[Rd] Non-free packages in R-Forge
djsamperi at gmail.com
Wed Nov 23 21:37:22 CET 2011
2011/11/18 Spencer Graves <spencer.graves at prodsyse.com>:
> Why do you want to reduce demand for Octave by forcing people who want
> to link to a commercial product to abandon Octave?
> Are you familiar with Shapiro and Varian (1998) Information Rules: A
> Strategic Guide to the Network Economy (Harvard Bus. Sch. Pr.)? Varian is
> now the Chief Economist at Google, and his ideas seem to have contributed
> substantially to their success. The book explains that if you want to
> increase the market for your product, you need to make it as easy as
> possible for potential users to use (for as many different purposes).
The book focuses on penetration pricing and lock-in as tools to
maximize profit, not as a means to support the free software community.
In a paper written a few years later the same authors note: "The very idea
of having "conditions" accompanying a "free" good
confuses some people, and opponents of open source software have done
their best to amplify that confusion." (See Linux Adoption
in the Public Sector: An Economic Analysis, Dec. 2003.)
I think what the original poster (Jordi) is concerned about is that there
are conditions (spelled out by FSF/GPL) that do not seem to be followed
in practice. See, for example,
It is important to remember that the FSF/GPL conditions are in addition
to Copyright law, that Copyleft is an extension of Copyright, not in
opposition to it. In particular, the Copyright holders of software released
under GPL are free to release the same software (possibly with a
few extensions) as a proprietary product, without including any
source code. The Copyright holders are also free to permit others
to violate the terms of the FSF/GPL, because only the Copyright
holders are in a legal position to enforce those terms. (In practice
things are complicated by the viral nature of GPL and by the
growing set of Copyright holders over time.)
This may not seem "fair" to some, but I think it is legal (check with
a lawyer to be sure).
> I've used Matlab, and I want to start using Octave. If I can connect
> from only one of these products to some third party software that I'd like
> also to use, that's a reason to use the more flexible product.
> On 11/18/2011 10:32 AM, Jordi Gutiérrez Hermoso wrote:
>> Let me give a little more context of why this is important.
>> As you can read in this thread:
>> The author of MOSEK basically created a non-free library and wants to
>> link it to both Octave and R. Normally this would be a GPL violation;
>> however the author of MOSEK has worked around the GPL by making a
>> wrapper and making the user do the linking, effectively neutering the
>> copyleft of the GPL (and yes, the GPL is not nice, and this
>> non-niceness of the GPL is a feature).
>> I am trying to reject this in Octave. We do not want to condone the
>> proliferation of non-free software. Instead, I invite the makers of
>> MOSEK to make the library free. However, the author has pointed out
>> that R has accepted his plugin, why can't Octave?
>> And this is why I appeal to the GNUness of R, if it still has it. If
>> Octave and R are part of the same organisation, we have to stand
>> together on this, and together pressure the maker of MOSEK to release
>> MOSEK as free software and stop trying to work around the GPL with
>> wrappers and avoiding binary distribution.
>> I am inviting R to work together with Octave on this. If we are both
>> using the GPL and both part of GNU, what good is it if the GPL can be
>> worked around and if we don't both stand for the same principles?
>> This isn't about prohibiting R from running on Windows or Mac (Octave
>> also runs on both because it's the only way to reach those users), nor
>> about meaningless ideology, but about bringing about a very practical
>> result: more free software for the community, more source for
>> So, please, users and developers and overseers of R, work with us. If
>> we are on the same team, can we work towards the same goals?
>> - Jordi G. H.
>> R-devel at r-project.org mailing list
> R-devel at r-project.org mailing list
More information about the R-devel