[OSM-talk] OpenRailwayMap Electrification Status vs tag electrification=no

[stevea]

On Feb 23, 2022, at 6:31 AM, john whelan <jwhelan0112 at gmail.com> wrote:
> My feeling is locally in Canada is every rail line that is electrified, and they are rare in Canada, will be tagged.  Yes there is some limited value to the tag but I don't think it is very high and there is a cost in the size of the database.

We (individuals in OSM) walk a very slippery slope (tread into areas where a simple, careless slip can cause us to fall down or create damage) when we "decide" that some tag has "some limited value" or whether that value is "very high" or not.  Especially an existing tag.  While it is true that we should avoid or actively eschew truly "bad tags" (especially when they don't already exist or are newly proposed), saying that an EXISTING tag "doesn't have very high (value)" is fraught with peril.  Let's be careful when we both do this and recommend that others do this.

I believe it's OK to passively ignore tags which don't affect you or those tags you don't care much about.  But to declare that a tag has "limited value" is tantamount to harming our map data, even if in so doing you simply state your opinion.

Yes, a tag like "oneway=no" seems to have limited value, but in certain contexts (difficult to describe here, though Martin offers us a good start), it actually ADDS value to the map (used sparingly and/or for emphasis, whether to humans or downstream map users like routers or renderers).  There are many other such cases for many other such tags.

The "size of the database" should be of little or no concern to us, except that we should be aware that truly superfluous data doesn't belong in our map (database).  I agree that it isn't always easy to discern when this is true vs. false.  However, let's not wave away as meaningless tags which without serious analysis (often worthy, yet seldom performed), we cannot say anything truly meaningful about them.