[Talk-us] Importing CobbLinc Public Transit Routes from Government Data

stevea steveaOSM at softworkers.com
Thu Feb 9 09:24:16 UTC 2023


Dear Sam:  On a spectrum, there is a gentle and kind method by which I could ASK you not to do this import and there is a blunt and rude method by which I could TELL you not to do this import.  I will refrain from the latter, and based on what I see here, I will simply advise that you not do this import.  Mike has kindly pointed out our Import Guidelines as well as that the license being used is likely not compatible with OSM’s license (ODbL), yet that does not dissuade you (and it should, yet you say “you could give credit just to be safe.”  Mmmm, that’s above the qualifications of even the most experienced OSM Contributors:  are you an attorney?)  Meanwhile, you have not answered Mike, despite your linking to ten mebibytes of XML with thousands of nodes, which quite literally at the top of the file — well, Line 8 of the file — says “DO NOT UPLOAD TO OSM.”  Have you even looked at these data in any sense?  Mike asks a very important question:  how are you going to align these data (data are plural) with existing OSM data, in the guise of ways which make up the route relations you purport to want to upload?  You obviously cannot upload the data directly (as thousands of nodes) and you have not answered Mike’s (quite legitimate) question about discovering the existing OSM data “ways” that must become members of your route relations.  This is known as “conflation with existing data” and simply must be answered for every import.

Licensing issues aside, no Import Guidelines written aside, not even demonstrable knowledge of what relations are aside (in a relational database, where relations ARE the data structures of the very route relations you propose to import!), you have the community responding to you in no uncertain terms that you are not ready to do this import.  I mean this gently and quite kindly.

Because I prefer people be gentle and kind with me, I’ll do the same with you:  please, don’t attempt this import (yet).  Gain some practice entering route relations as an OSM Contributor (volunteer) so that you may become familiar with this particular kind of data structure within OSM.  Do so one at a time, as well as "many at a time," perhaps as related groups of them.  Learn how these affects routers, renderers and other downstream use cases in OSM.  Document these in, say, a statewide or countywide wiki where it might make sense to build these into a table or otherwise reference them in the context of similar entities in your region.  (These could be road routes, bicycle routes, railway routes…there are quite a number of different types of route relations, each having both similarities to each other in structure and differences in how they affect particular renderers or routers).  Then, read our Import Guidelines and fully "drink in" what an undertaking it can be to import data into our project:  it can seem daunting for a relative novice, but it is not so steep a hill that it cannot be climbed.  But if you don’t know what a route relation is, you have no business importing a bunch of them into OSM, period.  And that’s as blunt as I’ll get.

I don’t wish to dissuade you from participating in OSM, really, I want to be encouraging.  But it appears you wish to “bite off more than you can chew.”  I don’t know if your experience with OSM is crawling or walking, but I can tell it is not the sort of running or even flying that it takes to successfully get community buy-in for a data Import.  Please, gain some experience (especially with route relations, which are at least an “intermediate” topic beyond rank novice) and perhaps you can “work up” to doing an import.  But do be advised that you can try to do too much too early (before you are ready) and that’s how this appears to me.  There’s nothing wrong with being ambitious, but you do need to be able to walk before you can run.


More information about the Talk-us mailing list