[Tagging] RFC 2 - addr:interval
pelderson at gmail.com
Thu Jan 7 21:04:09 UTC 2021
> I did see that some of the examples given led me to
> conclude that regexes would be one way of doing it.
Possibly, but if you are not proposing to do that, then nobody is, I think.
I think it would be overdoing it just a little.
> Take 27-35. Is it 27, 28, 29,,..,35? Or 27, 29, 31, 33, 35? Or the
> address 27-35, with the address of the next building being 27-37? Of
> the atomic address 27-35 with the address of the next building
> being 29-35? I haven't seen any example given here of that
> .last one, but I've seen the others. I can conceive of weirder
How is 27-35 interpreted now?
Would it alter the evaluation of this value if you would e.g. specify that
a suffix (2) may be added to a housenumber string to indicate that it is a
range with increment 2?
Would it alter the evaluation if e.g. single quoting was an added option
to force interpretation as a single housenumber, rather than a range?
You see, if it is ambiguous now, it remains ambiguous even if these options
were there. If one data user interprets it one way, another the other way,
and the third just ignores it, you lose nothing by adding the options.
Nothing is lost by not using extra options, but you may gain something by
> If we're going to be able to specify these in a way that software
> tools can understand, indicating which parts change and which
> parts don't, regexes would be a precise, concise way to do it.
> They'd just be very difficult for most ordinary mappers to enter
So, let's not go there. Let's go for extra's that add something if
applied, without demanding retagging of existing data.
> Tagging mailing list
> Tagging at openstreetmap.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Tagging