Best Practice for Ohio Fast Food/ Bank drive-throughs?

After perusing the Ohio mapping conventions wiki, I have a few questions.

I understand we’re trying to reduce map clutter by having gas stations retain the small rectangle they were populated with, and only marking “permanent” type landmarks such as parks, major museums, cemeteries, etc (what about large hospital campuses?)

My main question revolves around fast food establishments. Should we or should we not denote the drive through lane, with the “parking lot road” type?

I was an AM for Stark County originally, but lost my privileges after I fell off for 90 days. I’m just getting back into using the editor, and want to follow best practices, as a lot has changed since I was the only person in my area doing edits.

I’ve fallen behind a few weeks in keeping up on the forums, but the last guidance being offered was to generally not map minor parking lots (and drive-thrus) unless/until Waze updates the app to make the display of parking lot roads be less intrusive. This change in display was something promised earlier this year.

The notable exceptions that I recall:

  • when the stop-and-go traffic around a parking lot bleeds over into an adjacent roadway (causing spurious slow/stopped traffic indications). Other than being there in-person when this happens, the only way to find out about this condition is if there is a UR about it.
  • when something (e.g., a turn restriction) would significantly affect navigation into/out-of the parking lot

OK, thanks for the clarification.

I would have initially thought that a small “stub” of a road into a drive through would suffice to denote a turn restriction, but that may cause more confusion when looking at the map when driving. It’s a double edged sword. Someone stopping at McDonalds can cause a road to appear to have a traffic jam when there isn’t one, but at the same time, in order to fix this, you may have to add numerous segments to the map to show access roads to the drive through, directionality can become complex, etc.

With that said, I’ll keep an eye out for a situation where erroneous traffic jams are reported. It appears that most parking lot roads at the major shopping areas around my general area are already mapped. The GPS points show heavy use of them, so I don’t feel that removing them would be prudent. https://www.waze.com/editor/?lon=-81.28&lat=40.71738&zoom=0&layers=BFTFFTTTTTTTFTTTTTTTTTTFT

In the blog section, there is this article: http://www.waze.com/blog/waze-helps-find-best-route-to-the-hospital/

I say it is of value. For region 3 (Note below*), I have adopted what Jason and company have been working on for landmarks as my rule and guide for landmarks in this region. Here is an excerpt:

"Public Service

Hospital/Clinic
Hospitals Mapped at the fence line, including associated parking lot(s) (if attached) in one landmark."

That can be found here: https://www.waze.com/forum/viewtopic.php?f=594&t=44408#p401456

I hope that helps answer your hospital question.

In general, my recommendation is “no.” This stems from a zero-tolerance policy because of the intense clutter that would result in major metro areas. For example, in Columbus-- if you made every city/state/county government building (including libraries, fire departments, etc.), every hospital, every library, every school, every museum, every gallery, etc. into landmarks, it would be a mess-- the names of the landmarks would obscure roads and you wouldn’t be able to tell one landmark from the next. Many county seats would be the same way.

If folks are trying to find relief during an emergency, they shouldn’t use waze–they should dial 911.

If they’re trying to find directions somewhere, they should use the routing engine.

The exceptions to “no hospitals” would be something like, say, maybe the sprawling main campus of a famous clinic or medical center (which is, in effect, a university, if not a separate city). But rather than try to explain to every new editor why mapping their favorite stripmall minute-clinic is bad but showing Walter Reed is OK (or, worse, have them be lead by bad examples on the maps and not even bother to read the wiki for guidelines), it’s easier to say “no” to the vast majority of potential landmarks (and to delete them as soon as they pop up). The editors that stick around can learn the subtleties of when exceptions make sense (and how to avoid obscuring the surrounding streets with names, etc.)

Not everyone uses it for navigation for relief of an emergency. Just because my wife was in labor, doesn’t mean it was an emergency. They told us to go to the mall and walk around until her contractions were closer together. (before Waze experience)

Someone has to get to the hospital for an appointment, such as a children’s hospital or burn center. Shriner’s hospital burn center isn’t an emergency scenario, but one where they go for treatments and therapy for sure.

I can agree it can get cluttered, but I don’t think we are asking to map every doctor’s office in existence here. You draw a double edged sword when you say map this but not that in reference to a medical campus such as KU or Mayo compared to a Children’s Hospital or ER.

Another example of an emergency that isn’t necessarily for a relief where 911 would definitely be used for the patient, the family members and friends wanting to navigate to said place afterwards.

Granted, as we incorporate Google search more integrated into the app, it’ll get better for POI and search results. Until then, hospitals in general are pretty large masses that aren’t small and have bleeding names to obscure roads for the most part.

I don’t get it – you’re saying that a user is like “I need to find this place”, so, instead of using Waze’s navigation capability, they’ll scroll around a city until they find what they’re looking for?

I can search for an airport with the landmark, are hospitals not searchable?

Landmarks should be considered unsearchable. Yes, we have some searchable airport landmarks, however we have not been able to duplicate that consistently. The ones I created work great. The ones Alan created worked great, and the ceased to work at all. No other landmarks have been consistently searchable. The testing has been inconclusive.

So no landmark placement should be based on searchability. That’s what Google’s POI database is for.

iOS Tapatalk 2 sucks.

I recognize there hasn’t been activity on this thread for a month, but as a “baby editor”, I ask the following:

I’ve jumped in with best intentions to respect the community, and with regard to the topic, I’ve followed the “intro to editing” videos. The videos to would suggest that when waze detects users not following the route because they stopped at a drive through, or a trash truck emptied the dumpster out back, that the solution would be to outline the property in an unnamed parking lot to allow freely flowing traffic. This should cease the auto-generated map errors while eliminating the confusion of the additional parking-lot road types. I’ve done about three of these so far. Is this truly the best practice and/or should I revert and delete the one’s I’ve done?

The thought would be on an as-reports-are-generated basis. Nonetheless, still might seem to add clutter to the editor and servers, just not the end users.

Openly welcoming advice from the pro’s.

-Skip

Remove them, they clutter the map & are contrary to the near final US Landmark Guidelines.

Remember, MPs are “something to look at” not necessarily “something to fix”.

What AndyPoms said, and also please understand that the videos were made (1) by marketing people who typically do not edit the map or participate in the forum, and (2) over a year ago, partly using old or bad information.

Thank you both… I’ve read AndyPoms guideline link and undone what I promised. I noticed a few others that were recently done by other bold Level 1’er’s. Needless to say that, where appropriate, I uncluttered two others that were established within the last month.

I’m aware the videos are about a 1% jumpstart, and the other 99% is the editing manual, best practices pages, FAQ’s, the wiki, junction style guide, state, regional, national forums topics, as well as experience and forum feedback, etc, etc…

Very close to topic, but slightly off, I found this one of a small shopping center of maybe only 18 or less businesses. Per AndyPom’s linked Guidelines from the champs, these four separate yet touching lots should be removed in the same sense as the bank/fast food/small business lots?

https://www.waze.com/editor/?zoom=5&lat=39.98316&lon=-83.14756&layers=BFTFFTTTTTTTTTTTTTTTTTFTT

-Skip

IMO those roads are fine. The number of stores really doesn’t matter; the lot is large enough that users might benefit from being directed to a specific entrance or exit. Also, one of the entrances has a turn restriction, so even though that alone doesn’t require mapping it, it does support leaving the roads there.

We are asking the developers to make parking lot roads less visible in the client, so eventually you’ll probably see fewer objections to including them.

Thanks, and not that my opinion means anything, but I thought the roads were fine too. It was the blue shaded area “parking lot”/landmarks that I thought should be removed, and the fact that the editor placed four individual 4-sided areas instead of one irregular polygon over the entire lot (generating additional map clutter and server load). As a rookie, and as it would pertain to drive-thru’s: if large enough, such as this, leave the parking lot roads, but remove the parking lot landmark area(s)…?

I’m not the editor on this example, but likely done by another well-intended rookie who may not be doing equal due-diligence on learning the ropes. It was done earlier this month.

Again for reference: https://www.waze.com/editor/?zoom=5&lat=39.98316&lon=-83.14756&layers=BFTFFTTTTTTTTTTTTTTTTTFTT

-Skip

Oh, ok, sorry. I must have had the landmark layers off. No, parking lot landmarks are definitely unnecessary. If you find any that are named, leave them alone for now–someone may be testing whether or not they’re searchable–but unnamed parking lots perform no known function currently.