On Thu, 7 Jul 2011, Frederik Ramm wrote:
On 07/02/11 14:50, Stephan Knauss wrote:
The doCheck() in BasicLicenseCheck is in my opinion not right regarding
nodes that have no tags. As all information these nodes carry is their
existence and coordinates I see no reason why any other status than the
one of the last editor of that node could be relevant.
It has just occurred to me that there's a serious flaw in that
logic/implementation. While we could debate whether or not we might be able
to keep a node that has never been tagged, your check only looks at whether
or not the node is currently tagged.
This means that if someone created a node using e.g. a "created_by=blah blah"
tag, and someone else later removed that tag, your change would treat this as
if the new person had somehow acquired the copyright for that node which is
clearly wrong, and asking for trouble!
That would be wrong. I'm not sure if he described it that way, but the
condition must be:
* no tags of previous editors are included in the current node (which is
true for untagged nodes)
and
* the position of the node is different.
I think Stephan assumed that the second condition is always true for
untagged nodes, but your example shows that this must not be.
Essentially the idea is that if no original information remains, then also
the original authors should have no influence. Untagged nodes are probably
the only case, where detecting this is possible in a reliable way.
Ciao
--
http://www.dstoecker.eu/ (PGP key available)
_______________________________________________
josm-dev mailing list
[email protected]
http://lists.openstreetmap.org/listinfo/josm-dev
