[Imports] Permission Statement

[Brian May]

On 3/12/2020 7:30 PM, Greg Troxel wrote:
> Brian May <bmay at mapwise.com> writes:
>
>> We (OSM-US) should really have a state by state guide on open records
>> laws per state. That would make the "license check" process
>> easier. Look up the state, get an answer. At least a quick answer to
>> start from.
> Agreed in principle, but I think "open records" is off base.  The real
> issue is copyright, not whether we have a legal right to demand map data
> held by the government.
>
No, its not off base. Here's a reference for you: 
https://en.wikipedia.org/wiki/Microdecisions,_Inc._v._Skinner
>
>> This site seems to do a decent job of explaining public records per
>> state in a standardized format -
>> https://www.rcfp.org/open-government-guide/
> Yes, but for the Massachusetts page (where I'm reasonably familiar with
> the rules, having been subjected to mandatory trainign), it does not
> mention copyright at all that I noticed.

This is the grey area and sounds like up for debate. However, as I 
understand it, the omission of copyright assertion is a waiving of 
copyright assertion. I am informed by the Microdecisions vs. Skinner case:

"Additionally, the Court confirmed "Florida's Constitution and its 
statutes do not permit public records to be copyrighted unless the 
legislature specifically states they can be."

But there could be further nuances I am missing.

>
>> But the state statutes are where you find the details and final
>> answers. If copyright is not mentioned at all in the statutes, then
>> its public domain, right?
> I completely fail to understand how you conclude that.  These state laws
> and federal copyright laws act independently.  Perhaps more importantly,
> creative works (other than old or by the federal government) simply are
> copyrighted.
Not according to the Skinner case, in Florida at least.
>
> In MA, one can get a copy of public records, and one can disclose the
> information, but I see no right to copy, distribute, or to create
> derivative works.  As an example my Conservation Commission publishes
> trail maps.  Copyright is held by the town under the work-for-hire
> doctrine, and there's no license to copy granted, other than the "it's
> ok to print this out and use it" implied grant.
So are you saying since MA public records laws do not specifically state 
that agencies can copyright data, that towns default to everything is 
copyrighted? I don't think so. But this is a grey area that needs to be 
further explored, maybe state by state. What is the spirit of the law in MA?
>
>> In my view, if the state statutes do not allow governments operating
>> within the state to copyright public records OR don't mention it at
>> all, then you don't need to ask the government data provider for
>> permission at all, like in Florida.
> That view might be right given enough case law, but it is remarkable and
> in need of citations.
Skinner
>
> The phrase "allow governments to copyright" does not make sense since
> that's not how copyright works.  If it were "prohibits the government
> from enforcing copyright" and "prohibits the government from acquiring
> data unless the government is authorized to grant licenses to
> redistribute under CC0", that's something else.
Yes it does make sense. States decide whether agencies can copyright 
data. Some specifically state you can, others state you can't and some 
don't mention copyright at all. The ones that don't mention copyright 
usually talk a lot about how data is open and there's no restrictions on 
the data, which means no copyright. But if they don't specifically 
address copyright, it muddies the water.
>
>> In this case, sounds like Colorado allows governments to copyright,
>> but this government chose not to. From what I've seen, the only time
>> they do try and exert copyright is if they are trying to make money on
>> the data. And even then, it seems like there's ways of skirting that,
>> e.g. grabbing data from a public facing ArcGIS Server.
> I don't follow "grabbing".  If there isn't a license, the fact that you
> can get at the data doesn't seem to matter.  That would be like copying
> frmo google maps because you can.
Grabbing, aka writing code to extract data from a server. There has been 
previous discussion on this topic, specifically regarding OpenAddresses. 
I'm pretty sure they concluded copying data from a government server 
that is publicly exposed to the Internet means any copyrighting has been 
waived. That could be wrong, I don't know. Google is an obvious case as 
are most private companies. We're talking governments here.
>
>> This seems like a topic for the OSM US Board to provide direction on.
> Agreed.  I think most states intend to grant
> CC0/PD-with-attribution-requested terms, but the agencies are not
> necessarily adequately copyright-pedantic, and a high-level contact
> could be useful.
>
> (I and others in .ma.us have discussed licensing with MassGIS and the
> conversations have been highly reasonable.)
>
>
 From what I've seen over the years, governments are moving more and 
more open, but if the government started out with some overly protective 
laws over "their", I mean the public's data, it can take a long time for 
things to change. Georgia is an interesting example. The statutes 
specifically cited GIS data as copyrightable, if the agency so desired, 
and allowed charging whatever fees they wanted. So you have this 
mish-mash of counties that are wide open / public domain and others that 
charge $20k for a copy of parcel data and say its copyrighted. And at 
the same time, private vendors are selling the same data for a few 
hundred bucks. Not sure what is going on there. But I've talked to high 
level people there and no-one likes the current situation, except the 
agencies who can squeeze out a few $20k checks from big companies, and 
its difficult to get such a mundane issue on the agenda for the 
legislature to debate. But everyone knows the data needs to be more 
open. Agencies can't even share data between some agencies.

We should probably move this discussion to imports-us, since its US 
specific - agree?

Brian