Will Google ever use OSM data?

[Bráulio Bezerra da Silva]

With OSM data getting more and more quality, I guess it will be a logical move to Google (and others) to use it for its needs (for example in cities where they have NOTHING mapped). Though, I've never seen anyone discuss that. Thinking about this, some questions arise.

Will it ever happen?

Will this be a good thing for people who use routing services?

Will this be a good thing for OSM?

All of that happening, will Google get more involved and help OSM (for example, with Google Summer of Code, etc)?

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Bráulio Bezerra da Silva]

Hmm, an update: some GSoC OSM projects [1] were already accepted.

[1] http://socghop.appspot.com/org/home/google/gsoc2009/openstreetmap

2009/11/2 Bráulio Bezerra da Silva <brauliobezerra@...>

With OSM data getting more and more quality, I guess it will be a logical move to Google (and others) to use it for its needs (for example in cities where they have NOTHING mapped). Though, I've never seen anyone discuss that. Thinking about this, some questions arise.

Will it ever happen?

Will this be a good thing for people who use routing services?

Will this be a good thing for OSM?

All of that happening, will Google get more involved and help OSM (for example, with Google Summer of Code, etc)?

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Martin Koppenhoefer]

2009/11/2 Bráulio Bezerra da Silva <brauliobezerra@...>

With OSM data getting more and more quality, I guess it will be a logical move to Google (and others) to use it for its needs (for example in cities where they have NOTHING mapped). Though, I've never seen anyone discuss that.

on the German ML this is indeed a topic (or was some time ago). Yes, we're hoping they will use our data, why not?

Cheers,
Martin

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Jonathan Bennett-2]

Bráulio Bezerra da Silva wrote:
> Will it ever happen?

Yes.

--
Jonathan (Jonobennett)

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Ævar Arnfjörð Bjarmason]

2009/11/2 Bráulio Bezerra da Silva <brauliobezerra@...>:
> Hmm, an update: some GSoC OSM projects [1] were already accepted.
>
> [1] http://socghop.appspot.com/org/home/google/gsoc2009/openstreetmap

This doesn't mean that Google is plotting to use OSM data in the
future, just that someone submitted a well-formed GSOC proposal that
was accepted.

Google sponsors lots of things through GSOC that they probably won't
use anytime soon, or ever.

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[acrosscanadatrails]

Some parts of this message have been removed. Learn more about Nabble's security policy.

Once the Open Database Licence is sorted out at our end.

http://wiki.openstreetmap.org/wiki/Open_Database_License

I think that they will :)

Cheers,
Sam

2009/11/2 Jonathan Bennett <openstreetmap@...>

Bráulio Bezerra da Silva wrote:
> Will it ever happen?

Yes.

--
Jonathan (Jonobennett)

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Kai Krueger]

Sam Vekemans wrote:
> Once the Open Database Licence is sorted out at our end.
>
> http://wiki.openstreetmap.org/wiki/Open_Database_License
>
> I think that they will :)

Unfortunately I am not so sure and I would interpret Googles recent
activity as making it less likely for them to use OSM anytime soon. With
the move to drop TeleAtlas as map provider in the US and switch to use
their own map data including a "report a problem" feature, as well as
their increased activity with Google Mapmaker in many other countries it
seems Google is trying to enter the map data owner business with all
their might. Google has thus turned in my opinion from a potential
future user of OSM data into a major competitor. Google might even turn
into a tougher competitor for OSM than Navteq or TeleAtlas, as they are
using some of the same advantages as OSM, crowd sourcing and providing
mapping data for free, plus the added advantage of nearly limitless
marketing power and resources. I obviously still think that OSM is going
to succeed in becoming the most complete map data and has major
advantages over google's data which is the free as in freedom aspect of
the maps which google doesn't seem to have any intention in following
looking at the rather reluctant opening of the mapmaker data. However it
will make OSMs life more interesting having to focus even more on the
free as in freedom rather than free as in beer aspects to highlight the
benefits of OSM over Google to new contributors. So my guess would be
that OSM is not going to get much help from Google until it is
inevitable that OSM is unstoppable. I think it is much more likely that
one of the other mayor players like Yahoo, Microsoft or perhaps even
Garmin or Navigon embrace OSM, to fend of Google. So Google's move could
even turn out to be an advantage to OSM.

