[aklo]

I mean like getting the images for map view, who drew out those maps? Is it even possible to draw a map manually?
If not drawn manually, is there even a software that can convert satellite pictures to line vector drawing (I'm guessing it is vector because it scales beautifully).

Even from satellite, some buildings are grainy, can the software detect the building shapes/lines?

Im an average IT guy myself, not great just average.

[modi123_1]

I assume google's stuff is proprietary, but if you just bing "map software" you get plenty of options... like say imapbuilder.

[Syngenetic]

I don't know how to answer the question, but I want to point out that MapQuest and few other tools were created before Google Maps came out.

[DarenR]

I am guessing some little ninja draw those maps so fast that no one saw him do it...

This post has been edited by DarenR: 10 July 2012 - 06:28 AM

[jon.kiparsky]

To be quite honest, google developers didn't develop google maps. Very little of what google offers is actually original in-house development, they mostly slap their interface on companies they buy.


Sez Wikipedia:

Quote

Google Maps first started as a C++ program designed by two Danish brothers Lars and Jens Rasmussen at the Sydney-based company Where 2 Technologies. It was first designed to be separately downloaded by users but the company later pitched the idea for a purely Web-based product to Google management, changing the method of distribution.[25] In October 2004 the company was acquired by Google Inc[26] where it transformed into the web application Google Maps.

I don't know the details of the code, obviously, but for something like this you might start by worrying about representation and rendering. What does the world look like to the computer, and how do you convey that to the user? Delivery is also a problem. Google maps will let you scroll from Boston to Buenos Aires, in minute detail. Clearly, there's more data here than you can get in a page load, so you need an open connection to feed data. Given the time frame when this was deployed, this should suggest Ajax, which suggests that the rendering part is javascript and the representation part is coming in as xml - though it might have started out life in some other form.

So with that sort of half-speculation in hand, you can think about two questions:
1 - given some appropriate XML representation of a small piece of the world, say, Copley Square in Boston, how would you do something like that? Never mind the little half-grayed-out silhouettes of the buildings, just think about getting the streets in the right places and the parks distinguished from the buildings. What sort of a representation would you want?
2- Given a world to represent, how would you represent just the streets, buildings, "other land surface" and "water surface"? Remember, you want this to be continuous and scalable (remember those little images of the buildings? leave room for those) and it will ultimately be delivered in XML.

Having thought through all of that, or having thought other thoughts along other lines, you can start making guesses about how the data was acquired. At that point, your speculation will have given you a hypothesis, which you can use to extract facts from the world, which will either support or conflict with the story you're telling yourself. Repeat until satisfied.

[Nikitin]

jon.kiparsky, on 10 July 2012 - 08:44 AM, said:

Very little of what google offers is actually original in-house development, they mostly slap their interface on companies they buy.

You're making it sound like Google mostly acquires ready-made products and adds little of their own before pushing them to the public. This, of course, would be an incredibly stupid thing to say (unless you meant something else).

[blackcompe]

Quote

Given the time frame when this was deployed, this should suggest Ajax, which suggests that the rendering part is javascript and the representation part is coming in as xml - though it might have started out life in some other form.

It uses AJAX and JSON.

@aklo: I'd imagine the application is some depiction of satellite data. Have you read the Wikipedia entry?

This post has been edited by blackcompe: 10 July 2012 - 08:23 PM

[jon.kiparsky]

Nikitin, on 10 July 2012 - 10:10 PM, said:

jon.kiparsky, on 10 July 2012 - 08:44 AM, said:

Very little of what google offers is actually original in-house development, they mostly slap their interface on companies they buy.

You're making it sound like Google mostly acquires ready-made products and adds little of their own before pushing them to the public. This, of course, would be an incredibly stupid thing to say (unless you meant something else).

That is in fact what I meant to say, and it's been google's pattern. Maps, earth, youtube. Google documents is their implementation (I think it's their implementation) of a Microsoft product. Picasa was bought. Hell, they bought Android when they decided to go into phones, though they did a lot of work on Android. Google is really good at painting companies white and making their buttons look like Google's buttons. And search. But really, innovation? It's not their specialty. They're a marketing company. They sell ads. Everything else is there to bring eyeballs to ads.

It's not an unusual pattern - certainly you wouldn't be surprised if I told you Microsoft flourished by buying and rebranding other peoples' work. Why do you not believe it when you hear it about Google?
Well, I suppose it's pretty obvious, really. They're a marketing company, and the first thing they market is Google.

[Nikitin]

I never argued that Google didn't buy companies/ideas. That stuff happens everywhere.

However, once Google acquires a company, it does *not* receive a product that takes little development before it is put into the public. Saying otherwise (and is what I thought you meant) would be completely false.

The end-product that you see may contain little of the original code in it (sometimes even zero). It is Google that takes an idea/product and turns it into something much much better. 99% of Google Maps that you see today is a true Google product. It is *not* something that Google simply acquired and "slap their interface on".

[Logomachist]

I think the OP was asking how Google gets its data and how that data is transformed into a map, not asking about how the ompleted maps are presented to the user. IIRC there were already companies that had the data digitalized who lisensed it out to other companies who created electronic mapping products such as Mapquest and pre-Internet cartography software. Google just lisensed or bought this data for their maps. Then the digital data is automatically translated into graphics to be presented to the user.