[Tagging] maxspeed:signed=no - new proposed tag

stevea steveaOSM at softworkers.com
Fri Jan 8 06:57:32 UTC 2021

On Jan 7, 2021, at 10:46 PM, Shawn K. Quinn <skquinn at rushpost.com> wrote:
> On 1/8/21 00:12, stevea wrote:
>> This point is indeed a “first step in a multi-step process of determining that answer.”  We should all take a pause to realize how much that exact sentiment is very, very frequently used as a basis for mapping in OSM.  It is indeed “not such a bad idea,” as I know personally and from and with others I have mapped with in this project that if I / we can get 85%, 90%, 95% of “what is right” into the map (leaving the other fraction not WRONG, but perhaps “mildly incomplete,”) that is a worthy contribution to OSM.  Getting it to 98%, 99% and yes, 100% “another day” is OK.  It isn’t ideal, but I’ll take “95% correct” in a map compared to “0% and I’m lost…" any day of the week.  95% earned me an A in school, anyway, and while I do strive for an A+ or perfection, I’ll take 95% if I can’t achieve 100% “today.”
>> I do consider 85% to be my personal “floor” of completeness (not quality) as I add data to OSM.  YMMV.
> 85% of all data in an area that can be entered into OSM, or just 85% of,
> say, lane counts, speed limits, etc?

I know that I (and Minh) was / were speaking of speed limits and such signage, but I often map hiking trails and the like.  I’ve also done (but don’t really any more) serious and large-scale imports.  While I realize that getting to the 85th node out of 100 on a hiking trail would “meet my threshold” of “enough data” to make a comprehensive and useful addition to the map, that isn’t always translatable to other “completeness” quantities I (or you) might map, like lane counts or speed limits.  I could imagine entering 17 out of 20 speed limits signs (as that’s how many I could punch into my GPS as I whizzed by them) on a particular run along a highway, and “I’ll get those other three the next time I drive this stretch…” as being “complete enough” for my 85% criteria.

So, it’s an objective threshold number on a usually objective quantity of data, though there is some subjectivity there, I’ll admit.  I suppose another way I could say “if it isn’t ‘solid B’ work” (as far as “completeness for this edit session”) I won’t enter it.  To underscore, I’m talking about completeness (quantity of a whole) at 85% or greater, not quality.  Quality should always be as absolutely high as possible.

I hope I’ve answered you, Shawn, if not, I’ll do my best again.


More information about the Tagging mailing list