Re: libofx license
Benoit Grégoire <[email protected]> Tue, 31 Oct 2006 13:48:21 -0500
| Newsgroups | gmane.comp.finance.libofx.devel |
|---|---|
| Message-ID | <[email protected]> |
On Tuesday 31 October 2006 02:55, Thomas Viehmann wrote: > Hi Benoit, > > I'm writing to you as the (main) copyright holder of LibOFX > (I've found yourself, Peter O'Gorman, Ace Jones, Martin Preuss, and > Kasper Peeters in AUTHORS and copyright notices). > > So here is my question: Would you be willing to consider including an > explicit permission to link your work to OpenSSL? > > This may seem a strange request, but here it is: > > As you probably know, Martin wrote a banking library AqBanking which can > make use of LibOFX. Both LibOFX and AqBanking are licensed under the GPL > and both are packaged in Debian, where I comaintain packages for > AqBanking. We occasionally receive requests to enable OFX functions in > AqBanking. > Unfortunately, as some (notable the Free Software Foundation and Debian) > deem the GPL as incompatible to the OpenSSL license and as AqBanking > depends (rather strongly) on OpenSSL, we cannot presently link LibOFX > into AqBanking. > Now, you may share or not share this interpretation and I certainly > neither can nor want to impose any licensing on you, but I would really > like to include your work into the packaging and so would be very glad > if you generously granted explicit permission to link to OpenSSL.[1] Unfortunately, I do share the interpretation that it's impossible to respect both licences at the same time without an exception (with some nuances, but those nuances do not apply to this specific case). I fully realize that for the end user, not granting an exception means less availability of OFX support in the short term, and that granting permission would be very easy for me. But I simply do not want to compound the problem by making OpenSSL's advertizing clause even more entrenched, no matter how little difference my decision may actually make. I used to have the opposite viewpoint, but I've clearly been proven wrong. I now feel that having easy workaround for licencing issues since "the user just doesn't care" has directly resulted in the lack of developement of truly open, full featured drivers for just about all commonly available video and wireless cards (though the latter is improving). When the end users can do everything they want if they don't care about licencing (and most don't), the developpers don't spend much time working on the alternatives. I feel we are shooting ourselves in the foot in the long run. Now off course, the sadder thing is that OpenSSL actually IS free software. It's too bad that as a community we haven't been able to hamonize all copyleft licences with each other, and make all non-copyleft licences compatible with all copyleft one. In this specific case, where there is a mature, full featured and widely available alternative (GnuTLS, which even has an OpenSSL compatibility layer), I simply cannot see a reason to make an exception. I truly am sorry to have to say no to a polite and perfectly reasonable request, -- Benoit Grégoire http://benoitg.coeus.ca/ -- Benoit Grégoire http://benoitg.coeus.ca/ ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642