Hi,
I’ve made a forest but maplint sais it is having a “not in map features” error.
All I have is natural=wood,wood=mixed as stated to be possible here http://wiki.openstreetmap.org/wiki/Tag:natural%3Dwood
This is my forest I’ve made: http://openstreetmap.com/browse/way/32462384
Looks good to me. Maplint is possibly throwing an error about the wood=mixed tagging.
I’m getting the same kind of thing. Perhaps Maplint is not aware of all the features?
Maplint doesn’t recognise all tags that are used, only those that it can learn about from someone running a script to pull the information out of the Map Features wiki page (at irregular intervals I believe, though recent code changes may have changed this). I’d agree with Alex that it is likely that wood=mixed isn’t mentioned on the Map features page (I can’t see it), although it is mentioned on the natural=wood page http://wiki.openstreetmap.org/wiki/Tag:natural%3Dwood
PS: I just added wood= to the Properties section of Map Features. It’ll get picked up sooner or later…
PS to PS: To be clear… Now it’s in map features at some point the tiles at home clients will be aware of the key when generating the transparent Maplint tiles. After this happens the tile will need to be rerendered to update the maplint layer. That’s why the changes aren’t instantly recognised.
Probably. If oneway is anything other than yes/no or -1 then it will give a warning. Updating whatever it is set to will cure the warning.
Why does Maplint give this warning for Silver Crescent (http://openstreetmap.org/?lat=51.49412&lon=-0.2778&zoom=16&layers=B000TTF) when it has oneway=yes then?
A good question. I thought at first maybe that there were two ways on top of each other, and the other one had the issue, but that wasn’t the case. I then realised all oneway roads, including those near here, and highlighted as problems, so I got the latest Tiles@Home client code and checked the maplint tests and oneway isn’t currently listed. This probably stems back to a recent change to use a standard Map Features library rather than parse it within the .pl script, so I’m now trying to find the relevant code and see why oneway is excluded.
Given that the database is designed to be extended, I wonder whether the very idea of a whitelist-based key check is flawed and could tempt people to remove legitimate keys Maplint doesnÂ’t know about just to keep it happy. Might it be better to have a blacklist of deprecated synonyms and misspellings found on Tagwatch instead?
Yes, Maplint does seem to encourage homogeneity rather than good practice. I think a blacklist would be better.
Ed, thanks for your further replies to this thread. I learned something too - I didn’t realise Maplint imported the map features page.