Hello @all,
I am completely new to this board so please forgive stupid questions.
I want to use OSM streetnames&coordinates to calculate mighty border lines of postal code areas.
I also want to correct streetnames automaticaly where its possible (eg.t “stret” "instead “street”, or capital/small letters failures…
My question is
If I have to give OSM the corrected data back - where do I have to send it to fullfill license?
E.g. if I find 300 street which are spelled wrong , I have to correct them all manually - BEFORE I am allowed to use it? Or is it okay to post it for all & ask for help?
If i create condensed data which is not good readable, am I allowed to produce a better readable file which contains exactly the same data but which is much larger and more readable? Or am I asked to document the condensed file which I am using? Of course the source of data will always be correctly mentioned!!!
What about bad guesses - in my opinionmy calculated data is too unprecise & for a very special context. I think its definitly too bad for all others.(especially if I see the very high correctness of the GPS position coordinates in OSM) Do I really have to send my estimated borders to OSM - and where exactly? Its impossible to enter it manually, I have too many of the “estimations” and dont think there is a layer in concept for such unprecise data.
some people wrote me, its not wished that I write letters to officials if there are streets where official naming is unclear. (often “service” tagged streets) - I should only look for the official roard signs. Is this correct - even if it might give uncorrect data?
E.g. highway=service often seems to be anciant description and not correctly in postal environment…but I saw many to advertise for their companies with such a street name. In my opinion the official answer from town office would result in best quality - but this is not intention of OSM - right?..Its more the live tagging idea, right?
hint: for germany the “statistisches Bundesamt Wiesbaden” has many lists and data is often offered for free and any - even commercial - usage when mentioning the source. E.g. they have a large list of villiage names…what about combining such list for internal purposes/correction?
OSM requires when using such a combination data has always to flow back - is this correct? (here also question, entering all corrections manually before its allowed to use them - or post it to OSM?)
final and most important question:
I think best & easiest for me & the community would be to announce the source OSM - my thanks & describe the way of interpolation with link to a programming tutorial. But is this covered with the license?
It is not so easy to answer you when you post a bunch of long sentence questions … but I will try:
For what country do you want to calculate postalcode boundaries? Have you verified before whether there are those boundaries already in the OSM database, like it is for 100% in Germany?
my questions are all about germany. As far as I understood by definition from german post there is no real geographic postal boader. Its more a list of streets and house number ranges that define a postal area.
(large customers get there own postal number within geographical range of others)
I already found plz-gebiete.kmz (http://www.suche-postleitzahl.org/downloads) but was not able to convert it with any on&online converter tool to something I could read. Do you have another source?
Interesting with the bot, I’ll check this.
So lets reduce my question:
If I calculate only things from OSM data - what letters are needed by license to publish?
A) of course mentioning the source & warm thanks
B) description extraction method?
C) Result of Extraction? (in my case a rather size optimized file, but could offer a more readable one for publishing)
D) description estimation method?
E) Source of estimation code?
F) Estimation results?
-H) if I make a “correction” of errors like the bot would do (with risk that these are wrong), Do I have to return this data to OSM?
(I already found a small villiges far away from me, where streetnames have a prefix of the village name - but I cant drive there and check it with road sign) As far as I understood corrections and addtions always have to return to OSM, but I dont make “corrections” its more a rough estimation which is enough for my purposes.
The time consuming part of fixing street names is actually going out and checking them on the ground. (Note you must not use Google maps for this.) The most you should really do without a real survey is to add a fixme tag, although a better approach these days would be to add map notes, as they are more likely to get noticed.
Please note, though that automated map notes have a bad reputation too, as most of these have turned out to be wrong. E.g. there was a burst of reports of wrong speed limits, presumably from some phone app, that nearly all turned out to be wrong.
There are also some specialist maps that highlight discrepancies between other sources and OSM. ITO did a lot at least at one time.
On postcodes, at least in the UK, there are no hard boundaries and the official list of addresses in each post code is a proprietary database which must not be used as as source for OSM.
EN: I will try to take Goetz over again to the German subforum …
Hallo Goetz,
Wenn du genaueres über Postleitzahlen in Deutschland im Zusammenhang mit OSM wissen möchtest, dann empfehle ich dir dafür weiterhin das deutschsprachige Unterforum hier.
Wenn du dort deine Fragen Stück für Stück stellst, und auch die Antworten der anderen User dort beherzigst, dann kannst du dir eigentlich alles herleiten und dich schlaulesen zu dem was du wissen möchtest.
Probier vielleicht doch einen neuen Anlauf dort, und lies auch intensiv im OSM-Wiki nach.
