🚀 go-pugleaf

barmar@think.com (Barry Margolin) writes:
>>Compare two hypothetical situtations:
>
>>Situtation 1:
> [Can't use Microsoft source because it's not available.]
>>Situtation 2:
> [Can't use FSF source because hired to write a proprietary program.]
>
>>Looks the same to me.
>
>The difference is that in situation 2, your employers could conceivably
>decide to copyleft the resulting program, getting around the problem.  But
>it's much harder to get around the problem in situation 1.

Actually, they could probably throw money at Microsoft until a license
to source falls out.

>In many situations it might even make economic sense to make software
>non-proprietary so that you can incorporate FSF code.  Suppose it saves you
>a man-year of work in development, and a man-month per year of ongoing
>maintenance, and you charge $5K for the program.  Calling a man-year $120K,
>you break even if the non-proprietariness of the program results in 26
>fewer sales the first year and 2 fewer sales/year in the future, and you
>come out ahead if there were fewer sales lost.  For specialized
>applications, those numbers may not be a problem.

In short, I have to consider the cost of a source license from FSF, just as
I would if I wanted to use Microsoft source.  The FSF license just requires
different economic analysis, and its cost is harder to calculate.

But there is hope.  A few posts ago, someone mentioned a public domain
Scheme they are going to release, and I got email a while back from someone
who was finishing up a public domain C compiler.  TeX and X are quite
unrestricted.  So the list of software that is almost really free is
growing.  Now if I could just get time to work on my public domain OS...

--Tim Smith