the world. Doyle reports that he didn't find it distracting and that this combination is very common at high tech events. He did note that at traditional conferences chat tends to be about the topic about which the presenter is speaking. At this meeting, it was about reporting what was being said, so it was helpful. He particularly valued being able to keep an eye on other breakouts while participating in just one.
Doyle couldn't be more pleased with how the day went, a sentiment shared universally by the blogs I read. "Everyone was impressed by the entire group everything just flowed there were no arguments." He did note one ten minute period where things went "off track" but which was quickly straightened out. He noted how upbeat everyone is about the future and suggests that the announcement of the MapServer Foundation back in November may well have prodded people to think harder about what they wanted from such an organization. That in turn, may have made Saturday's meeting and the future, all the more focused and successful.
Blog ReportsGary Sherman (QGIS), Spatial Galaxy
Schuyler Erle, Mapping Hacks
Matt Perry, PerryGeo
Paul Ramsey (Refractions Research),GeoTips - Ramsey shares his concern over whether to add PostGIS, gestated at Refractions, to the Foundation's projects
Howard Butler (Hobu, Inc.), Geospatial News from Ames, Iowa - Butler offers high praise to Autodesk for its leadership
Jody Garnett, Java.net - Garnett offers a Java perspective, among other things
Other ResourcesIRC Logs of Meeting (courtesy Gary Sherman)
Images (Courtesy Tyler Mitchell)