While I have started a global discussion and vote on how to tag narrow residential roads, someone favoring a service=alley solution edited the wiki page. They tweaked a paragraph to fit a certain narrative without letting anyone in on it in the discussion or mentioning it on the wiki talk page.
I’m not pointing fingers, and I’m not totally in the loop on the process, but is this really appropriate? The wiki page has been confusing, and it’s causing different opinions, but shouldn’t the community agree on these wiki changes first?
It depends. Sometimes the edit can serve the discussion by clarifying language that was simply poorly worded. (Most of the wiki wasn’t written by professional technical writers!) And eventually someone will need to edit the page to reflect a consensus – identifying and acknowledging a consensus on a subject is one of this community’s biggest challenges. But if someone clearly knows there’s a discussion on the topic and there’s no consensus yet, they should acknowledge the disagreement and/or inform the discussion’s other participants as a courtesy.
For example, I recently made an edit to a page that came up as a point of contention during an RfC for a tagging proposal. I didn’t expect the edit to be controversial across the OSM community in general, but I knew it would come as a surprise to the discussion’s participants, so I left a message for them on the relevant talk page. Even so, it did upset them, which was not my intention.
Acknowledging competing mapping styles in the documentation can avoid hurt feelings, but in the long term, these explanations can outlast the original controversy, confusing mappers and data consumers who can’t tell whether they have to worry about the differences. So ideally someone would come back later to reassess the debate and acknowledge a consensus later on.
The proposal affected by the changes its mine. It is a proposal about traffic signs. If you consider two special ones ( highway=stop and highway=give_way ) we are talking about 2 million items in OSM (at last these two are out of this proposal because there are so much items without this consensus and generates controversy, for now.)
Probably changes are done with good intention but then It could affect the rest of the proposal because if you are basing part of the proposal in a idea with some “same exactly sentences” you can find in the wiki and this wiki has change you have a big problem, specially with other people who reads the wiki and the proposal. If you add also some not general consensus there is a problem. OSM is big (all over the World) and consensus is difficult but when you are thinking you have one…other person in other language can read (or understand) other, and when there is a proposal in the middle it makes difficult to re-find this consensus. For example , in our case changes affects to the summary of the page, but also all the translations (8-10 pages). As a so-used item these wiki pages are read by lots of mappers.
I think the best option is wait reasonably to change the wiki after proposal voting. After all, probably Community should discuss this new consensus, also with the situation of a new proposal approved (if rejected it is work of the Proposer to remake the proposal). But Proposers have to be able to refer to wiki without fear that everything in that bad week you are inside of the process had changed.
Is this “tweaking” a paragraph? I thought from the discussion it was clear that “old” is not a useful distinction for alleys.
You are free to make votes about tagging narrow residential roads, but take into account that the current scheme (highway=service with service=alley) is in use for a long time and for example in Italy there are 25k uses of this tag: service=alley | Tags | OpenStreetMap Taginfo Italy
in Germany there are 47k and in all of Europe 385k: service=alley | Tags | OpenStreetMap Taginfo Europe
UK and Ireland together are about 56k, in many other places this is likely about narrow streets more than about access from the back. By the way, some buildings only have a back
You’ve mitigated this problem using permalinks to specific versions of a page. This is a smart approach. In the future, other mappers will be able to return to your proposal, follow this link, and understand what you were discussing, even if things completely change in the meantime.
I fully agree that the wiki should accurately represent the current usage of a tag. However, why initiate this update in the midst of an ongoing discussion after numerous individuals have already cast their votes?
Many including me have formed their assumptions that this high-level summary paragraph you modified referred to the “Old or medieval ways” section below, which serves as an exception in Europe and deviates from the main purpose of the tag (“An alley is a service road usually located between properties for access to utilities”).
By eliminating the “old” reference without specifying its location/use-case (Europe) or introducing a corresponding section below, you are effectively leading readers to believe that they can apply service=alley globally for any narrow streets, which, based on the discussion and vote outcome, is not acceptable.
…And then , in talk page you would find sentences about your proposal is so long or why do you talk about that if it is not the strictly aim of the proposal and probably you have extra work to change the proposal. Proposer has to be able to be calm, it is so difficult to make a big proposal with a RFC and answer it all to complicate it more with the other users changing the outside context it could be based on.
This is true. I don’t have a good answer for this, other than using permalinks. But please also understand that, at any time, the wiki has hundreds of in-flight proposals. Those of us who maintain the main tagging pages are already navigating a minefield of political tagging battles, even before considering proposals in various stages of completion. At the end of the day, the wiki needs to serve mappers and data consumers with accurate information to the best of our ability. Everyone needs to be flexible and transparent in this environment.
Also, if discussion related to proposal found issues with Wiki documentation then waiting (sometimes years or more) for completion of proposal should not be needed.
Though in such case controversial edits should be even more avoided or be more discussed.
In my case there will be not years. RFC is working (I try to answer every day some of them) and when it passes some days without new comments on all the discussions places about that I will go forward to the vote process. So changes in the wiki can wait some days also.
I’m also currently taking a proposal through a (second) RfC. Technically, it has been in RfC for a couple years, since there’s no process for ending the first one. Fortunately, the pages it affects have been quite stable, yet it also cites actual usage statistics in the database, which are much less stable.
When I first drafted the proposal, the tagging style I wanted to defend was more common on natural=peak than it is today. It’s still common enough on other kinds of features, but this has made it more difficult for me to find examples. There was even an active effort to eliminate this tagging style, technically during the RfC. It was a bit frustrating, but ultimately it was my fault for taking so long. At least it was an opportunity for me to raise awareness of my proposal.
