For Approval: Academic Citing License

[Rick Moen]

Quoting John Cowan (jcowan at reutershealth.com):

> Only 948.  Public domain projects don't have OSI-approved licenses
> for the very good reason that they aren't licensed at all.  But 
> they are clearly Open Source.

You are (obviously) correct in the abstract, but I investigated quite a
number of these as I was vetting PalmOS codebases for inclusion in my
collection of all open-source PalmOS packages
(http://linuxmafia.com/palmos/).  It turned out that basically _all_ of
the packages I checked had various encumbrances by upstream copyright
rights (e.g., prior authors), but that, in each case, basically the
project lead developer was using "public domain" as shorthand to mean "I
don't understand licensing and/or want the problem to go away".

> I investigated just a trivial number of those projects.  First of all,
> some have (or claim to have) both OSI-approved and other/proprietary
> licenses.  Filtering those out (manually), I found miscategorized
> projects, projects with Creative Commons licenses, and idiosyncratic
> but plainly Open Source licenses like the license of gnuplot (which
> is BSD-ish but allows only patches, like the QPL).  So assuming that
> all 948 are in fact under proprietary licenses is incorrect.

Actually, all _three_ of three of the categories I reported are riddled
with miscategorised projects, etc.  So, you could equally well have objected
that "assuming that all 55076 projects are OSI Approved is incorrect".

I called this, too, to SourceForge.net management, and suggested that at
least a cursory review that submitted projects' licensing was being
accurately recorded might be a good idea.  They said no.

-- 
Cheers,                        Open-source SourceForge retakes the lead:
Rick Moen                      http://gforge.org/  Thank you, Tim Perdue.
rick at linuxmafia.com