How to map Key:smoothness

General talk Tagging general discussion
182

Let me just get my time machine out of the shed … :smile:

5 Likes
183

Perhaps an editor view that would ask what you were driving/riding before adding the value would be better in that case?

Basically first question.

Were you driving a car/4WD/skateboard/bicyle etc ?

And then you give your opinion of the surface based on your experience as good/okay/bad.

The editor then assigns the correct value depending on your ride.
So for examples:

if skateboard and good then value could be excellent
if bicycle and good then value could be good
if car and good then value could be intermediate

this might help to break the bad influence from the meaning of the words?

184

With gaps this large, if they were also deeper, I would probably have voted for smoothness=bad as it would be inadequate (mostly unusable) for city bikes. If the gaps were deep but narrow, then it would still be smoothness=intermediate because the top of the blocks is quite flat and evenly aligned. If the gaps were narrow or shallow but the blocks where not flat or were misaligned, I would tend to vote for smoothness=bad for the same reason.

185

For a moment, I was unsure if this would fall into the smoothness=very_bad category, as a car with high clearance might be necessary, not because of the smoothness/unevenness of the surface along the tire tracks, but because the central part is slightly elevated with some slightly taller rocks on it. This is an aspect that perhaps needs clarification on the wiki.

186

To me it looks like the best three values (excellent, good and intermediate) are bike-centric (they make very little difference for car routing choices) and the rest is car-centric (they make very little difference for bicycle routing choices).

2 Likes
187

The Highway_Tag_Africa wiki already documents something close to this using conditional restrictions:

seasonal=yes
smoothness=horrible
smoothness:conditional=very_horrible @ Dec-Apr

This works well for capturing the worst case, but leaves open what the base smoothness value should represent. On a surface like laterite the two states are not minor variations: a track that is bad or very_bad in the dry season can become horrible or very_horrible after rain, and in the wet season that state recurs frequently even if heat restores it between events.

“Most common state” is not a useful guide wherever strong seasonal variation exists: wet season in the tropics can cover half the year or more, but spring thaw, autumn mud, and snow or freeze-thaw cycles in temperate and northern climates raise the same question.

Two questions for the community:

1. For seasonally variable unpaved surfaces, the base smoothness value should reflect…
  • Dry-season / best-case conditions
  • Wet-season / worst-case conditions
  • The conditions at time of survey (no convention needed)
  • The most common / typical conditions
  • None of the above (comment below)
0 voters

Note: “best-case” and “worst-case” here mean the best and worst regularly recurring states, not rare extreme events (e.g. a once-a-decade flood).

2. Should smoothness:conditional be documented on the Key:smoothness wiki page?
  • Yes, with guidance on what the base value represents
  • No, coverage in Highway_Tag_Africa is sufficient
  • None of the above (comment below)
0 voters
1 Like
188

Always put an extra answer in the poll “none of the above”, otherwise you’re creating false dichotomy and skewing the results. Anyway I answered the “closest” answer, but I don’t really stand behind them :slight_smile:

For me:

  1. I’d use raw smoothness for most common situation, i.e. one that the random driver is most likely to encounter. Only when the “good” and “bad” situations are about equally common, would I choose worst-case condition (i.e. it is better that the driver is pleasantly surprised that the conditions are better then expected, then for them to arrive there and find unexpected unsurmountable barrier)

  2. I’d say “no need at a moment, really” for documenting as Conditional restrictions; but the reason is that those apply widely and it does not make sense to mention them for all wiki pages for all tags, and with only tiny 22 uses worldwide it is basically just noise. I’d expect at least 2-3 orders of magnitude more usages before it makes sense.

Also related: Surface: Worst Case oder Average Case?

2 Likes
189

Added a couple more options to both polls :grinning_face:. Turns out author can add options after posting if “Allow options to be added or removed after posting” is enabled when creating the poll.

190

I would put worse of typical states - not overall worst possible

if road is impassable for few days in decade due too flood I would ignore it

but if state exists each year for weeks…

2 Likes
191

Indeed smoothness assumes that it is constant (or at least as stable as many other things we map (surface, opening_hours, etc.) and the same for all wheeled vehicles. This is often a false assumption and a major weakness of the tag, and smoothness:conditional could be a way to work around it (as is surface=clay, surface=laterite and surface=?).

Although Highway_Tag_Africa mentions an example where the best case is tagged with smoothness and the worst case with smoothness:conditional, it is probably not too late to reverse that and edit the few uses of smoothness:conditional accordingly. As stated by @Matija_Nalis: better be pleasantly surprised than to get stuck or cause damage, injury or worse.

We have to be careful not to expand the smoothness wiki much more because some mappers are already daunted by the amount of information there and decide they are not capable of mapping it (@supsup). But I think we can afford to mention it in a short “Seasonal variations of smoothness” paragraph and then link to a dedicated wiki page for smoothness:conditional.

1 Like