On the other hand, google has given a significant contribution towards
the new server fund back in April, sponsored SoTM in the past and
sponsored OSM projects through GSoC. So who knows what Google is
planning. I for sure don't have a clue and just thought I'd share by
uninformed 0.02$ ;-)

Kai

>
> Cheers,
> Sam
>
> 2009/11/2 Jonathan Bennett <openstreetmap@...
> <mailto:openstreetmap@...>>
>
>     Bráulio Bezerra da Silva wrote:
>      > Will it ever happen?
>
>     Yes.
>
>     --
>     Jonathan (Jonobennett)
>
>     _______________________________________________
>     talk mailing list
>     talk@... <mailto:talk@...>
>     http://lists.openstreetmap.org/listinfo/talk
>
>

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Liz-25]

On Tue, 3 Nov 2009, Kai Krueger wrote:
>  Google might even turn
> into a tougher competitor for OSM than Navteq or TeleAtlas, as they are
> using some of the same advantages as OSM, crowd sourcing and providing
> mapping data for free,


We need competition
We should be thrilled that we are also providing competition

Remember that we are all trying to improve, and we are seeing ourselves taken
being taken very seriously by a competitor. When mapping starts in earnest in
a third world country on the OSM map, then google thinks they have potential
mappers and opens mapmaker for that area.
This competition should improve both maps.



_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[yellowbkpk]

On Mon, Nov 2, 2009 at 6:42 PM, Kai Krueger <kakrueger@...> wrote:

Sam Vekemans wrote:
> Once the Open Database Licence is sorted out at our end.
>
> http://wiki.openstreetmap.org/wiki/Open_Database_License
>
> I think that they will :)

Unfortunately I am not so sure and I would interpret Googles recent
activity as making it less likely for them to use OSM anytime soon. With
the move to drop TeleAtlas as map provider in the US and switch to use
their own map data including a "report a problem" feature, as well as
their increased activity with Google Mapmaker in many other countries it
seems Google is trying to enter the map data owner business with all
their might. Google has thus turned in my opinion from a potential
future user of OSM data into a major competitor. Google might even turn
into a tougher competitor for OSM than Navteq or TeleAtlas, as they are
using some of the same advantages as OSM, crowd sourcing and providing
mapping data for free,

I know you prefaced this phrase with "might," but Google is almost certainly not going to give away the map data they built up for free. This is the difference between Google and OSM. Google's business model is based on the work they do collecting the world's data and making it easily searchable. There is definitely room in the world for OSM -- where we ask for crowd-sourced data and give it back, too.
 

On the other hand, google has given a significant contribution towards
the new server fund back in April, sponsored SoTM in the past and
sponsored OSM projects through GSoC. So who knows what Google is
planning. I for sure don't have a clue and just thought I'd share by
uninformed 0.02$ ;-)

There are two things I've repeatedly heard from several Google employees (not speaking officially, of course):
1. OSM is interesting and should not give up simply because Google is pushing into the Geo space. They are emphatic about this. I've heard it from managers and engineers both.
2. Since so much of their business is based on open source software, Google supports open source software itself as much as possible. They spend millions of dollars and hours to support open source.

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Anthony-32]

2009/11/2 Bráulio Bezerra da Silva <brauliobezerra@...>:
> With OSM data getting more and more quality, I guess it will be a logical
> move to Google (and others) to use it for its needs (for example in cities
> where they have NOTHING mapped). Though, I've never seen anyone discuss
> that. Thinking about this, some questions arise.
>
> Will it ever happen?

