On Sat, May 9, 2015 at 11:31 PM, Mateusz Konieczny <[email protected]> wrote:
> > Also, it would break all current data consumers. > I think the concern about data consumers in general is far higher on this tagging list, then among actual data consumers. For example: Any decent data consumer needs to process *both*: * phone=XXX +* * contact:phone=XXX* Else they're missing 100,000 data points. So even if *phone* was mechanically retagged to* contact:phone* (or the other way around) data consumers would* not even notice.* It's the 18 pages of tag soup from http://taginfo.openstreetmap.com/search?q=phone that hides phone numbers from data consumers, not a potential well discussed and documented improvement to the tagging architecture. In fact worst case is not all that bad with a mechanical retag process: if a data consumer breaks, it's because they're years out of date on following evolving tag preference. ---------------------------------------- The access tags, and contact tags are both large tag spaces created before namespaces. If invented today, they'd problably use namespaces. There are strong advantages for data processing, to have them groupable.
_______________________________________________ Tagging mailing list [email protected] https://lists.openstreetmap.org/listinfo/tagging
