Re: ATLAS license discussion, please

> You might look at the C++ Boost licence, this is specifically designed to allow
> both free use of the Boost library, Standardisation of the intellectual content,
> and commercialisation.
>
> My thoughts in general are:
>
> You must decide what you as the author you desire. If you want to support
> the notion that your software is *freely* available you need an open
> licence.
>
> If you want to *enforce* open-ness at the expense of freedom, GNU is the
> traditional supplier of licences. If you take the GNU path you're
> *endorsing* the kind of legal system that also supports software
> patents and authoritarianism. GNU chose to leverage the legal system
> to satisfy its primary goal: open-ness. GNU licences are not, by
> any stretch of the imagination, supporting "free" software as in
> "freedom of use for any purpose".
>
> You must think: do you approve, for example, Apple taking BSD Unix
> and making a proprietary product with it? Do you approve Apple
> being "forced" to throw out GNU Gcc support in favour of LLVM just
> because of GNU's licence policy?
>
> Now, the other thing to think about is: what possible litigation could arise?
> ANY litigation is a major negative. You neither want to sue someone nor
> be sued.
>
> It is not so easy, despite what some courts think. In Australia in some
> states if you leave your keys in your parked car it is a criminal offense! you are
> considered to be "inciting theft".
>
> By analogy, if you put software in which you claim intellectual property rights
> on the Internet and do not provide a reasonable way of enforcing your
> rights, it may at least be that you have violated your own ability to claim
> any rights at all. After all, if I can simply download your software, without
> signing a paper saying I read and accept the licence terms,
> who is to say I have broken any laws? What if the download is done
> by a robot? What if that robot is a system backup?
>
> My point really is: what's the point of trying to enforce a copyright with
> possible exemptions which in fact you not only cannot enforce due to
> lack of funds .. but probably don't even want to enforce?
>
> It is generally considered that in the woeful US and European legal
> systems you actually need a licence to prevent someone taking
> your work, copyrighting it or patenting something YOU invented,
> and then suing YOU for breaching their rights.
>
> Luckily, the primary motivation in such cases is commerical gain: if you're
> not making a commercial gain or interfering with someone else making
> one there are no grounds for a suit.
>
> Ed Borasky wrote:
>
> "My recommendation is
> to consult an attorney. He / she may advise creating an "entity" to
> house the intellectual property before devising a license."
>
> and there is much sense in that: a separate entity holding the
> property my provide your private assets some protection.
>
> If Google stole my software and commercialised it I'd be ecstatic!
>
> IMHO the major concern is not people doing math with your software,
> but people releasing a product (such as a programming language)
> which has the kind of licence THEY want. If you want such people to
> use your software you MUST provide the most liberal licence possible.
>
> For example if my programming language Felix were to incorporate
> Atlas as a "standard" component, your licence would have to be
> compatible with mine (BSD camp: my actual licence is FFAU:
> free for any use). Any GPL licence would conflict with that and exclude
> such use, at best I would be forced to distribute and build your product
> in a special way to ensure your software was kept "at arms length"
> from mine so your licence didn't pollute mine .. or simply not use
> your software.
>
> I have in fact spent considerable effort ensuring I did NOT incorporate
> any software polluted with Gnu licences. Unfortunate since some
> of it is very good. For example to utilise Gnu's Multiple precision
> arithmetic package I was forced to write the interface (header
> file equivalent) by hand, since generating one automatically
> from the C header would render the product "a derived work".
> A hand written interface supporting the required protocol is
> not a derived work because it is derived from the intellectual property
> but not the embodiment of it.
>
>
> The point here is to strongly consider indirect use of your product:
> it is OTHER people's licensing desires you need to think about when
> constructing your own, if you want your product to be widely used.
>
>
>
>
> --
> john skaller
> skaller@...
>
>
>
>
>
> ------------------------------------------------------------------------------
> Special Offer -- Download ArcSight Logger for FREE!
> Finally, a world-class log management solution at an even better
> price-free! And you'll get a free "Love Thy Logs" t-shirt when you
> download Logger. Secure your free ArcSight Logger TODAY!
> http://p.sf.net/sfu/arcsisghtdev2dev
> _______________________________________________
> Math-atlas-devel mailing list
> Math-atlas-devel@...
> https://lists.sourceforge.net/lists/listinfo/math-atlas-devel
>