OK, so if I’m understanding, the justification for the TIO is not because of any particular desire for the additional instruction to rejoin the freeway, but rather because it will postpone issuance of the long-haul “continue” instruction until after one is one the freeway.
In other words, the principle is not to add a “nice confirmation” at all, but rather to prevent a long-haul instruction from occurring at what could be a confusing, or at least awkward, moment.
Seen that way I could be OK with this. It is consistent with the (evolving?) principle that the Continue TIO is primarily for improving instructions associated with transitioning to long-haul roads. But I’d sure want the guidance to be clear about the underlying principle not being a “nice confirmation”. It would look like that to editors who might take it as a role model.
I always expect at least one instruction at the foot of an exit ramp or at some point along the way unless it’s the type that simply deposits you onto a road without any choice in the matter. And I have always edited that way. The only other option in this case as far as I am concerned is “stay to the right”.
At first I thought “continue” for sure, because the right hand of the initial ramp fork is straight as an arrow. But then three thoughts occurred:
The “continue” instruction at intersections is everywhere and always a forced override (i.e., neglecting the automatic continue for x miles instruction). Waze by default never says to continue at an intersection. Therefore it is not something Wazers are going to get used to except when tailored for specific locations, under (hopefully) unique circumstances. It’s a “special” instruction.
For those same reasons, we don’t want to edit as if “continue” is in the same class as “keep left” or “keep right” – that is, we don’t want to do anything to support the habit of thought (either in ourselves or in drivers) that “If I’m supposed to go straight, Waze will tell me”. If we go that way, basically fighting Waze’s default, we’re creating a world of hurt for ourselves.
If there was no fork, and the driver had no choice but to continue straight onto the surface road, we would not add a “continue” instruction, because Waze is not a chatty app and we don’t typically add instructions only for the purpose of “nice confirmations”. So the only reason for it would be to distinguish it from the other choice of the fork. In that case, the opposite of “keep left” is “keep right”.
I think it is somewhat fallacious to believe that we should decide how to use TIOs based on how things have worked in the past and how things work in places we haven’t got to yet. You could say the same about any of the little fictions we use on the map, like micro-doglegs or median U-turn prevention… A user might get “stay to the right” for some AGCs in old-map pre-mDL areas, but we don’t leave the rest alone for the futility of it, oh no.
Junctions aren’t always going to be parallel. In this example I would almost prefer a “turn left” for the other option. The opposite of a “turn left” is a “turn right”? Not doing that.
We should do what we can to make it the most understandable in each particular situation to the best of our abilities.
I think you misunderstand me. I’m all for supporting drivers. For me, consistency is really important. If Waze talks one way here and another way there, even when the circumstances are substantially the same, that’s confusing. Confusing is bad.
For better or worse Waze has set the default. Then, rather than re-evaluate their default behavior, they threw a one-off override capability in our general direction and then left town to have beer.
I worry we need to be very careful about using that one-off override capability to redesign Waze one intersection at a time. We may not like Waze’s default behavior, but inconsistent behavior is also dislikable.
I get that, but I really don’t think either option is particularly confusing, I just think “continue straight” is a little bit better. I don’t think anyone is going to lose their marbles over this.
If one option is nearly 90° and the other is exactly 0°, and given that it is desirable to have an instruction one way or the other, to me, “continue straight” makes a lot of sense.
The important difference here is that my examples are ramps that end with cross streets and a continuation. This is more of a split were I don’t think anyone would say there shouldn’t be some instruction for both directions.
I’m OK with a Continue here since it better represents the driver’s path. It’s similar to how I’m handling westbound instructions at this junction.
I’ll confess that, if I were only using Waze for one thing in the world that and that was to get from the top of the ramp to the bottom of the ramp, I’d be fine with a Continue.
My concerns are at a more global level. Waze seems to be designing tools for us that are making it increasingly difficult for us to verify and maintain correctness and consistency. I’d prefer to use TIOs (not to mention secret Place stop points – ugh!!!) sparingly because the more we have the harder it gets.
In my darker moments I fear I’m the one who’s fighting Waze! It seems pretty clear they’re not prioritizing for verification, maintenance, or consistency.
Anyway, if in principle we want to limit the extent of TIO usage, I would not say an override here is warranted. If in principle we don’t care that much about limiting the use of overrides, given that Waze has provided the tool and why not optimize every intersection we edit using all the tools in our toolbox, OK.
