On Thu, 2008-07-10 at 10:08 +0100, Daniel P. Berrange wrote: > On Wed, Jul 09, 2008 at 02:59:48PM -0700, David Lutterkort wrote: > > I am thinking about relicensing Augeas under the LGPLv3, one of the > > reasons being that I want to use at least one gnulib module > > (canonicalize_file_name) that is only available under the LGPLv3. > > What does that do ? Can you use realpath() instead ?
It's the non-broken version of realpath. See the BUGS section of realpath(3) > > As far as I understand the matter, there should be little impact on the > > users of Augeas, in particular what can be linked against libaugeas, but > > IANAL, and I'd like to hear from others before I take that step. > > Well the obvious impact is prevention of its use in GPLv2 only programs, > and any (L)GPLv2+ programs which use it, result in a LGPLv3+ combined > work. Personally I take a conservative view & prefer to keep things > (L)GPLv2+ licensed. Though clearly in a year or two there'll be a tipping > point where too many useful libs are v3+ and it thus ceases to be pratical > to care about v2 compatability. We're not there yet though Yeah, that's the biggest concern; the main issue though is that reliable licensing information is almost impossible to come by in bulk. > The question is how many of these would be interested in Augeus ? The > most likely candidates are KDE, Java, HAL and Ruby since they're general > purpose libs/apps sitting as a foundation for many other apps The RPM licensing information is seriously unreliable; just spot-checking two off your list, the Ruby license is compatible to the GPL (both versions) and enscript is actually GPLv2+. This whole licensing business is an insane mess ... David _______________________________________________ augeas-devel mailing list [email protected] https://www.redhat.com/mailman/listinfo/augeas-devel
