Re: Please choose a free license for fastml
Hello Tal,
I hope you are fine. I'm afraid you might have missed my mail from July
this year when I was trying to give reasons for a license change of your
program FASTML. We have now assembled all other preconditions to
package Gubbins for Debian and the license of FASTML that requires
contacting you in case of modifications is the only blocker. It would
be great if you would consider to answer my mail below to let us know
whether we might have chances for an integration of FASTML into a free
operating system like Debian and its derivatives (Ubuntu, BioLinux)
which is used by lots of biologists.
Thanks for considering
Andreas.
On Mon, Jul 06, 2015 at 04:15:03PM +0200, Andreas Tille wrote:
> Hi Tal,
>
> I'm writing you on behalf of the Debian Med team which has the objective
> to package all free software that might be relevant in medicine and
> biology for official Debian. You might like to have a look into the
> software in the field of biology inside Debian[1].
>
> Recently we received the request to package Gubbins from Sanger
> institute[2] (Andrew Page as one of the authors in CC) and I noticed
> that it depends from your software FASTML. (Actually Gubbins is using
> an adapted version of FASTML which is called fastml2 and maintained at
> Github[3].)
>
> To fulfill the dependency requirement of Gubbins we also would need to
> package FASTML but I noticed that the website states the following
> Copyright:
>
> To modify the code, or use parts of it for other purposes,
> permission is requested. Please contact Tal Pupko
> Please note that the use of the FASTML programs is for academic use
> only
>
> The choice of this license is a bit unfortunate since the first part
> would not enable a distribution by Debian since we ship only software
> that allows unrestricted modification to our users. For instance in
> case of a potentially detected security problem it would not be
> acceptable for the security team to ask for permission to fix the
> problem quickly since this would be an inacceptable delay.
>
> The second restriction you are applying is the academic use only.
> Software with such kind of license unfortunately needs to go to the so
> called non-free section in Debian which means the package does not
> belong to the official Debian distribution. The reason is that it is in
> conflict with item 5. and 6. of the Debian Free Software Guidelines[4].
> The non-free section has several consequences for the package itself
> (for instance it is excluded from typical quality assurance means like
> regular fresh builds and test suite runs) and for its dependencies - in
> this specific case Gubbins. Since Gubbins would depend from a package
> in non-free it also can not be included in the main Debian distribution
> even if it is covered by a free license. It could at best target at the
> so called "contrib" section (=free license but depending from package in
> non-free) with the same consequences that the usual quality means are
> not applied.
>
> I wonder whether you might be willing to reconsider the license of
> FASTML for the profit of Gubbins and to my personal experience also to
> the profit of FASTML itself. In the past we have discussed licensing
> issues with several authors of software (latest prominent example was
> Joe Felsenstein with phylip which is now delivered under BSD license).
> Besides the extra eyes that might look onto your code and provide
> enhancements the inclusion into Debian might bring you some extra
> users - or at least some estimation how many Debian users are using
> FASTML (see on [1] the numbers of the so called popularity contest.
>
> So it would be really great if you would consider a free license. Feel
> free to ask us for advise which one of the usual licenses might fit best
> in your case.
>
> Kind regards
>
> Andreas.
>
> [1] http://blends.debian.org/med/tasks/bio
> [2] https://sanger-pathogens.github.io/gubbins/
> [3] https://github.com/AidanDelaney/fastml2
> [4] https://www.debian.org/social_contract#guidelines
--
http://fam-tille.de
Reply to: