Hacker News Reader

Comment by themodelplumber | Hacker News Reader

OSS's greatest leverage points have always been in the niches, the cases where _if_ it's useful in your situation, you are a very lucky person indeed.

Having read the article I think it's using a great little leverage point: Do you have a website with a map? Does the map not work anymore? ("Yikes, will my map stop working," people think)

OSS doesn't have to win big popularity battles; the little cases here and there just add up over time. Soon it's in the background of everything, even proprietary software, because it's so useful that migration and contributions to the source code make sense.

View on HN

More like anywhere vaugely rural, Google Maps seems to fail. Going out to the San Juans? Have fun with missing business listings, incorrect road traces and names, or worse.

Just in the past decade Kitsap County (just across the water from Seattle) has renumbered 300k addresses. Google Maps is buying their address database from a vendor that rarely pulls this publicly posted data, thus leaving Google a few years behind (except for spotty user corrections). OSM comparatively gets this data right from the county every week, which is great since they're renumbering 30k more addresses currently. Google will take a few more years to get any of this data if they stick to their current operational model.

Entire countries and continents are far more detailed on OSM than Google. There's a feedback cycle where having lots of active users to passively harvest data or receive user-submitted corrections from improves your map quality. Well, Google Maps is a really bloated app that can barely open on the budget Android phones people use in most developing countries (hell, GMaps pegs the CPU on a 2-year old flagship for several seconds just starting the app).

When nobody uses Google Maps, Google Maps will never have good street/address data. In developing countries there's no official government office to get this data from, just user data.

It strikes me as obvious that what Google should do is support OSM upstream and use OSM data in their product, just as what they do with Wikipedia.

Google feels it has a superior product, which is true for certain types of data in large cities. Why contribute to Wikipedia when your the publisher of Encyclopedia Britannica, you'd eat your own product alive for no notable benefit.

> use OSM data in their product

OSM has a share-alike clause, which would require G to share all their own data with everyone.

Reply on news.ycombinator.com