[OSM-dev] new fixme-like check for keepright.ipax.at and alike

Eddy Petrișor eddy.petrisor at gmail.com
Fri Jan 16 16:38:34 GMT 2009


2009/1/16 Jochen Topf <jochen at remote.org>:
> On Fri, Jan 16, 2009 at 03:13:37PM +0100, Simone Cortesi wrote:
>> On Fri, Jan 16, 2009 at 8:30 AM, Jochen Topf <jochen at remote.org> wrote:
>>
>> > Please, please, do not use this convention. If there is no name or no
>> > ref thats enough to know that something might be missing. If one uses
>> > 'fixme' and the next one uses 'tbd' and the next 'missig' and the next
>> > one uses 'unknown' as "special values", the map will be full of strange
>> > names and nobody will be able to find anything!
>>
>> Still I do not understand why one has to use name=FIXME in the first
>> place. Just not adding a key/vaule is not enough?

The only thing I can think of (note I am not adding such tag values) is that
it is really easy later to query a database for ways which should have a name,
but need a survey to establish that, since not all ways need to have a name.
So, it seems, no, it isn't enough to have a key missing, since that would blend
into the background of ways which don't need a name or tag.

As I said, I *don't* use such values for tags, but I can only assume what people
think. I am trying to highlight such wrong tagging practices in automated
checkers and validators in a constructive way by locating them, this way
helping to fix the problem.

> That is just what I was saying. Don't use anything like this. If you
> want to add a note use the "fixme" or "note" key. Something like
>    fixme=please add name
> or
>    note=missing name
> would be better. But thats really only needed for more complex cases like:
>    name=FoObaR
>    fixme=Check spelling of name

I agree this would be better.

-- 
Regards,
EddyP
=============================================
"Imagination is more important than knowledge" A.Einstein




More information about the dev mailing list