User talk:Ulamm/Mappers, evaluators and feedback

From OpenStreetMap Wiki
Jump to: navigation, search

If somebody considers this short article to be wrong, he may move it to User:Ulamm/Mappers, evaluators and feedback.--Ulamm (talk) 12:34, 8 October 2014 (UTC)->One month after its start the initial modesty has expired.--Ulamm (talk) 09:39, 8 November 2014 (UTC)

As far as I can see, this page comes from a single user's opinion, and was not born from a discussion with the OSM community, therefore it would indeed be more appropriate as a user's subpage. --Jgpacker (talk) 13:42, 9 November 2014 (UTC)

"almost every paramater pragmatically has at least three values, plus, minus and not yet recorded"

I am confused by this statement. Most objects have some implied values (no one is seriously proposing adding toll=no to all footways, cycleways and roads without toll), numerous tags have more than two values Mateusz Konieczny (talk) 15:00, 8 October 2014 (UTC)

The necessity depends on the circumstances.
Example: In a village without any sidewalks you need not write footway=no. In an urban residential district, where all streets have sidewalks, you don't mention them, too. If in a region, where it is fifty-fifty, you begin to mention sidewalks, you ought to tag either "yes" or "no". This way, any reader sees (or could see if it were rendered) where he really can expect a sidewalk and where he really has to expect none. And your fellow mappers in the same city know, where there are still "white spots".--Ulamm (talk) 15:34, 8 October 2014 (UTC)
This is not what happens in practice, though. Either things are explicitly mapped as being there, or they are not mapped. The latter might be because they don't exist, or because they haven't mapped yet. Often there is not even a possibility to map the absence of a feature. --Tordanik 16:51, 9 October 2014 (UTC)
That up to now many if not most mappers have not been aware of this aspect of working in an opensource project and therefore haven't recorded if they didn't find what they have researched, is the main impairment of the usability of OSM.--Ulamm (talk) 10:04, 8 November 2014 (UTC)

Confusing page

I think this page is quite confusing. It's like it's trying to cram a dozen different messages from other pages into the same one. Doesn't pages like Tagging for the renderer and Any tags you like already say the same thing? If I understood this correctly, the main point is that mappers should be more active in trying to get mapped features used by some tool. (which I think people already do a lot, but it's not exactly bad to document this, however it should be done with a more self-explanatory title) --Jgpacker (talk) 18:43, 9 October 2014 (UTC)

This article is the late attempt of creating something like a preamble of the wiki, showing the rules behind the rules.
In special:
  • The terms "tagging for the renderer" and "tagging for the router" are often misunderstood as an invitation to neglect feedback.
  • The article "Any tags you like" is not bad, but something like the second step before the first step.--Ulamm (talk) 09:53, 8 November 2014 (UTC)

The exactness of the data must be limited to practical limits

You say that no mapper is allowed to limit the exactness of data to his personal limit of interest, but historically the OSM community does limit the exactness of the data. The limit is when the trouble to add and maintain the details are not worth the benefits. --Jgpacker (talk) 13:35, 9 November 2014 (UTC)

  • The subject of this chapter are details that are not interesting for everybody, but relevant for some purposes. It is not in favour of excessive mapping, see the last but one section of the chapter above, "no mapper should add records without an idea, for what or whom they could be useful".
  • The different depth of mapping in OSM as an opensource project is not only a challenge for mapping discipline ("Note if you didn't find what you have been looking for!" – see first chapter) and for renderers, it is also a matter of freedom. A good example is 3D tagging, for some famous buildings it is nice, but nobody will map all the world in 3D.--Ulamm (talk) 18:30, 9 November 2014 (UTC)

Copy out of:

Thu, 18 Dec 2014 17:14:08 +0100 From: Mateusz Konieczny
This advocates adding [oneway=no; toll=no] to nearly all roads (just because some are with toll and oneway). I consider this as a bad idea.

Ulamm, 19. Dezember 2014 11:58:50
The principle "yes vs. no vs. unrecorded" is no total ban of default values.
But if both, "yes" and "no" have a certain likelyhood, you mustn't use "no" as a default value.
Toll is a good example:
In a region without toll roads or on a type of roads that is always for free, you need not tag toll=no.
In coutries like France and Italy, where most motorways are tollroads, but some are for free,
you ought to tag toll=yes to the pay sections and toll=no to the free sections.
But motorways tend to be the best recorded part of a road system.
Residential streets often are not, nor tracks in the fields.
In old narrow urban districts more than 50% of the streets may be oneway roads – there you'd better tag oneway=no, if a section of a street is bidirectional.
In the outer suburbs and the scattered settlement around, some streets may have sidewalks but some not, some may be paved but some not, some may be lit but some not. The ratios may be 95%/5%, 50%/50%, 5%/95%, or anything in between. Such are classical conditions where you have to note "no" as well as "yes".