[RouterGeocoder] Dual-licensing

[Sampson, David]

First off, I would argue FOSS libraries are already in Commercial
products. Infact full FOSS applications are sometimes found within
Closed source applications.

I find it interesting while some of the biggest software and technology
solutions are moving their products to the open source market, we are
still talking about moving open source into the closed source realm. We
might not be closing the source, but we are closing the model that makes
FOSS successful.

Admittedly FOSS is not the proper solution for all software, but where
it does make sense I see evidence that it works just fine. And the FOSS
field is growing everyday into health, law and education.

If you do a little digging around you might find that FOSS libs exist in
many commercial products without dual lic. To the best of my knowledge.

Most of the examples of dual lic. I have seen have been driven from a
business perspective. So the business starts with its own code, open
sources a community version and every so often comes along, takes a
snapshot. The snapshot has some "other stuff" added (value?) and then
packaged, boxed, included with a manual and then sold to clients.

The duality comes in that sometimes this has proven most effective
without a dual licence, when you look at value add as a service, not a
product. The software is a product.

OSGEO has a clear mandate (IMHO) to support its community of developers,
however I don't see a clear mandate to MANAGE business models... Their
revenue stream should remain through membership and sponsorship, not
lic' of IP. Unless things have changed.

I think if some of these dual lic models existing means more access to
FOSS4G tools then that is an acceptable compromise. To dual lic. in the
hopes that some company will pick it up and champion it dismisses the
positive impact of MANY companies driving development. IMHO.

Too much forking of any project kills it. Privatizing or publicizing are
merely two forms of forking.

Not that this article has all the answers, but reading "The cathedral
and the Bazar" may shed some light on how open source has successfully
worked
(http://www.catb.org/~esr/writings/cathedral-bazaar/cathedral-bazaar/)
this can also be easily dismissed by counting up all the failed or dead
projects on sourceforge, but that is inline with any human endeavour, we
fail more often that we succeed, so lets make our success count.

To match the philosophical waxing perhaps follow up the reading with
some FOSS Economic Modeling http://perens.com/Articles/Economic.html

Between what is presented here and other sources I would say a single
license, a strong community and filling a niche for people will make the
geocoder and router libraries successful... IMHO

Cheers

> -----Original Message-----
> From: routergeocoder-bounces at lists.osgeo.org 
> [mailto:routergeocoder-bounces at lists.osgeo.org] On Behalf Of 
> Anton Patrushev
> Sent: Wednesday, December 03, 2008 00:46
> To: routergeocoder at lists.osgeo.org
> Subject: Re: [RouterGeocoder] Dual-licensing
> 
> Hi Steve,
> 
> Thanks, I knew you'll be the first to comment :)
> 
> Actually, my intention is not to find a way how to make money 
> with dual-licensing, but how to make it easy to use an Open 
> Source library with proprietary products. I think it would 
> work even if the project is immature - while nobody sees a 
> value, nobody buys it under terms of commercial license. 
> That's fair and doesn't prevent the developing community from 
> trying hard and making it better as an Open Source project. 
> But once somebody wants to use it together with proprietary 
> software - no problem, we have our second license for that.
> 
> I don't know if OSGeo will work here (but I believe that the 
> license fee can be judged as a contribution). That's what 
> OSGeo guys should tell us.
> 
> 
> Anton.
> _______________________________________________
> Routergeocoder mailing list
> Routergeocoder at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/routergeocoder
>