Wednesday, September 18, 2013

GeoTools IP Review Update

There we go, a blog title designed to scare away all but the most dedicated blog readers.

GeoTools Review Status update

After several weeks of email, and a couple days of fixing, I am in the clear to make a GeoTools 9.0-M1 release. Basically the IP review discussion we have been having with LocationTech is complete for GeoTools.
Here are the links:
Only the subset of GeoTools that is used by uDig was reviewed, and there are still many dependencies left to be contacted. But GeoTools is by far the largest stepping stone.
Thanks to all the project leads who will be (or have already been) contacted. We really appreciate you taking the time to answer a few questions.

Wait Didn't OSGeo do this Already?

GeoTools has gone through the OSGeo Incubation Process, and has an extensive Code Provenance Review.
This review was focused on minimising the risk to the OSGeo Foundation by taking on GeoTools as a project. Some risks the foundation is comfortable with, ie listing all known problems, are not necessarly sufficient for all users (who would like to see known problems fixed).
The OSGeo Foundation does not require projects fix all issues in order to give interested parties a chance to fix (or fund) the work.

Why this is Cool

So why is this cool? The thing is these intellectual property reviews happen all the time, but behind closed doors. If we are really lucky us open source projects get told about the result, or a few private emails with questions.
And if an IP review fails, it fails quietly and we never hear anything again. Indeed if we ask what happened often the organisation conducting the review is forbidden to talk to us about it (for fear of the conversation being taken as advice).
Contrast this with recent developments:
  • Google was in position to talk to us about our contributor agreement being unusual. As a result we were able to role out a contribution agreement, OSGeo wide.
  • LocationTech IP review is in position to talk with the GeoTools development team, allowing us to answer any questions, and fix any issues that are of concern.
  • IBM China was able to ask a few questions via private email which were passed on to the project team as bug reports
These developments lower the barrier of entry for using GeoTools in a corporate setting.

So how did we do?

We kind of nailed it, a big thanks to:
  • All those who helped dig into the past
    (ie sort out where the code comes from).
  • And Brett Walker for putting together our github pull request policy
    (ie looking into the future and sorting out where our code will come from)
Many project have sorted out how patches work, but have not extended that idea to cover github pull requests.
The only area where we could do a bit better is with the EMF generated code which currently says:
It would be nice to configure EMF to fill this in next time we generate stuff. As an example for CSW I ended up with:
 * (C) 2012, Open Source Geospatial Foundation (OSGeo)
 * (C) 2004,2010 Open Geospatial Consortium.
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation;
 * version 2.1 of the License.
 * This library is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * Lesser General Public License for more details.
 * $Id$