www-legal-discuss mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Henri Yandell (JIRA)" <j...@apache.org>
Subject [jira] [Comment Edited] (LEGAL-183) Is Apache software allowed to distribute the EPSG database?
Date Sat, 18 Jul 2015 02:55:04 GMT

    [ https://issues.apache.org/jira/browse/LEGAL-183?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14630750#comment-14630750
] 

Henri Yandell edited comment on LEGAL-183 at 7/18/15 2:55 AM:
--------------------------------------------------------------

Jochen - users would expect that what would? That the EPSG Dataset would affect our Apache
Licensed code?

Martin - The content of other directories would stay under their original license (afaik this
is always true, regardless of the OSS licenses in question). Sure a lot of Apache 2.0, much
of it contributed to the Apache Software Foundation, but also MIT, BSD, CDDL etc. It's not
abnormal for pieces of a download to not be under AL 2.0.

The challenge here is identifying what the problem is. If we look at the guiding criteria
(on resolved.html), this would appear to pass all three. It's a bit surprising in places,
but, for example, I can't see an obvious reason why it fails the Open Source Definition. Unless
a person trying to sell the original bytes is being discriminated against, or their field
of endeavour is.

I think the problem is that we have a piece of an Apache product that, while it is within
the Apache product, behaves roughly in line with the Apache License, but when you remove that
piece no longer behaves in line with the Apache License. That's the moment of user surprise,
and the question is whether that user surprise is unreasonable.

If you look at the binary-only use cases, it discusses this same issue: "By attaching a prominent
label to the distribution and requiring an explicit action by the user to get the reciprocally-licensed
source, users are less likely to be unaware of restrictions significantly different from those
of the Apache License. "

If this was included as a binary file, it would seem to fit the above. If it's a series of
SQL scripts, then the question is what can be done to reduce user surprise. 

Four paths jump to mind:

1) Include in the current source/binary with notices to the point of the licensing difference.
2) Include as an optional download from Apache, that calls out specifically the difference.
3) Include as an optional download from elsewhere.
4) Point to the EPSG themselves along with instructions. This has the advantage of meaning
users will have the latest data.

More I think about it (and length of this comment hopefully shows a wee bit of pondering)
- I favour #2 or #4.


was (Author: bayard):

Jochen - users would expect that what would? That the EPSG Dataset would affect our Apache
Licensed code?

Martin - The content of other directories would stay under their original license (afaik this
is always true, regardless of the OSS licenses in question). Sure a lot of Apache 2.0, much
of it contributed to the Apache Software Foundation, but also MIT, BSD, CDDL etc. It's not
abnormal for pieces of a download to not be under AL 2.0.

The challenge here is identifying what the problem is. If we look at the guiding criteria
(on resolved.html), this would appear to pass all three. It's a bit surprising in places,
but, for example, I can't see an obvious reason why it fails the Open Source Definition. Unless
a person trying to sell the original bytes is being discriminated against, or their field
of endeavour is.

I think the problem is that we have a piece of an Apache product that, while it is within
the Apache product, behaves roughly in line with the Apache License, but when you remove that
piece no longer behaves in line with the Apache License. That's the moment of user surprise,
and the question is whether that user surprise is unreasonable.

If you look at the binary-only use cases, it discusses this same issue: "By attaching a prominent
label to the distribution and requiring an explicit action by the user to get the reciprocally-licensed
source, users are less likely to be unaware of restrictions significantly different from those
of the Apache License. "

If this was included as a binary file, it would seem to fit the above. If it's a series of
SQL scripts, then the question is what can be done to reduce user surprise. 

Three paths jump to mind:

1) Include in the current source/binary with notices to the point of the licensing difference.
2) Include as an optional download from Apache, that calls out specifically the difference.
3) Include as an optional download from elsewhere.
4) Point to the EPSG themselves along with instructions. This has the advantage of meaning
users will have the latest data.