As long as the eventual guidance makes it clear which principle we’re following – we can’t adhere to both.
I would like to say that Waze would much rather prefer better instructions for the 20M users, rather than the possibility of sub-optimal instructions, or less ease-of-use for the few editors which can also self-service and script away the concern*.
If there’s a lack of editors with sufficient rank to maintain these, then it would be up to the community to downlock or promote enough editors.
*There isn’t any network calls right now when clicking a segment or node, so it would not be unreasonable for JAI or TB to show more information about a TIO.
Hey, I want the best instructions possible. If the map becomes more difficult to verify and maintain, that will eventually lead to poorer instructions unless the volunteer community generates increasing effort to compensate.
When TIOs get to that point, then I think we should raise and push this back to Waze. It’s hard to justify further development of a feature, when there is no usage of the feature.
Good point. I think that primarily speaks to the lack of close cooperation and mutual understanding between Waze and the volunteer community. But that’s a different topic.
What I’m getting here is that the prevailing attitude towards the Continue TIO consists of two categories:
Locations where an instruction would issue anyway. In this case editors are free to change the default instruction in any way they judge would improve on the default.
Locations where an instruction would not have issued without a TIO. In this case editors should not add a continue instruction unless the road name and number change, or it is necessary to insert a new instruction to postpone issuance of Waze’s long-distance “continue” instruction.
Given how the conversation is going, I’d like to suggest we abandon the growing list of situations where the Continue TIO is suitable, and instead simply list situations where it isn’t. Maybe that list would be a lot shorter?
Apologies for the double-post but I couldn’t help myself. I reviewed all the comments in this thread about when NOT to use the Continue TIO, and it is a short list. I recommend we go in this direction for maximum editor flexibility.
The wiki article should have a section on naming for Continue, but that is a separate issue from when it is to be used, so I’m not discussing that here.
WHEN TO USE A CONTINUE TIO
When continuing straight is not an obvious choice due to road geometry, naming, lane or traffic patterns that may suggest best continuation is in some other direction.
To override an existing keep, turn, or exit instruction to better match actual roadway geometry.
To delay a long-haul “continue” instruction until after one has rejoined a freeway (typically only used at rest stops)
If (1), (2), and (3) do not apply, to alert the driver to changes in road name or number when Waze would otherwise remain silent, EXCEPT:
Do not place when there is any continuity of the source highway number or its formal or informal name (e.g. do not use when a shared alignment joins or departs the highway)
Do not place on transitions to local roads, collectors, or short-haul arterials, regardless of name or number changes
Do not place where business routes, beltways, or other major spurs rejoin the associated through highway
Do not place when preceding instructions already convey sufficient information
Do not place on minor toll booths
How does that sound?
I didn’t think the inclusion list was an issue and I think it leads to better understanding of this somewhat complex combination of issues.
This particular exception list might be hard for some people to visualize and in my mind that will lead to compliance issues. Maybe we can expand their descriptions or give examples to better illustrate. Also there seems to be a logic or grammar issue in the #4 sentence, but I am going to ignore that for the moment and focus on the exclusion list.
(I changed to letters for easier following in my notes.)
I agree with what kentsmith9 has said. (I was writing my own response).
I’m doing a bit of digging on the I-605 scenario, but I’m failing to find a significant number of those scenarios (i.e., short spur routes that terminates in an unidirectional interchange). The majority have all directions covered. I agree that it should be Continue-d, and as always, local guidance can be provided for this specific interchange (as I’m having trouble finding other ones!)
I will however try to re-write 1) and 2) to better fit the suggestions.
Generally re-wrote it, should be much more condense, with examples:
I will remind everyone a Wazeopedia article is not intended to be the definitive and exhaustive rulebook, local area consensus can be used for exceptional or unintended situations.
I continue to think a short, basic statement of principles and exceptions, combined with a list of positive and negative examples, would make the clearest guidance. I agree the list of exceptions to my Item 4 needs adjustment. Partly I was just making a point that those were the only “when NOT to use” examples I could find in this whole thread, and it was quite a short list!
However, if consensus is forming around a list of positive usage examples, OK.
If we don’t want to explicitly forbid any particular usage of the Continue TIO, I think we need at least a list of negative usage examples. Otherwise I fear editors will not understand that we want some limits on the use of the Continue TIO (we do, don’t we?), what those limits are, and why anybody would be concerned about misuse or overuse.