On Mon, 2009-01-19 at 16:37 +0100, Daniele Romagnoli wrote: > Hi list, > I would like to move the gt-imageio-ext-gdal module from unsupported > to plugin module.
Daniele, That I even write this mail, you will take the wrong way. You will assume, no doubt, that I am attacking you, that I want GeoSolutions to fail, that I am simply taking pleasure in criticizing you, that ... Well so be it, perceive my intentions as you will. That is actually, profoundly, not my intent but everyone is so paranoid in this community that I have given up caring how I am read. So why do I bother writing this? In a broad sense, my motivation appears to be my desire to build a world class geospatial library of free software that I can depend on for ecological research and distribute to my colleagues and students. In the narrow sense, I suppose I am composing this to help myself understand better the bitter taste in my mouth tonight so that I may learn from it. It will take me a good few hours to write this mail, then I will dwell on it all evening, and eventually I may even decide not to send this. Still, I presume I will learn about myself in the process. I also write you to address directly the issues that your recent call for a vote and lightning commit of the imageio-ext dependency bring to the forefront. Since you said your imageio-ext project had hit 1.0.0 yesterday and today you called for its inclusion into GeoTools, I took some time out from my work this evening to check it out. I wanted to see what a 1.0.0 number brought to the code, and see if I would recommend having GeoTools depend on it. Not knowing the code, I opened files perfectly randomly. The first was a utility class (seems like a good place to start, no?) which contained only static methods but had a chaotic view of inheritance: it's not final but not instantiable and then has some public, some private and some package-protected methods. Okay, that's not a blocker problem but suggests a lack of care or understanding. The second class I opened, lists only Martin as the author but is taken into your namespace and has two separate headers for the same license. Seems weird, again not a blocker but suggesting a lack of care and confusion about licensing. The third class I opened has a Sun header which imposes additional restrictions beyond the LGPL on all subsequent distributors and has an (L)GPL incompatible field-of-use restriction. My review had only hit three classes and I had struck out. I conclude that GeoTools and its derivatives legally cannot depend on your imageio-ext library until the license issues have been resolved and that, at that point, we would be well advised to look more deeply at the code to assess its quality. You will remember that I raised the issue of the Sun license incompatibility several months ago. At that time, I took a similar brief look at the project's code and discovered a similar (the same?) header. (Maybe it's just this class, have you taken lots of code from Sun?) At that point you suggested that I should (1) list the problematic classes and (2) propose fixes to the code. That is not my job, not my project, and clearly not a task that falls on my shoulders. I have done enough work in the licensing area to last me for quite a while. Aside from that, to my chagrin, you have bulldozed your proposal into reality. After a long phone call helping another member of our community, I turn around to find you have, a mere two hours after calling for a vote, committed our project to your new dependency, to this new course, and to all the issues that will entail. I am surprised. I find it surprisingly disrespectful to not even wait for one spin of the globe to see if anyone in Australia, anyone busy this afternoon, or anyone away lying sick in bed would have anything to add. But perhaps you did not really want any opinion, but rather seeking an excuse to commit. If so, it seems silly to make a show of asking the community for their opinion but then committing before any productive opinion could possibly have offered. If you don't actually want any opinion, then save yourself the trouble of asking for it. Now, I don't actually assume that you did any of this to be disrespectful intentionally; I merely think that you want to integrate your code as a direct dependency of Geotools, no matter the opinion of anyone else. So now, I am not sure how to proceed. Do you have any suggestions about resolving the legal situation with regards to the Sun copyright code? Do you not care? Do you care and have no time to resolve it? Do you have any other ideas? I know what I will need to do to keep my code legally distributable but I am not sure if I need to do this completely outside the SVN. Aside from that practical issue, your work today has left a bitter feeling in my mouth and given me less energy to keep arguing with my local community that geotools the project is worth the struggle. Do you have any suggestions for how we could actually collaborate on this project given our different approaches? sadly, --adrian custer ------------------------------------------------------------------------------ This SF.net email is sponsored by: SourcForge Community SourceForge wants to tell your story. http://p.sf.net/sfu/sf-spreadtheword _______________________________________________ Geotools-devel mailing list Geotools-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geotools-devel
