[OSM-legal-talk] Substantial meaning

Frederik Ramm frederik at remote.org
Thu Apr 23 23:34:06 BST 2009


Hi,

SteveC wrote:
> Has there been any discussion on what people here feel 'substantial'
> means in the context of the definitions of the ODbL? I've banged
> around the wiki looking but might might have missed it.

It hasn't been discussed a lot. I guess you have read the 
often-referred-to pp28-32 of 
http://edina.ac.uk/projects/grade/gradeDigitalRightsIssues.pdf which of 
course are not specific to OSM.

RichardF once summarised;
 > Dr Waelde also notes that under EU database right it's not just a
 > matter of percentage, but also of a "qualitative analysis". To me it
 > appears clear that a single street name isn't substantial, the whole
 > of Cambridge is. We could spend hours discussing this but I suspect it
 > comes down to "don't take the piss".

I could think of a number of ways of defining substantialness. Any 
workable definition must

1. be compatible with the EU database directive
2. be absolute (no "more than 0.01% of data means substantial")
3. be practical

For the sake of 2. and 3., I would define something like this:

"A substantial extract is an extract that covers more than one square 
kilometre *and* more than 25 features, where a "feature" is either a way 
with all its nodes and tags, or a single tagged node."

This would mean that an area of 1000 x 1000 metres could always be fully 
extracted as insubstantial and used for any purpose. That's roughly the 
size of Green Park in London. It is not something of a huge value; it is 
something that can be mapped in an hour, or make it half a day if you 
want all house numbers and pubs in a densely populated area. For a 
densely populated area, the 25-feature constraint will always be met 
within a square kilometre and thus ineffective. For a sparsely populated 
area, it might be possible to fully extract a larger area as long as it 
contains less than 25 features.

The 25-feature constraint becomes interesting when someone does queries; 
one could extract the 25 highest mountains or the 25 largest buildings 
or 25 catholic churches that are within 100 metres of a pub. Because 
whole ways are allowed, the 25-feature constraint might even allow the 
extraction of the whole motorway network of a small country.

When you think about what you can do with 25 features, always bear in 
mind that you must not combine multiple insubstantial extracts - if you 
manage to extract the motorway network of a small country in 25 features 
then you cannot also get the coastline or the primary roads, even if 
someone else extracts them for you or if you do it another time.

Of course the numbers can be changed but I think that a definition like 
this is easy to understand and apply. There is some potential for abuse 
because someone could quickly connect all things of interest to him into 
one large way and then extract that but I think such behaviour could be 
ruled out by proper wording.

Our goal in defining "substantial" must not be to allow as little as 
possible. The database directive allows free use of insubstantial 
extracts, and says the database owner must not disallow such use. If we 
try to circumvent the rule by issuing a definition of "substantial" that 
basically makes everything that is useful also substantial then we are 
in breach of the database directive.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"




More information about the legal-talk mailing list