Marius,
If you want to start with OSM, you may end up without any issues at all switching to Google later if needed. You may also end up with several more cities, and some of them may not be named as you expect, but merging them isn’t hard, and that is one of the reasons that feature exists. Switching to Google is never hard if you have complete street address data because you can always just go back and geocode your export CSV in bulk, too.
I have been working on OSM and OSM apps and projects using OSM for many years now. I would say it is definitely AWESOME to use OSM, especially if you are doing something in a small town or a limited area.
I would say that, for a global directory, your geocoding results are likely to be more of what is ‘expected’ with Google. That doesn’t mean you will get bad results with OSM. Members have written in about geocoding issues with both APIs/datasets. With OSM you can go and make changes directly; with Google, not so much.
But don’t let that talk about accuracy push OSM out of the running for right projects. It’s cheaper, more hands on, and fixable. The best thing about OSM is that you can literally go and edit the map. I have traced many thousands of kilometers of roadways across Africa, SE Asia, and more. I love drawing in the details of parks and beaches, and I definitely recommend joining up with the OSM list serves for the area/country etc for some interesting conversation. The non-profit geotourism programs I work with uses OSM heavily and even sent a participant to last year’s conference in the Philippines. FUN!
It is definitely a good idea to consider how many listings you will have, and what you intend to do with the data in the long run. If your idea is to contribute the point data back to OSM, then using OSM is a
way to go, and I will give you a round of applause, too. OSM is a great way to setup a project for a community and to be able to hand it over without a thought about the API key or ongoing costs.
directoryfan 7 years, 1 month ago.
