[OSM-legal-talk] ODbL: How obscure/inaccessible can published algorithms be?

[Matt Amos]

On Sat, Dec 12, 2009 at 8:20 PM, 80n <80n80n at gmail.com> wrote:
> On Sat, Dec 12, 2009 at 8:09 PM, Matt Amos <zerebubuth at gmail.com> wrote:
>>
>> On Sat, Dec 12, 2009 at 6:30 PM, 80n <80n80n at gmail.com> wrote:
>> > On Sat, Dec 12, 2009 at 4:13 PM, Matt Amos <zerebubuth at gmail.com> wrote:
>> >> On Sat, Dec 12, 2009 at 3:43 PM, 80n <80n80n at gmail.com> wrote:
>> >> > What kind of duck test can you use to be sure that a derived database
>> >> > is
>> >> > involved in the process?
>> >>
>> >> if you suspect that someone is using a derived database, and isn't
>> >> making an offer of it, you are suspecting that they are in breach of
>> >> the ODbL. this can be tested by asking the company and, if they don't
>> >> provide a satisfactory response, legal proceedings could follow.
>> >>
>> > Exactly.  On what grounds would you suspect that either company was
>> > using a
>> > derived database?
>>
>> by whatever grounds you'd suspect that a company was providing
>> services based on AGPL software, or distributing a binary
>> incorporating GPL software - gut instinct ;-)
>
> In the scenario I described you'd have no grounds for suspicion.

yes. and you'd have no grounds for suspicion if a company were using
modified AGPL software, so you have to rely on gut instinct.

>> let's assume it's known that this company is definitely using OSM data
>> - determining that can be difficult, depending on exactly what it is
>> they're doing with the data. in general, it's very difficult to do
>> anything directly from the planet file alone, so i'd suspect that any
>> company doing anything with OSM data has a derived database of some
>> kind and, if there's no offer evident on their site, i'd contact them
>> about it.
>>
> You're going to do that for every single organisation that publishes some
> kind of OSM data?!!  Good luck.

no, i'm going to assume that most organisations and are going to read
the license and abide by it, the same way they'd read and abide by any
other open source/content license.

>> it's a similar situation to looking at a site and thinking they're
>> using OSM data to render a map, without respecting the license. it's
>> entirely possible that they have some other data source, or have
>> collected the data themselves. so it's a gut instinct whether or not
>> you think any of the data has come from OSM and should be followed up.
>
> Not at all.  The lack of attribution is self evident.  A derived database is
> not at all evident.

company A: publishing a map with no attribution, but it's at least
partly derived from OSM.
company B: publishing a map with no attribution and it's all their own data.

a lack of attribution is evident, but whether they're using OSM data
isn't. you have no grounds for suspicion, but you might have a gut
instinct. what do you do?

cheers,

matt