[Tagging] New sections added to "Good Practice" page?

Fernando Trebien fernando.trebien at gmail.com
Mon Jul 1 17:28:43 UTC 2019


On Sat, Jun 29, 2019 at 7:58 PM Joseph Eisenberg
<joseph.eisenberg at gmail.com> wrote:
> 2) "Don't map insignificant, perishable and mobile object"
>
> This is certainly true, but it duplicates advice in the previous
> section, so I've removed it.

I think the existing advice was better indeed, so +1 for removing this.

> 1) "Check the history of important objects"
>
> "Before making significant changes to important objects (in particular
> settlements, administrative boundaries, major buildings, tourist
> attractions,long route relations etc), check their history. Who did
> this and why? Was this an experienced contributor or a newbie?
> Previous editors may have valuable insights to offer on why things are
> currently tagged the way they are."
>
> I don't find this information helpful for new mappers.
>
> What is the "history" in this context? I don't think it would be clear
> to new mappers - it seems to suggest changeset comments
>
> What is an "important object?"

I think I see what concerns that mapper as important. I've seen new
mappers deleting place nodes by accident and reinserting them,
creating objects that are not linked to the edit history of the former
objects. Of course, I've also seen this with less salient features
too.

While OSM does not have the concept of notability, some kinds of
mistakes prompt the community more readily than others. For example,
accidentally removing part of a motorway is likely to be discovered
and fixed very fast (also making some people a little annoyed), while
accidentally removing part of a residential road may take years to be
noticed and fixed, depending on the level of OSM contributions (very
high in Germany, low where I am in Brazil). Errors on administrative
boundaries often show up while converting the data for offline usage,
which some apps do daily, but with JOSM and iD today this is not easy
to do unintentionally. Deleting a tourist attraction also prompts a
quick reaction, as the attraction is searched way more often than,
say, the median household building.

I've often been told to explain certain kinds of changes in changeset
comments. The only way mappers can become aware of such comments is by
reading the history. Thus, reading the history is good practice, in
fact for all objects, but it adds a lot of extra work, mostly because
our current tools cannot display the history of every small object
(changes to geometry, tags, and changeset comments) in a way that is
easy/quick to read.

I've often recommended using note=* and source:*=* tags for the more
important justifications as they are more visible to mappers, but a
careless mapper may make changes ignoring those as well, as they are
not very visible in all editors. The wiki prescriptively discourages
the use of source:*=* since 2016, but the OSM Tag History service
tells me that usage of source:*=* tags continues to increase
worldwide, so I'm note sure it can really be described as a
"historical" practice.

-- 
Fernando Trebien



More information about the Tagging mailing list