Subject: Re: For Approval: Open Source Hardware License
From: Rick Moen <>
Date: Fri, 6 Jul 2007 16:51:12 -0700

Quoting Allison Randal (

> In terms of practical applications, I see two extremes, and a range in 
> between:

I note with approval Michael Tiemann's polite but judicious reaction
when asked, recently, once again whether OSI shouldn't evaluate one of
Microsoft's licences for OSD-compliance.  He said, paraphrasing:

"Is someone actually using the licence on an actual piece of software?
If so, would that person consider submitting it?"

I know coders, and so I know they have a tendency to want to test
anything that looks like a Turing machine against a variety of inputs.  
The OSI certification process has, alas, looked to many people like a
Turing machine, with the result that they try to hurl goofy licences 
at it for sundry theoretical purposes or bits of intellectual
thumb-sucking^W^W^Wexploration -- most often licences not _even_
actually in use for real pieces of software.

(And then, of course, we hear blatant non-sequitur reasoning like "Well, 
we borrowed that language from the Attribution Assurance Licence, and,
surely, if we frankensteined together hunks taken only from
OSD-compliant licences, the result must be, too."  But I digress.)

Getting back to my earlier point:  I greatly doubt that OSI would refuse
to examine a licence for no better reason than its title (or text) 
making reference to hardware.  However, expecting that the licence
be one _actually used_ for software seems entirely reasonable, and
a good heuristic for filtering out time-wasting exercises in theory.

Time-wasting exercises in theory would include -- not to put too fine a
point on it -- determining how "completely separate" hardware and
software are.  That might be a fascinating discussion, but it has
no obvious relevance to OSI certification.

Cheers,                                     The Viking's Reminder:
Rick Moen                                   Pillage first, _then_ burn.