If you haven’t already done so, consider adding a link to your proposal on each of the pages that would be most affected by the proposal. Then, no matter how the status quo should be documented, mappers can see that some changes may be coming soon.
For widely used tags, generally I think it’s a good idea to let a discussion settle, then propose a change, and see if anyone objects, and if they don’t, proceed. At least that’s what I’ve tried to do so far, to get a select few Wiki pages into slightly better shape. It takes a lot of time and effort though.
@dieterdreist, I’ve undone your wiki edit since there was no response to my comments.
Using the “current scheme” in one region doesn’t necessarily mean it’s the right way to go or something we should fully endorse globally. Otherwise, we might end up with every private road getting tagged as highway=service and every dirt roads as highway=track.
It’s interesting to see that the country your referenced here doesn’t even have specific guidelines for service=alley:
Maybe it’s worth looking into that before starting making changes to the global wiki?
I agree the service=alley wiki needs some work, but perhaps we can take a more collaborative approach, like suggesting changes through a wiki revision proposal and RfC, instead of diving into a wiki edit dispute.
A formal tagging proposal, RfC, and vote may be overkill if your intention is to clarify the definition of a “de facto” secondary key within a specific country. The local community can choose a less formal route, such as a straw poll on the forum. There is some precedent for a regional proposal on the wiki (Japan, Portugal, U.S.), but usually only for a complex tagging scheme that requires a detailed document.
No worries at the country level; for instance, Thailand conducts regular informal discussions and votes in the forum to reach a consensus on road usage. We then update our country-specific wiki guidelines accordingly.
I was referring to changes made to the global service=alley page that didn’t include any references to a specific region. I believe any global changes should be made in a more official way.
In my understanding, the point of the proposal process is to agree on how to tag stuff, so after the proposal the Wiki gets updated with prescriptive instructions on how things should be mapped.
The point of a Wiki page on a de facto tag is to describe to people how things are mapped, without being prescriptive.
Changes to the descriptive documentation should be possible without a formal proposal, because that’s not the goal of the proposal process, but ideally they still only be made after the discussion has settled.
For example a proposal for service=alley that tries to define it as “a narrow service road for rear access to buildings; also a narrow passage between buildings such as can be found in medieval European city centres, but only if it is drivable; also a tree-lined path in a park or garden” would never be successful, still I would think it’s useful to have a Wiki page that informs mappers and data consumers that when they see service=alley it might mean any of the above.
if you think putting back a sentence that we already discovered to be factually questionable will improve the situation, then you have done the right thing.
For service=alley I have been looking at the map in Morocco and elsewhere and it seems they also use the tag for narrow streets. In general the situation was “mixed” but I might also have missed something because I did not know what is there on the ground.
I do not understand why you bring highway=track into the discussion?
Generally I believe we should strive to have uniform tagging conventions globally. I know there are some areas where people enjoy doing everything differently to everybody else, in particular our homeland, but overall they are in a special situation (island, self isolated by choice) so cooperation with neighbours is less crucial. We couldn’t afford if mappers in each of the 200+ countries would invent their own conventions.
I didn’t expect my edit to be controversial, or I would not have done it. I am certainly not going to dive into a wiki edit dispute about this.
I get what you’re saying, but how can we make sure a tag doesn’t go off track and cause confusion? It seems like anyone can come up with their own system, throw it in the global wiki, and convince people to roll with it.
This just convince me more that each country should have its own tagging rules because global pages might shift and create a mix-up with conflicting info.
At the very least, if the service=alley tag had a clearer description like below (additions in bold), users could make their own call instead of getting pushed around by controversial global statements:
service=alley
Description: An alley is a service road usually located between properties for access to utilities.
An alley or alleyway is a narrow service road usually located between properties to provide access to back gardens, rear entrances, fire exits, and storage areas. Alleys are normally found in urban areas and often run between the rear sides of buildings such as houses, and commercial premises.
Other region-specific definitions:
in USA, alley streets are alleys by definition, i.e., they serve the rear of buildings, but are named.
In Europe, Old or medieval narrow streets that provide access to the main entrances of buildings may also be mapped as alleys, distinguishing them from wider modern streets mapped, for example, as highway=residential. In some countries, this case is much more common than the rear access case.
In Brazil, the tag service=alley is applied to denote extremely narrow streets that permit bi-directional traffic, featuring a width of less than 3 meters, where two vehicles cannot pass simultaneously.
In Taiwan [and Thailand until now], assigning service=alley to narrow residential roads is discouraged. Local mappers are advised to utilize the relevant width and lanes tags instead.
Always refer to your country’s wiki to ensure alignment with any specific guidelines.
Case in point: what you thought was the definition in the U.S. was just one mapper geeking out about a peculiarity in their own city, possibly overstating the case. In fact, named alleys account for only 11% of service=alley ways in the U.S. and only 14% in Baltimore specifically. To provide a bit more balance, I added a simple gallery with more representative examples from elsewhere in the country, including some unnamed alleys. I would encourage mappers from more countries to add locally relevant examples. At least it would be an opportunity for the community to identify and discuss unusual local practices.
I think some tags naturally vary by country more than others. For example, motorroad=* has a different explanation for each country, because a global definition can only speak in generalities. Highway classification is a prime example of this phenomenon: one only needs to look at the “international equivalence” of highway=trunk to understand that global consensus is sometimes a pipe dream.
