After two years of using the tag Key:trailblazed - OpenStreetMap Wiki, I noticed today that I probably use it wrong. I tend to put trailblazed=yes for colour markings, because the first picture of a painted arrow is associated with yes. I assumed symbols is reserved for very identifiable symbols (comprising of at least two colours, say, or something more recognizable than a splash of colour).
But now my understanding is that yes is a legacy value used before the proposal to add more values to trailblazed and its high useage is a historical issue. Any thoughts? If not I would try to clarify the wiki (and reorder examples, possibly trying to get a picture of an axe mark or something like that for yes instead of the arrow).
I wouldn’t say you used it wrong, just lacking detail. yes is correct, but symbols would be correcter. I wouldn’t be surprised if many of the 39% trailblazed=yes should be =symbols, and the number keeps rising.
Reordering the wiki table would probably solve the issue.
Hmm. I think I’ve been interpreting it the way you had as well, with trailblazed=yes for “there are lines of colors painted on trees” (not necessarily rectangular)
and trailblazed=symbols for something that’s clearly a more distinct shape.
Honestly I don’t think it’s that unusual for some trails to use a mix, with painted lines on trees acting as reassurance markers, with more significant turns or tricker places using a more prominent actual symbol.
I wouldn’t mind there being some more-specific term than symbol for “trees are regularly marked with colors”, but I don’t know as there’s much harm in leaving them as trailblazed=yes (or even trailblazed=symbol) either. Are there hiking maps/apps that distinguish between different trailblazed= types in some useful fashion? I suspect that showing a user of paths the value of colour= and symbol= would be more helpful than more details of the trailblazing= value (beyond it being something other than no or not), though my hiking is much more casual than some other people’s are so others might find different things useful.
I think that the left picture for trailblazed=symbol is an edge case if it is an symbol at all.
The third and fifth one in the first picture row of Tag:information=trail_blaze - OpenStreetMap Wiki would be better examples in my opinion.
It looks to me like trailblazed=* follows a common tag value pattern where one can simply tag trailblazed=yes meaning “the trail is blazed in some way” or one can provide a more specific value (poles, cairns, symbols) indicating what type of trailblazing there is. A data consumer that doesn’t need the specifics but does need to know if a trail is blazed or not will interpret all values other than no as meaning yes. This is the same pattern used by Key:crossing:markings - OpenStreetMap Wiki.
So trailblazed=yes is not wrong for a simple paint blaze. The yes value just has a broader meaning and doesn’t specifically indicate only a paint blaze. It looks to me like of the more specific values, trailblazed=symbols would be the correct one to cover a simple paint blaze if more specificity than yes is desired. It is a symbol despite it’s simplicity.
I’d only use “trailblazed=symbols” if the paint has a distinct pattern to it, such as “yellow rectangle” or “red stripe over blue stripe”. Simple blotches of color would be “trailblazed=yes” (or a theoretical “trailblazed=paint”).
I actually did a big change in the end. I removed this sentence:
Tag trailblazed=no indicates that blazing is expected, but doesn’t exist. This could e.g. be the case in snowy areas where the trail is visible on the ground only in the summer.
This got added post-aproval in 2021 and does not make sense at all and does not match what is in the values in the proposal (aproved definition was “No trail marking at all.”)
Then I removed this:
trailblazed=poles or trailblazed=cairns can be used as combination with waymarked route tagged by a osmc:symbol=*symbol=*wiki:symbol=*colour=*.
as those symbols are supposed to be used on relations, not ways as trailblazed is. I fixed this for the whole article, noting it is discouraged for ways. I also fixed the language and simplified and removed duplicities.
In the end, I added this sentence:
For more than ten years, the tag had only a “yes” or “no” value and the other values were only approved in 2021. For that reason, trailblazed=yes can mean all the other positive values too. It is not clear how distinctive the markings must be in order to classify as trailblazed=symbols. Do splashes of paint or arrows clasify? In the table below, a sprayed arrow is used for trailblazed=yes and a splash of paint for trailblazed=symbols, which does not seem to make much sense.
Thanks for working on the wiki. Looks like an overall improvement, but some of the changes I don’t think are quite right.
Although it’s true that yes/no came first and then other values were added, this is not really the reason that yes represents all the other positive values. They could have all been introduced at the same time and still work the same way. It’s just a common OSM tag value pattern where yes is a non-specific umbrella value and all other positive values subdivide it. Have a look at how building=yes, bridge=yes, and crossing:markings=yes are described:
“Use this value where it is not possible to determine a more specific value.”
“Non-specific bridge tag”
“An unspecified or unknown crossing marking”
trailblazed=* follows the same pattern as these other tags so its description of yes in the values table should say something similar, like “the trail is trailblazed in an unspecified way”. To avoid any confusion, it’s probably also best to remove the photo examples from the yes value, just as is done for the other yes tags mentioned above.
The “Definition of marking” wording “Any trail marking which will not fit any definition below” that has existed since 2021 also misleadingly suggests that yes is only used if one of the other values doesn’t fit. This isn’t how yes values end up getting used in practice though. This should be changed to something like “An unspecified type of trail marking”.
A benefit of this tag value pattern is that it is allows for extensibility without being too disruptive. Data consumers are expected to treat any value other than “no” as “yes” so that as new values are added (such as the potential trailblazed=paint for example) they will automatically fall back to the default. Otherwise they’d be entirely unsupported until data consumers noticed the new value and added explicit support for it. This is hypothetical for this specific tag as I’m not sure if any data consumers currently do anything with trailblazed. I’m just trying to explain the intent of the pattern in general.
I understand the need for continuity, yet I am hesitant to change an approved definition. Maybe some compromise as in “Any trail marking which will not fit any definition above. Also stands as an umbrella term for mappers who do not want to specify the type of markings used.” or something like that?
An approved proposal just means some people agreed it was a good idea to update the wiki in a certain way at that point in time. In this case it was 22 people. Approval, especially by a relatively low number like this, does not mean the proposal was 100% free of mistakes.
Yes that generally captures the meaning. It’s not just about mappers not wanting to specify though. There a number of reasons why a mapper may choose yes instead of a more specific value. For example, the mapper might know the trail alternates between cairns and symbols, but not know the exact positions where the trailblazing type changes. So they are not able to split the trail way and tag each section with trailblazed=symbols or trailblazed=cairns. So for now they tag trailblazed=yes on the whole thing.
This would be good wording to describe trailblazed=yes on the the wiki:
“An unspecified type of trail marking. Used when mappers are unable or unwilling to choose a more specific value.”
If I’m looking at the right version of the page, this was in the Usage section, while the Values table still said “no trail marking at all”. I understood it as a genuine attempt to explain how a “no” value is typically used in practice, and why there may not be all that many of them even if the tag becomes popular. It’s true that it could perhaps have been worded better to avoid sounding like a redefinition of what “no” means. But the writer may have been wrestling with the fear that 45 million paths and footways would eventually end up being tagged as trailblazed=no.
Well, trailblazed=no means that it is established there is no trailblazing. When trailblazed is untagged, it is unknown whether it is unmapped or untrailblazed.
Anyway, anyone is free to add something sensible there.
