[Talk-GB] OSM Analysis New Data and bot

[Ed Avis]

SomeoneElse <lists at ...> writes:

>>I think we all agree that reaching 100% completeness with a collection 
>>of people
>>doing diverse surveying methods (and even aerial tracing) is much better than
>>reaching 90% completeness by importing.

>Grr.  "100% road name completion" has become in this thread "100% 
>completeness".

That's why I was careful to talk about 90% completion as being the maximum that
can be achieved by importing from OS.  The last 10% of completion relative to
what's on the ground can only be achieved by survey combined with high-quality
aerial photos.  Of course these numbers are arbitrary: a footpaths maniac might
consider the map only 50% complete if footpaths were missing even though
everything else were perfect.

>BTW sorry Ed - I'm not attacking you directly - it's just that the 
>general thrust in some of the mails here seems to be that mapping the 
>names of roads is all that matters.

I think that in this particular thread (and it is hard to stop discussions
wandering or to keep track of what was originally proposed) we were discussing
the addition of road names to existing roads from OS Locator.  There is far more
to a map than road names but they are an important part, and something that OS
could help us with.

We're trying to explicitly exclude other map features from discussion to allay
people's fears about large-scale trampling all over the map adding features that
don't exist on the ground, or deleting people's existing work, or other bad
decisions that certainly have affected imports made in the past, but are not an
inevitable feature of every import.

At one time there was also a proposal for a different 'OS bot', which would add
rural roads to blank areas of the map.  I think that also has something to
recommend it and some points against, but it would need to be discussed
separately from street names.

-- 
Ed Avis <eda at waniasset.com>
of the map