No.  I was going to say "not unless OSM abandons CC-BY-SA in favor of
public domain", but that's not going to ever happen either, so no.

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[John Smith-131]

2009/11/3 Anthony <osm@...>:
> No.  I was going to say "not unless OSM abandons CC-BY-SA in favor of
> public domain", but that's not going to ever happen either, so no.

Google has no problem with saying where the data comes from, they
already do this by commercial companies.

This isn't the same thing as something like GPL where if you
distribute a binary you have to distribute the source code too.

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Anthony-32]

2009/11/2 John Smith <deltafoxtrot256@...>:
> 2009/11/3 Anthony <osm@...>:
>> No.  I was going to say "not unless OSM abandons CC-BY-SA in favor of
>> public domain", but that's not going to ever happen either, so no.
>
> Google has no problem with saying where the data comes from, they
> already do this by commercial companies.

True.  I should have said "a non-copyleft license" rather than "public
domain".  CC-BY would probably be easier to implement than public
domain, actually.

> This isn't the same thing as something like GPL where if you
> distribute a binary you have to distribute the source code too.

It's geodata.  The binary is the source.

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[John Smith-131]

2009/11/3 Anthony <osm@...>:
> True.  I should have said "a non-copyleft license" rather than "public
> domain".  CC-BY would probably be easier to implement than public
> domain, actually.

Actually that's a good point, with CC-BY-SA it's obvious where it came
from, public domain can take a lot of effort to prove it's public
domain.

> It's geodata.  The binary is the source.

It's data, not source code, however that wasn't my point, my point was
GPL requires you distribute the source code if you distribute
binaries, CC-BY-SA doesn't require any such thing, you can convert the
data into raster format and still not be required to distribute the
modified data, although it is encouraged.

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Dave F.-2]

Bráulio Bezerra da Silva wrote:
> With OSM data getting more and more quality, I guess it will be a
> logical move to Google (and others) to use it for its needs (for
> example in cities where they have NOTHING mapped). Though, I've never
> seen anyone discuss that. Thinking about this, some questions arise.
>
> Will it ever happen?

It will be great if it does, but I think there's a fundamental problem
with a point you make.

IMO, OSM doesn't have more quality.
More /quantity/, yes, but quality, no.

I think G may baulk at that.
I hope I'm wrong.

Dave F.


_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Kenneth Gonsalves-2]

On Thursday 05 Nov 2009 7:52:35 am Dave F. wrote:
> IMO, OSM doesn't have more quality.
> More /quantity/, yes, but quality, no.
>

the reverse is true
--
regards
Kenneth Gonsalves
Senior Project Officer
NRC-FOSS
http://nrcfosshelpline.in/web/

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Nick Whitelegg-2]

>IMO, OSM doesn't have more quality.
>More /quantity/, yes, but quality, no.

>I think G may baulk at that.
>I hope I'm wrong.

There are obviously many areas where OSM data is absent, but.... in the
areas which have been surveyed by OSM, what way does OSM have less quality
than Google? I'd say Google's data quality is poorer, in the sense that
many features are simply not there (you try finding an urban footpath or a
rural path on Google), or footpaths are sometimes shown as roads, etc.

Nick

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Anthony-32]

On Thu, Nov 5, 2009 at 8:07 AM, Nick Whitelegg
<Nick.Whitelegg@...> wrote:
> I'd say Google's data quality is poorer, in the sense that
> many features are simply not there (you try finding an urban footpath or a
> rural path on Google), or footpaths are sometimes shown as roads, etc.

Yep, OSM will probably be king of the (global, vectorized) micromaps
for a long time to come.  And in my opinion it'll likely never be able
to compete with the likes of Google in terms of massive database
imports.

So that's why I don't see a marriage ever happening so long as the OSM
database is copylefted.  One possible exception to that is if Google
figures out a way to mix its proprietary data with OSM's free data
without triggering the share-alike clause.  And if that happens, it'd
probably be good for everyone - Google most likely would still
contribute back modifications it makes to the actual OSM data, they
just wouldn't release everything (much of which is not even copyright
by them) under a free license.

