Bussireittejä on tässä alueella muutenkin korjailtu.

Mielestäni olen lukenut, että liikenteenjakajien kohdalla tiet jaetaan erikseen. Voi toki olla, että tämäkin on jotain toista maata koskeva tieto ja Suomen osalta tätä käytetään vain osittain. En osaa sanoa.

No tämäpä se ongelma onkin, kun Wiki on niin paljon jäljessä ja vajaana monessakin asiassa... toki ymmärtäähän tuon, että ensin karttaa vapaaehtoisesti naperreltuaan, ei herkästi tule näperrettyä vielä Wikiä päälle... :)

Hmm, muistelisin lukeneeni jostain, että nuo tiedot pitäisi olla relaatiossa (mutta ei itse rakennuksen outlinessa). Laitoin niitä nyt kuitenkin tuohon testiksi, kun tuntuu ettei OSM tunnista jos tiedot on pelkissä relationeissa.

Voi nyt toki olla, että muistan väärin ja/tai ymmärtänyt väärin niin, että nimenomaan type=building relaatiossa niitä ei tarvinnut/pitänyt olla. Taidanpa vähän tutkia, kun tulee sopiva väli!

Meinaa sitten vain, että täytyy purkaa ja/tai korjata aika kasa type=building relaatiota. Apua x)

I disagree. I'd have to drop my monitor resolution to 640x480p to be able to tell the different grey blobs apart.

But there are plenty of other volunteers to do the work.

Meh. Terrible source imagery doesn't help.

Ok, kopioin tämän jostain Pietarissa olevasta rakennuksesta, joka oli lueteltu hyvin piirretyissä rakennuksissa.

Tuo porrastyyli löytyy myös Finlaysonin kirkosta ja Aleksanterin kirkosta. En tosin tiedä portaiden korkeuksia.

Joo, en vain keksi mistä tuo syntyy.

type=building vaatii building tagin ulkoreunalle, ja osiin building=part - ainakin ymmärtääkseni. Tuolla on siis jossain tod. näk. päällekkäinen node tai node joka kuuluu vain toiseen building=part tjsp., mutta vaikka kuinka mergeilin, en löytänyt sitä oikeaa.

Thank you, I have indeed forgotten to add them! Will make a note for future!

Having now slept the night, I am fair bit less upset than I was last night all sleep deprived :)

However, my issue is with removing information/data, without providing a fix. While natural=wood is incorrect in the literal wiki definition sense, the result is still the "there is trees here!".

Arguably, I could use a better tagging for it, or even add Key:fixme for it as well.

Point being, instead of requiring a 100% accurate, perfect fully populated information of every single tree at the get go, those areas show tree cover roughly around the area nobody has surveyd and added as of now. But it's much more useful than just plain grass cover.

Now, OSM is far from perfect in the tagging department, so what is the option? landcover=trees doesn't really seem to be a supported option, but I could use that in the future for marking tree areas that have not been surveyed down to a tree and would require future resurvey / more detailed survey. But logging individual trees live is fairly time consuming process even when using satellite imagery as well :)

You brought a good point about the natural=wood in Tiergarden. Probably should have spent more time looking at the examples. Berlin is very detailed and well mapped, and I looked around a lot of the parks and saw them marked as natural=wood. I took a quick look at some of them via Google Earth (not optimal) streetview to see what's actually going on, but it never dawned on me that there is indeed a logic in the groundcover part of the tags in use in Germany as well. There's also a fair bit of landuse=forest in some of the parks, but I haven't checked their age.

Now this might be because I am more used to a different definition of forest floors in parks elsewhere in the world (where I lived, parks with trees often have all sorts of "foresty" growth under them). I suspect this is why the difference between Finnish grass and trees vs German trees with forest floor didn't register in my inspection.

In short, I think we are in agreement that the natural=wood is indeed wrong tagging scheme to use here. Solution is to use landcover=trees and perhaps fixme or some other style to denote that further work is needed, yet still retaining information that there are trees in the area, rather than open air grass. I am open to further ideas.

As to tagging genus, in hindsight I wish I had actually marked what the trees are in the area, instead of just locations roughly. As you can probably see, I've made quite a lot of various changes in a broad area - my survey is not exactly fully comprehensive unfortunately! I've marked a variety of features and their locations by hand on a map in relation to major features so that I can refer to them when tracing the aerial data. It didn't occur to me to actually note if they are leafy etc., though for some trees this would be possible to see from aerial imagery.

However, this being an opensource project, I trust that other participants ccan and will add further information and detail in the future. I consider my work mainly an updated groundwork for future adjustments, additions and fixes :)

Also I don't think there's a need to revert the deletion; it's well reasoned and has a point, even if I disagree with how it was done.

Ket's keep the discussion going on in the future! There's certainly a lot of issues that need to be discussed and due to the state of the wiki and ever changing nature of the tagging, vigilance in the correct current tagging is apreciated on my part :)

Instead of removing them out of principle _when they provide information_, see the typical well mapped parks elsewhere in the world. Berlin parks are a good place to check as well.

Or - go and mark individual trees yourself. Right now you are removing useful information in lieue of getting to a outdated wiki practice.

See - for example - Großer Tiergarten in Berlin or Kensington Gardens in London.

Feel free to replace with more accurate information, rather than remove.

area:highway=footway

Legendary. Separate schemes and styles here and there.

Sure; there's best and then there's good enough (for now). I spent quite a bit of time beforehand looking at how other parks all over the world are done, and natural=wood is pretty standard, with outliers marked with individual trees.

I spent significant time marking trees down and checking where they were irl, but I ran out of time and who knows when I'll have time to count individual trees again... :)

In the meantime, now there's at least clear areas showing where trees are for anyone to go and count them and improve that portion of the map. Certainly better this way than having it just be void of anything.

Trees are placed roughly based on a general survey. Didn't have enough time for further, accurate placement. They will need future improvements. Rest done based on aerial imagery and Virastokartta.

Hopefully will serve as a future basis for more accurate drawing of the park.

I masy have broken busroutes on Satamakatu - Papinkatu area. Don't know how they work, debugging was a struggle.

Yeah, looking at my notes, I wrote them down when figuring out how to tag them but seems I've done something in between (probably the kids disturbed) and completely forgot to finish tagging that.

I am not sure taxi:conditional and service_bus:conditional are correct at all, I need to look at the conditional time tagging more.

Oh, it seems I did, it must be a placeholder from when I was figuring out the tag for "Palvelubussi" and simply forgot to change it to a correct one.

I don't think I did? Or did I? I have no memory of such tag, let me take a look :D

Well, don't know about perfect, but I feel it's at least better than before. Quite a few of these buildings have multiple addresses, so I tried to figure out a method to do them neatly. In the future it might require a more detaield aproach if someone is interested...

I'll add those in the next changeset, or the one after that.

Probably indeed. Easy enough to create a new parking lot area around those and change them.