I am missing something here, maybe it is because I have autism. Your project registration page assumes my project is already finished. In the spirit of divide and conquer, parts of it are finished, parts of it are workable, parts of it I have ideas on and some code written, and some I haven't started on.
I am working on something that is mostly written in Perl, and I may be able to find dependencies that already exist in CPAN. If a particular module that might be useful has a license incompatible with the GPL, I guess I would have to find an alternative or write it myself. I can't say my dependencies ARE compatible with my project license, because I have written them yet. They WILL BE. I haven't seen a Makefile that is copyrighted, or a test program, or lots of other things. License headers on data files? There are a number of parts of this that are useful for other purposes. For example: outliers in data are common in the kind of data most people would use this for. Most common methods of detecting outliers, are only meant to detect 1 outlier, and are not intended to be used recursively. I have a method (and a module) which can detect up to 9 outliers in up to 60 data points that is part of this, and is useful all on its own. As my project requires the data to be recursively partitioned in a quadtree, at some point a partition will fall below the 60 point maximum, and I can then look for up to 9 outliers. Is it not useful to get these sub-parts into the world? Or, do I have to have 30+ (actually I am guessing 45) different projects, that are all small parts of this project which I think people would be interested in? Some possibly enough interested to help? :-) Gord
