[reposting to list]
The perfect is the enemy of the good.... things will rarely be 100%
right. Be satisfied with "good enough". Now, if someone could just come
up with a workable definition of "good enough".... Or failing that, an
objective test to show whether A is better (closer to perfection,
without needing to define that) than B....
--colin
On 2017-06-11 19:30, James wrote:
> Well mapping from imagery would be "wrong" then as a lot of times it's
> misaligned with the "real world" short of buying a 100 000$ gps receiver with
> 1cm accuracy, there will always be "accuracy problems"
>
> On Jun 11, 2017 1:15 PM, "Colin Smale" <colin.sm...@xs4all.nl> wrote:
>
> On 2017-06-11 18:18, Eric Gillet wrote:
>
> On Thu, Jun 8, 2017 at 10:58 AM, Frederik Ramm <frede...@remote.org> wrote:
> I am concerned that reckless users will use your tool to basically go
> over the planet in a "task manager" fashion, running the matching for
> square after square, selecting all matches and hitting upload. (On the
> basis of "hey if 95% of matches are good then I am improving OSM, right
> - someone local can sort out the 5% bad apples".)
>
> Should changesets which are less than 95% correct be disallowed on OSM ? That
> would block a lot of contributions !
The easier it is to determine objectively whether data is "correct" or
not, the higher the bar should be. For example, for highway=motorway we
should expect nothing less than 100%. For many others, the criteria are
just too fuzzy to say with reasonable confidence if data is "correct" or
not so any discussion about tolerances is premature.
--colin
_______________________________________________
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk [1]
Links:
------
[1] https://lists.openstreetmap.org/listinfo/talk
_______________________________________________
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk