Get a sneak peek at whats next for Permanent Hazards on our April 7th Office Hours!
Post by DwarfLord
My efforts linked above to create separate guidance regarding the hazards of adding junction nodes -- to which this proposed addition to the RR guidance could easily refer -- are not working out at this time.

I appreciate there are aspects of my proposed change that could be considered vague and/or could use expansion. On the other hand the language proposed a few posts up (in the quote) is compact and I believe reasonably meaningful to most editors.

I'd like to go ahead with the proposed change to the RR section for now, because I think it's important that we have this caution in our guidance. As far as I know, new editors reading the current guidance are still under the impression that we want every grade crossing junctioned without exception, and as many of us here know, that is counterproductive.

Once the guidance is in place, we can continue to look for ways to improve it as always.
DwarfLord
Wiki Master
Wiki Master
Posts: 2512
Has thanked: 1065 times
Been thanked: 1451 times
Send a message


Post by DwarfLord
Typically I would wait some time. But, given the immediate positive response from Kartographer, and given that even the two editors who expressed any caution about the proposal were only suggesting additional clarification and not faulting the proposed content as far as it went, I've gone ahead and made it live.

https://wazeopedia.waze.com/wiki/USA/Ro ... s#Railroad
DwarfLord
Wiki Master
Wiki Master
Posts: 2512
Has thanked: 1065 times
Been thanked: 1451 times
Send a message

Post by DwarfLord
Hey, while reading the entire Railroad section I came across the following as the "third purpose served by the railroad road type" that I missed during our initial conversation about junctioning grade crossings:
Existing railroad section wrote:Finally, when a railroad crosses a drivable road segment at grade (same elevation) the routing server can better determine delays at that crossing.
This statement is not really true, is it? My understanding is that the benefit to which this sentence alludes has nothing to do with the Railroad road type. It is a byproduct of adding a junction node on a drivable road at or near the grade crossing (under the specific circumstances we've been discussing). Such a junction node could be with nearly any road type, a PLR, a Major Highway, maybe even a Runway/Taxiway; doesn't have to be Railroad. So I believe the quoted sentence is misleading.

I'd like to recommend either
  • OPTION A: we change the preceding language in this paragraph to "two purposes" instead of "three purposes" and remove the problematic sentence; or
  • OPTION B: keep the "three purposes" but change the language to read: "Finally, mapping railroads at grade crossings provides a convenient opportunity to add a junction node to the drivable road that, under specific circumstances[rr], will allow Waze to gather higher-resolution timing data associated with reaching nearby destinations. Do not, however, add such junction nodes under other circumstances."
I lean towards Option A, since Option B is largely redundant with the stuff we just finished working on. But if folks prefer Option B that's OK by me.
DwarfLord
Wiki Master
Wiki Master
Posts: 2512
Has thanked: 1065 times
Been thanked: 1451 times
Send a message

Post by DwarfLord
tonestertm wrote:..."Under certain circumstances, mapping Railroads can be beneficial to data gathering"...
The Railroad road type section actually already states (as benefit #2) the utility of this type in absorbing railroad-associated speed data and preventing them from contaminating drivable-road speed data. That's a great benefit and I support leaving it in guidance! But, it also means the third "benefit", if we do end up keeping something in that place, has to be more specific than saying the road type is beneficial to data gathering, as that's basically already been said.
tonestertm wrote:I think the "offending" sentence is poorly worded, but indicates a sometimes-appropriate reason for making sure tracks (or at least a data gathering node accompanying them: see below) are mapped.
Right, the current sentence for the third "benefit" of the Railroad road type gives the strong impression that junctioning RR grade crossings benefits routing not merely because of the new junction node, but because of the junction specifically with the Railroad road type.

Well, I believe that impression would be completely incorrect; it's 100% the junction node that creates the benefit. Any standard junction node, with any road type, or with no road type at all, will do. But editors who don't understand this might disregard any other junction nodes on a drivable road near a grade crossing, even those just a few feet away, if those junction nodes are not explicitly connecting a railroad.

Those editors could then go on to add another junction node specifically with the Railroad road type in order to create the correct "benefit". A situation we've already seen trainloads of ( :mrgreen: ).

My hope here is to remove or adjust text that could lead to this well-meaning but counterproductive technique. I fear that anything we try to say in an article about the Railroad road type that says this road type is beneficial when junctioning grade crossings is going to point readers in the wrong direction. Which is why I support Option A -- simply remove the misleading sentence.

I'm happy to discuss running the content of the footnote into the bullet point, but I think that's a different conversation and would love to finish this one first. (I only kept it as a footnote because that's what it was before and I was hoping not to rock the boat any more than necessary.)
tonestertm wrote:This may be heretical, but I'd even go so far as to wonder if the junction node should just be added to the driveable road AT the RR, without actually going though all the histrionics involved with actually junctioning the RR segment.
This would work splendidly, but I will courteously take a pass at advocating that it be formalized in guidance :mrgreen:
DwarfLord
Wiki Master
Wiki Master
Posts: 2512
Has thanked: 1065 times
Been thanked: 1451 times
Send a message

Post by DwarfLord
Based on the responses so far, there seems to be general agreement or at least acceptance of the proposed edit to the first paragraph of the Railroad Road Type section. I'll do this tomorrow (Sunday) unless anyone squawks. The edited section will read as follows:
First paragraph of Railroad Road Type section wrote:The railroad road type serves two three purposes in Waze. First, it provides drivers with visual orientation relative to railroad and light-rail tracks. More importantly, in the common case where passenger-carrying tracks lie parallel with roads, mapping the tracks allows Waze to recognize spurious speed data from people Wazing on the train and prevent it from corrupting speed data for the adjacent road. Finally, when a railroad crosses a drivable road segment at grade (same elevation) the routing server can better determine delays at that crossing.
DwarfLord
Wiki Master
Wiki Master
Posts: 2512
Has thanked: 1065 times
Been thanked: 1451 times
Send a message

Post by DwarfLord
Done.

https://wazeopedia.waze.com/wiki/USA/Ro ... s#Railroad

I took the liberty of adding a minor wordsmith to avoid associating a verb with the word "data". Technically the word "data" is plural -- "datum" is the singular -- but this is little known; I've even seen Wazeopedia entries "corrected" from "data are" to "data is". Rather than fight this effect I just reworded to avoid the situation altogether :ugeek:

This concludes my efforts with this section. Thanks to everyone who reviewed and contributed!
DwarfLord
Wiki Master
Wiki Master
Posts: 2512
Has thanked: 1065 times
Been thanked: 1451 times
Send a message

Post by hawkeygoal
While part of me is reticent to add a hard definition: It may preempt future discussion of how much room is implied with the phrase "room for cars to back up" if it were defined specifically or allusively. That is, replacing the verbiage with a specific distance (for example, "50 feet") or adding a parenthetically suggested minimum distance. I know it's silly, but devil is often in the deets. Otherwise, rock on.
hawkeygoal
State Manager
State Manager
Posts: 252
Has thanked: 91 times
Been thanked: 95 times
Send a message

Post by kentsmith9
DL, I think we might also want to consider adding warnings about adding the junction if there is a drivable intersection nearby that is subject to traffic backups in turn pockets which you already noted in the True Elevation thread. We can likely use similar language. If you already have that in your proposal, I missed it.
kentsmith9
Waze Global Champs
Waze Global Champs
Posts: 5766
Has thanked: 816 times
Been thanked: 1156 times
Send a message