I'm actually not sure quite how to apply copyleft to geodata.  If I
overlay a proprietary aerial photograph with OSM data, assuming I have
permission from the copyright holder of the aerial photograph, am I
violating CC-BY-SA, because I don't (and legally cannot) release the
derivative work under CC-BY-SA?  To the extent geodata can be
copyrighted in the first place, I don't see how I'm not.  That makes
the data fairly useless for Google's purposes.

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[MP-14]

> There are obviously many areas where OSM data is absent, but.... in the
>  areas which have been surveyed by OSM, what way does OSM have less quality
>  than Google? I'd say Google's data quality is poorer, in the sense that
>  many features are simply not there (you try finding an urban footpath or a
>  rural path on Google), or footpaths are sometimes shown as roads, etc.

Well, the point is, that quality of google data is quite uniform
across entire countries (while google have better data for USA then
for India for example, the quality/completeness of data in entire
India is approximately the same, Quality across USA is also
approximately the same)

But in OSM, you can have one suburb mapped to level of individual
houses and sidewalks and suburb next to it have only streets - some of
them even missing or unlabeled. While many areas are mapped better
than google, you can't rely on the area you'll need that will be
mapped well too. So while in average the quality may be better that
google maps, in worst case (when you stumble upon some poorly mapped
place) it is not. Once even the poorly-mapped places will be better
than google, then we can say we have always better data :)

Martin

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Dave F.-2]

Nick Whitelegg wrote:

>> IMO, OSM doesn't have more quality.
>> More /quantity/, yes, but quality, no.
>>    
>
>  
>> I think G may baulk at that.
>> I hope I'm wrong.
>>    
>
> There are obviously many areas where OSM data is absent, but.... in the
> areas which have been surveyed by OSM, what way does OSM have less quality
> than Google?

My worry is not so much the bits that are missing from OSM (I'm off out
this afternoon for a walk to try & rectify that), but in the
inconsistent  way in which the data that's already collected is tagged.

 From the many discussions on these forums it's clear that different
people tagged the same elements in different ways.
That's what I believe what Google & their ilk will not be prepared to
sort out.

Again, I hope I'm proved wrong.




_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk

[Peteris Krisjanis-2]

2009/11/5 MP <singularita@...>:

>> There are obviously many areas where OSM data is absent, but.... in the
>>  areas which have been surveyed by OSM, what way does OSM have less quality
>>  than Google? I'd say Google's data quality is poorer, in the sense that
>>  many features are simply not there (you try finding an urban footpath or a
>>  rural path on Google), or footpaths are sometimes shown as roads, etc.
>
> Well, the point is, that quality of google data is quite uniform
> across entire countries (while google have better data for USA then
> for India for example, the quality/completeness of data in entire
> India is approximately the same, Quality across USA is also
> approximately the same)
>
> But in OSM, you can have one suburb mapped to level of individual
> houses and sidewalks and suburb next to it have only streets - some of
> them even missing or unlabeled. While many areas are mapped better
> than google, you can't rely on the area you'll need that will be
> mapped well too. So while in average the quality may be better that
> google maps, in worst case (when you stumble upon some poorly mapped
> place) it is not. Once even the poorly-mapped places will be better
> than google, then we can say we have always better data :)
>

While undoubtedly in Google Maps there's lot of correct data, amount
of error I have witnessed during my mapping efforts indicates that
Google doesn't care about quality nor they plan to improve it. So I
will take OSM with unclassified streets in town over Google stuff.

And so far only OSM creates motivation and momentum worth to explore
and correct these things.

And to answer to main question - I think they won't and they don't care.

Cheers,
Peter.

_______________________________________________
talk mailing list
talk@...
http://lists.openstreetmap.org/listinfo/talk