[lug] Open Source Software Licenses & the University

[Daniel Webb]

On Fri, 6 Dec 2002, Andrew Gilmore wrote:

> I would love to see the research center I used to work for open source
> its software, but they are very focused on trying to recover just the
> maintainance costs. These costs are about 200K/year, and they are
> licensing the software to try to recover them.
>
> That, for me, is the question I have yet to find a good answer for. How
> do you pay the costs of maintainance of a large program? I'm talking
> hardware and people time, mostly.
>
> This may not be as big of a problem when the program has very wide
> appeal, but the software I worked on is large, complicated, and has a
> small set of applicability.
>
> Comments on this? If this had started as an open source project, it
> would perhaps be less difficult. Note that I am not talking about the
> costs of new development, which are running >1M a year.

  That is a tough problem, no doubt.  For new development on smaller
programs, grants and cheap label (graduate students) can make it possible
to open source it.  For me, the reputation boost from a good piece of
software that is known and used would be worth much more than 25% of
royalties, unless of course we're talking hundreds of thousands of dollars
;)

  Open source can work for maintenance only if there are people
volunteering to do the maintenance, which in your case sounds not very
likely.  It works well for things like Apache, where companies have a
vested interest in keeping it nice.

  The thing I never understood is why companies haven't figured out how
much money they would save if they just figure out a way to work
cooperatively on things like software.  Imagine how many times over
something like Microsoft Word could have been developed with the amount of
money that companies (or the government) has spend on Word.