More I think about it (and length of this comment hopefully shows a wee bit of pondering)
- I favour #2 or #4.

> Is Apache software allowed to distribute the EPSG database?
> -----------------------------------------------------------
>
>                 Key: LEGAL-183
>                 URL: https://issues.apache.org/jira/browse/LEGAL-183
>             Project: Legal Discuss
>          Issue Type: Question
>            Reporter: Martin Desruisseaux
>
> The EPSG geodetic parameter dataset is a freely available structured repository of data
used in geospatial applications. This dataset is of critical importance to the Apache Spatial
Information System (SIS) project. The distribution terms have some conditions for which I
would like advices. But before, some context:
> * EPSG is only data, no software.
> * The EPSG database is maintained by the "International Association of Oil & Gas
Producers" (OGP) and members are big companies like Shell.
> * Oil & Gas producers maintain and provide the EPSG database free of charge because
the cost of installing a drilling platform in the wrong location is too high. Since they rely
on maps and data produced by various actors (national map agencies, etc.), it is in their
best interest that those actors have access to the most accurate Map Projection definitions
when they create their data.
> * The EPSG database, or something equivalent, is critical to a Spatial Information System.
Apache SIS without EPSG would probably lost a lot of its interest. For example EPSG codes
are the the-facto standard in most Web Map Services.
> * I'm not aware of any freely available alternative to the EPSG database, and it would
be impossible for us to create one.
> * Open source and commercial products like Proj.4 (MIT license), PostGIS, GDAL, MapServer,
Geoserver (GPL license), ESRI, Oracle Spatial and many other all include the EPSG database
(or a subset of it) in derived forms. I think that basically all major GIS products around
the world include the EPSG database in one form or the other.
> The Term of Uses are there:
> http://www.epsg-registry.org/help/xml/Terms_Of_Use.html
> We can read "The EPSG Facilities are published by OGP at no charge. Distribution for
profit is forbidden" followed by "The data may be included in any commercial package provided
that any commerciality is based on value added by the provider and not on a value ascribed
to the EPSG Dataset which is made available at no charge". My interpretation is that anyone
is allowed to distribute Apache SIS + EPSG for profit, but would not be allowed to extract
the EPSG tables from Apache SIS and sell only them, without anything else. In other words,
the restriction would not apply on Apache SIS but emerge on what is remaining after one deleted
everything from Apache. Indeed, I have never hear about any legal dispute regarding other
open source projects that distribute EPSG data for years.
> We can read "Ownership of the EPSG Dataset by OGP must be acknowledged in any publication
or transmission (by whatever means) thereof (including permitted modifications)". Is it something
that we can handle in the NOTICE file?
> Next, the Term of Uses page lists the kind of modifications that users are allowed to
do and said "No data that has been modified other than as permitted in these Terms of Use
shall be attributed to the EPSG Dataset". My understanding is that they are protecting the
credibility of their name, in the same way that (I presume) the Apache foundation would not
let a badly broken fork of an Apache project call itself "Apache Foo". Better example: if
someone copy the ASCII table, then modify the definition of ASCII code 65 so that it stands
for something else than "A", then this is not the ASCII table anymore. It may have legitimate
use, but this is something else than ASCII. If we replace "ASCII" by "EPSG", we get the same
situation. E.g. EPSG code 4326 stands for "World Geodetic System 1984". If someone change
that definition, then this is not anymore the EPSG definition of code 4326.
> So would Apache SIS be allowed to bundle the EPSG database in its distribution, with
a restriction that emerge only if the user delete all Apache work, and with a restriction
to the changes that are allowed in the database? If the later point is problematic, would
if be an acceptable workaround to use a mechanism that verify the database integrity (e.g.
using a checksum) and automatically rename "EPSG" in something else if a change is detected?



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: legal-discuss-unsubscribe@apache.org
For additional commands, e-mail: legal-discuss-help@apache.org


Mime
View raw message