[OSM-talk] THIS is the kind of enthusiasm some would reject

[moltonel 3x Combo]

On 14/09/2015, EthnicFood IsGreat <ethnicfoodisgreat at gmail.com> wrote:
> I guess we're asking that an exception to the "verifiable features only" rule
> be made for these features.

IMHO the exception that you are asking for is not to the "verifyable
only" rule but to the "presently existing" rule. All the
abandoned/dismantled railroads I've seen in OSM were verifyably
"previously existing" but also (where the conflict arrises as far as
I'm concerned) verifyably "no longer present".

This is not a rejection of your plea, just trying to make sure of what
we are talking about.

> Simply confining abandoned railroad
> features to OHM is not a good solution, because without being able to
> view them in the context of existing features, they lose a lot of
> their value.

Agreed, OHM is currently not very usable.



I've suggested that early on, and again in my latest reply to Russ : I
think that maping the past in OSM would be acceptable, if done
properly. Some kind of "OHM done right". Doing things really right
might require a modification of the data model, a cross-db
synbchronisation tool, or some other cool technology... But that's
just too far off, too hypothetical. The next best thing is a tagging
system for the past.

If it wasn't clear already, railway=dismantled, end_date, or any
system that mixes past and present in the same namespace is IMHO not
acceptable. Consumers, editors and tools should be able to filter out
historical data with a simple rule. I've suggested using "past:" as a
key prefix, with an optional " @ date - range" as a value suffix.
Didn't see any reply, what do people think ?

As for opening the floodgates of historical mapping, I do not like it
from a very personal POV, but I can recognise that there is a need,
that OSM might be the best tool to fill that need, and that it might
ultimately strengthen the poject. I just hope (and believe and work to
make it true) that it won't be too much of a nuisance to my usecase.
And if we do open up to maping the past, I don't think that it should
be reserved to railroads.

I've argued against maping no-longer existing railroads in way too
many emails at this stage, but I suggested this escape route early on.
Nobody picked it up but I think that's the only thing that currently
stands a chance of reaching consensus. EthnicFoodIsGreat, can you see
the working compromise that Russ cannot ?

That's it for me, bye bye railroad thread, I hope. Of course I'm only
one contributor, not a highly prolific or influential one, not an
authority, just a voice. Others have been less noisy but more dogmatic
than me on the subject. The community as a whole must decide wether
"we map the present" is still a hard OSM rule.