LWN: Comments on "FSF reboots its High Priority list with a grant and call for input"

FSF reboots its High Priority list with a grant and call for input

pabs — Sun, 05 Oct 2008 17:33:35 +0000

FOSSology does something similar to what you want:

Never happy :-)

coriordan — Fri, 03 Oct 2008 12:02:58 +0000

I know of two companies that provide such tools, specialising in free software. They charge for the service, but then again, this type of multi-licence analysis of combinations of many software packages is generally only needed by companies.

There's a touch of irony in seeing this comment when FSF announces funding for (specific) software developement, given that I've often seen FSF explaining the importance of licences and people replying, asking why FSF doesn't put more resources into general software development :-)

(Of course, just for completeness, I'll mention that their reason for not focusing on general software development is the same as FSFE's reason: the market value of free software already causes companies to employ thousands of developers for general software development. Non-profits like FSF and FSFE who can employ a handful of people must focus on the necessary work that we *can't* leave to corporations or "the market": awareness of the principles of freedom and the licences that make everyone play fair.)

FSF reboots its High Priority list with a grant and call for input

yarikoptic — Fri, 03 Oct 2008 02:04:11 +0000

imho besides the projects trying to replace non-free software, transparent legal support needs to be done. FSF indeed goes after the bad guys who fail to respect FOSS licenses, but we have no automagic protection against our innocent failures.

Coming from Debian developer perspective of view:
Licenses: FOSS world is full of different licenses which might be incompatible one direction or both. Simple relicensing is an option if there is a limited set of copyright holders. We have no automagic tools which would at least try to guess all mentioned licenses in the parts of code, check for their compatibility and compliance with the license of the shipped end-product... not to say (judging form SCO-vs-Linux history) that theoretically someone could build a tool to extract signatures (hell simple sha sums would be sufficient for the starting point) of code snippets and comparing them (and corresponding licenseS) to the known once in a central DB.

Trademarks: I know about few projects which got contacted by lawyers of big companies since apparently FOSS software was infringing a proprietary trademark (although the names for the projects were quite different)... sure thing poor scared developers preferred to simply pay a little fee under agreement of non-disclosure of the "deal" and changing the name of the project. But little portion of developers knows about existing trademarks in a particular field of applications... and generally speaking it is easy to check either the trademark is already registered at least in the states... e.g. would you know that 'junkyard' was trademarked by Apple some time ago? http://tess2.uspto.gov/bin/showfield?f=doc&state=johk...

So if I were FSF, I would just announce a competition for the creation of such a tool with a prize of 10k$ (somewhat like a fixed GSOC project ;-)). In the long term imho it would have greater impact and be more valuable to FOSS community.