asymmetry?

[Michael R. Bernstein]

> From: Joel West <svosrp at gmail.com>
> 
> On 6:11 PM -0400 4/13/05, Michael R. Bernstein doth scribe:
> >
> >In the latter case, regardless of how many times a project forks
> 
> So a license that enables and allows forking is good? I thought that was the whole point of the GPL, to prevent forking.

*All* OSD compliant licenses allow forking.

This isn't the place to recap the topic "why FOSS projects don't fork more often", but I think it's self-evident that the disincentive for the ID inherent in an independent fork encourages (but doesn't mandate) good project stewardship.

> >Speaking only for myself, this isn't the sort of deal I would find attractive unless the modifications I intend to make are mostly trivial.
> 
> Well and that's the deal. The GPL allows several different companies to be in the Linux business. A dual license GPL only allows one company to be fully in, say, the MySQL business.

This isn't so. A dual-license business model rather can encourage a secondary market in value-adds that, should the ID want to include them in a proprietary release, the ID would have to license the modification from the contributor.

In a GPL dual-license model, no principal (whether the ID or a contributor) is in an inherently priviliged position WRT other people's code.

> So is it the license that's bad, bad, bad? Or is it the business model where some people get to make money off the software and not others?

I think you've drawn a false dichotomy here. First, no OSD compliant license actually precludes 'making money off the software'.

Nor do I think that a business model that is based on the *assumption* that the lion's share of benefits will flow to the project maintainer is bad.

I do have a problem with licenses that effectively enable a one-sided proprietary business model that is (potentially) parasitic on others' efforts, and which ensures that the ID never has to compete with non-proprietary forks (since they can always assimilate others' improvements and add them to their proprietary 'value added' version).

> Finally, if users get value from seeing the code and fixing short-term bugs, but have no desire to compete with the ID, isn't there some "open source" value in an asymmetric grant of rights? Open Source Definition #1, #2, #3 (the core of the OSD) would still apply.

Sure, *some* value, but not much (except perhaps PR). 

So, for some types of software (hosted web-applications, mostly) the software becomes attractive mainly to end-users who want to use it for free and competing low-end service providers who have no intention of making any modifications at all (except branding).

As a result, I am seeing increasing interest (on this list as well as elsewhere) in licenses that explicitly restrict the ability of others to make money from services based on the application as well (because the ID doesn't want to compete with low-end copycat services). Thankfully, this sort of license is beyond the OSD pale, but it shows why I consider the notion that an ID *ought* to be in any sort of priveleged position WRT contributors' code somewhat pernicious.

- Michael Bernstein