> So, it is legally possible to do a clean room reimplementation of this > code, but it does not look like that is what happened with the existing > jsmin.py. > > Basically, what someone would have to do is to write a version of jsmin, > looking only at an "algorithm description", but never jsmin.c. Ok, thanks for confirming this was a possibility. > Red Hat Legal compared this jsmin.py to the jsmin.c code, and it is > their opinion that it is not a clean-room reimplementation, but rather a > "conscious" translation from C to Python. > > If I had to guess, if pressed, the jsmin.py author will admit to having > looked at jsmin.c. I actually asked him, to be sure, and that's what he told me as well, that he simply translated into Python. > So, you can either find someone to make a jsmin in a clean room > reimplementation, then use it, or rework your package to not use it. I removed the jsmin tools from the source tarball of OpenLayers and rebuilt it for F-11 and Rawhide (I don't have the commit access in F-10, and it will be EOL-ed soon anyway, if that really matters I'll ask for the commit ACL and do it there as well). In the future, I'll try to see if another Javascript minifier couldn't be used instead of jsmin and what the OpenLayers devs think about it. Best regards, ---------- Mathieu Bridon (bochecha) _______________________________________________ Fedora-legal-list mailing list Fedora-legal-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-legal-list
