Bug#940185: src:pywps: Debian/copyright needs update

[Scott Kitterman]

On Friday, September 13, 2019 11:56:30 AM EDT Sebastiaan Couwenberg wrote:
> Hi Scott,
> 
> Thanks for finally reviewing pywps.
> 
> On 9/13/19 5:23 PM, Scott Kitterman wrote:
> > One of our ftp-trainees reviewed your package and made the following
> > observations.
> 
> It seems that the process is broken.
> 
> This is far from the first time where an anonymous ftp-trainee commented
> on a package in NEW, but no ftp-master acted on this.
> 
> What process does ftp-master use to review ftp-trainee comments?

The delay this time was caused by it being in an incorrectly locked state, not 
because there was a pending trainee comment.  The stale lock problem happens 
rarely enough that it took me some time to remember how to resolve it.

> > Compiled works are present in tests/ and it does not appear this data can
> > be rebuilt from the source package provided.  Please remove and repack
> > the tarball to remove any such artifacts.
> 
> Can you or or the anonymous ftp-trainee clarify which files they
> consider to be "Compiled works"?

I'll ask, but unfortunately the note didn't include it.  If you don't see 
anything there, then I wouldn't worry about it.

> There are data files under tests/data & tests/requests used in various
> tests. Why should these need to be rebuilt if those are the files in
> question?

There's no need to actually rebuild them.  We do generally require that it be 
possible to rebuild them from tools in Debian.  Of course the best way to know 
that you actually can rebuild them is to do so during the package build, but 
it's not required.

> > Data in pywps/schemas/geojson/ has no licensing information.
> 
> Its upstream states: "[...] either of the AFL or BSD license", but not
> which version. I've contacted the author to request clarification.

Thanks.  I took a look and those files seem to be based on https://geojson.org/
geojson-spec.html which is CC BY 3.0 US, so I have doubts.  

> > Who holds an actual copyright is confusing:
> > - LICENSE.txt claims copyright by "PyWPS Development Team"
> > - All source claims copyright by "Open Source Geospatial Foundation"
> 
> Why does this matter?
> 
> debian/copyright includes the holders as listed in LICENSE.txt and the
> sources. Are you saying that's wrong?
> 
> > - Source also (incorrectly) uses "and others" as a copyright holder
> 
> Why is this incorrect?
> 
> The sources have copyright statements like this:

>  Copyright 2018 Open Source Geospatial Foundation and others
>  licensed under MIT, Please consult LICENSE.txt for details
> 
> That's what's reflected in debian/copyright.

I've reviewed it more carefully now and I agree with you that it's fine as is.

> > - d/copyright claims "PyWPS Project Steering Committee" is a copyright
> > holder, but is> 
> >   not represented in source
> 
> It used to be a copyright holder, see:
> 
>  debian/share/pywps/processes/sayhello.py
> 
> > Files provided in d/patches have a copyright holder that is not present in
> > d/copyright This appears to be the new maintainer, they should be
> > included in the debian/* paragraph.
> The patches are trivial and cannot be copyrighted in my understanding. I
> wave any copyright claims on them if they can.

I think that's fine.

> Depending on how strict ftp-master is on the data files issue, it may be
> better to just remove this package from Debian as I don't use it myself
> and just co-maintain it because it's also included in OSGeoLive.

I should have checked more carefully before passing on all the note as it's 
not all correct.  The missing license for the schema is correctly serious and 
should be resolved.  I can't tell you if it's worth keeping in Debian or not, 
but I don't think (now that I've looked harder) there's a lot of work to do to 
resolve this.

Scott K