[OSM-talk] [Talk-GB] Coordinates in OSM. Really annoying

Lester Caine lester at lsces.co.uk
Sat Apr 22 17:19:28 UTC 2017


Looks like I fell fowl of another annoying 'standard' ... just how
replies to an email list are handled! If only people would use the same
standard everywhere ;)

On 22/04/17 16:23, Dave F wrote:
> In this context I'm unsure what you mean by "interface standards"

Working from the bottom ...
nik2img is a python application, so expects the list of parameters to be
listed with spaces ...
https://code.google.com/archive/p/mapnik-utils/wikis/Nik2Img.wiki it IS
actually using the box parameter, but in python scripts this is --bbox
mapnik supports several interfaces, and url style links prefer not to
use the space character, preferring the comma to break up strings.
Overpass actually comes with a number of interface standards and the
comma string can be used in OverpassQL (but the wrong pigging way around
;) ), but the XML standard requires each element has it's own name.
Osmosis is a java application and that brings another layer of
'flexibility' which requires every element of --bounding-box to be named
... and as with python each element is flagged by a space ...
essentially similar to the XML rules but with different element titles.

Each programming language has it's own well defined standards and there
is little chance we will be able to change those, so we end up with
wrappers which pass a coordinate set in the right format.
http://osm.duschmarke.de makes a number of those formats available in
parallel for those times when you may need to swap between languages ...

> On 22/04/2017 15:45, Lester Caine wrote:
>> On 22/04/17 15:28, Dave F wrote:
>>> As they're parameters, the format can be standardised for the end user &
>>> any programme should be able to sort out syntax behind the scenes.
>> Totally agree except passing box="w,s,e,n" is STILL a matter of the
>> passing mechanism the target application is built around. There would be
>> no problem with the box selector simply providing the right input via a
>> link to the target application, and a graphic interface providing a
>> suitable 'box' into which to put the coordinate string into for GUI
>> interfaces, but as I said ... the problem is the vast number of
>> interface standards currently in use and not a simple one to solve for
>> one piece of data. 

-- 
Lester Caine - G8HFL
-----------------------------
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk



More information about the talk mailing list