[OSM-talk] amenity=doctor or amenity=doctors ? [tagging]

[brendan barrett]

I have to say i'm siding with those in favour of validation as well.

Not for everything, but at the very least some kind of "spell
checking" utility when uploading from the editors. I've seen a number
of spelling mistakes when "residential" was clearly the intended tag
value. I have nearly uploaded incorrectly spelled residential tags as
well from Merkaartor (even though the correct spelling
auto-completes... if you just type in the box, sometimes you mess it
up - my own stupid user error). In fact I corrected a number of
clearly inaccurately spelled tags the other day (after finding them in
my local DB). Would a "spell check" type utility that could point out
possible errors perhaps make sense? I'm thinking of a "Did you mean"
kind of dialog used to validate the data before upload (like MS Word /
Open Office). The user can thus ignore the validation (add it to their
"dictionary" even) if they feel strongly about their implementation,
but it might catch some blatant errors.

We could also include OGC validation later on (like a grammar check -
to take the analogy further:P) if the users wanted to (I know the MS
Sql Server users of OSM would love this... even though we're in the
minority:P)

Regards,
Brendan

On Tue, Feb 24, 2009 at 7:47 PM, Guenther Meyer <d.s.e at sordidmusic.com> wrote:
> Am Dienstag 24 Februar 2009 schrieb Dave Stubbs:
>> > 1. ... every application trying to use the data has to deal with several
>> > taggings for the same thing. that's an unnecessary waste of resources.
>> > a script running on the database can minimize this waste, and
>> > furthermore, can fix typos in tags, which are also found all over the
>> > database.
>>
>> 1) Typos are a different beast entirely. If you've applied proper
>> context then some sort of fix bot might work. And some kind of
>> validation tool would be even better.
>>
> yes, this is something different, but technically very similar.
> you mean validation in the editors?
> ok, but not during editing but before the upload/commit.
>
>> 2) OMG two tags!!1! Trust me when I say this is a trivial thing to
>> include when you consider some of the other random tagging variations
>> that people keep voting in.
>>
> it may be trivial, but when you have to do this for every possible tag with
> some variations, it's a waste of time, that should not be necessary.
> parsing the osm xml files is already a ressource consuming task; every
> unnecessary work should be omitted.
>
>
> _______________________________________________
> talk mailing list
> talk at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
>