4 ms·

> It is about its pervasiveness and presence on too many markets at once, coupled with availability of too much data. It's precisely the pervasiveness and avai

> It is about its pervasiveness and presence on too many markets at once, coupled with availability of too much data. It's precisely the pervasiveness and availability of the data that drives the quality of the mapping service. I'm curious as to how you think you can have one without the other.

feanaro 7y ago
Do you have to have an ad service, a search engine, Youtube, an email service, a smartphone OS and a variety of other things to produce good maps, though?
- kristianc 7y ago
  Ad service / YouTube is 70% revenue for the search engine Smartphone OS and location reporting feeds location data in which improves and validates the maps Fact that other services are tied to the maps drives usage of the Maps themselves, increasing their quality further Other companies, like Nokia, have tried to create a business from Maps without the install base Apple has the install base and after five years billions of dollars is approaching a usable product Wouldn't say Nokia's approach was a complete failure, but to pretend the quality is anywhere similar is delusional https://en.wikipedia.org/wiki/Here_(company)#/media/File:HERE_Maps_coverage.png https://en.wikipedia.org/wiki/Here_(company)#/media/File:HER... Certainly you aren't going to be able to create a decent Maps service without centralizing and processing an enormous amount of data
  - feanaro 7y ago
    I do feel like you avoided my actual questions, perhaps accidentally. Do you think it would be impossible to have well-made maps without Google? Do you think Google's wide and heterogeneous basket of unrelated products and services is a necessary requirement to produce well-made maps, such that no company without it couldn't hope of achieving it? For instance, would the quality of Google Maps be appreciably decreased if Gmail was moved to another company? My position is that no, it would not, so that would already be a win. I think a lot of services could be shaved off in this way without loss of map quality. On the other hand, if Google's map product is superior to everything else by such a large margin, then certainly it is a viable product in itself because people still need a high-quality map. People would still install this app on their phones, providing opportunity for map validation. Importantly, people could then be asked whether they want to share this data to improve the service, which many would do, I'm sure. > Certainly you aren't going to be able to create a decent Maps service without centralizing and processing an enormous amount of data This also seems suspicious if we take it to mean user data, acquired incidentally, from users using other, unrelated services. We already know Google uses streetcars for mapping and in this case, it is a deliberate mapping effort and therefore the process by which it is converted to maps is easy to understand. But how are we imagining incidentally collected user data (from usage of other services) to improve map quality? Yes, smartphone OS location validates it, in the sense that you may notice something is wrong if people start walking through buildings on your map. It's hard for me to imagine that there is a hard requirement for the map company to be the smartphone OS company before this can happen. I mean, even if the map was produced by some other company, it would still get used by smartphone owners, providing the same opportunity. If this map was good, it would then get used by other services as well, achieving the same feedback loop you mention. By the way, I use OpenStreetMaps on my phone and its error rate is about as high as Google Maps on the places I've tried it on. I can't help but how much better it would be if it actually had a lot of money behind it.

feanaro 7y ago

Do you have to have an ad service, a search engine, Youtube, an email service, a smartphone OS and a variety of other things to produce good maps, though?

kristianc 7y ago
Ad service / YouTube is 70% revenue for the search engine Smartphone OS and location reporting feeds location data in which improves and validates the maps Fact that other services are tied to the maps drives usage of the Maps themselves, increasing their quality further Other companies, like Nokia, have tried to create a business from Maps without the install base Apple has the install base and after five years billions of dollars is approaching a usable product Wouldn't say Nokia's approach was a complete failure, but to pretend the quality is anywhere similar is delusional https://en.wikipedia.org/wiki/Here_(company)#/media/File:HERE_Maps_coverage.png https://en.wikipedia.org/wiki/Here_(company)#/media/File:HER... Certainly you aren't going to be able to create a decent Maps service without centralizing and processing an enormous amount of data
- feanaro 7y ago
  I do feel like you avoided my actual questions, perhaps accidentally. Do you think it would be impossible to have well-made maps without Google? Do you think Google's wide and heterogeneous basket of unrelated products and services is a necessary requirement to produce well-made maps, such that no company without it couldn't hope of achieving it? For instance, would the quality of Google Maps be appreciably decreased if Gmail was moved to another company? My position is that no, it would not, so that would already be a win. I think a lot of services could be shaved off in this way without loss of map quality. On the other hand, if Google's map product is superior to everything else by such a large margin, then certainly it is a viable product in itself because people still need a high-quality map. People would still install this app on their phones, providing opportunity for map validation. Importantly, people could then be asked whether they want to share this data to improve the service, which many would do, I'm sure. > Certainly you aren't going to be able to create a decent Maps service without centralizing and processing an enormous amount of data This also seems suspicious if we take it to mean user data, acquired incidentally, from users using other, unrelated services. We already know Google uses streetcars for mapping and in this case, it is a deliberate mapping effort and therefore the process by which it is converted to maps is easy to understand. But how are we imagining incidentally collected user data (from usage of other services) to improve map quality? Yes, smartphone OS location validates it, in the sense that you may notice something is wrong if people start walking through buildings on your map. It's hard for me to imagine that there is a hard requirement for the map company to be the smartphone OS company before this can happen. I mean, even if the map was produced by some other company, it would still get used by smartphone owners, providing the same opportunity. If this map was good, it would then get used by other services as well, achieving the same feedback loop you mention. By the way, I use OpenStreetMaps on my phone and its error rate is about as high as Google Maps on the places I've tried it on. I can't help but how much better it would be if it actually had a lot of money behind it.