[OSM-talk] Yahoo Aerial Imagery vs. OSM Spirit

Joshua Marinacci joshua at marinacci.org
Thu Mar 29 17:27:27 BST 2007


I can't contribute anything until after JavaOne, but I can tell you  
what I've learned in my research for the JXMapViewer  
(www.swinglabs.org).

First, I didn't really get anywhere talking to Google about using  
their stuff. They are bound by their own contracts to their providers  
which are very restrictive. I imagine Yahoo is the same. Google's  
core restrictions were that you had to use their JS apis and you had  
to host it on a publicly available website. Those may be too  
restrictive for the types of things we want to do.

NASA data is free (at least all of it that comes from NASA itself).  
And they request only an attribution to NASA when you display the  
imagery.   It seems that we have two options for using NASA data.   
Hook into a NASA server (probably WMS or some variant) or put up our  
own server using tile data.

== the WMS route:

WorldWind is NASA's open source 3d globe viewer (and is being ported  
to Java as we speak. It's a seriously amazing piece of software). It  
accesses NASA data through some sort of tile-like api, but I haven't  
decoded it yet. I'm not sure if it's the same api you find here:

	http://wms.jpl.nasa.gov/tiled.html

This is a WMS service that covers most of the globe.


== the hosted data route:

In my quest to have a stable map server for the JXMapViewer and  
JXMapKit I put up my own tileserver. It's very limited but effective  
for certain applications.  I wrote a little piece of (highly  
inefficient) software to take a large bitmap and chop it up into  
tiles then put it on my website here:

http://maps.joshy.net/applet/
http://maps.joshy.net/bmng_tiles_1/0/outputimage.0-1-1.jpg


This is 2km data which is obviously not enough for OSM work. However,  
they have 500m data available as a bittorrent download here and may  
have higher res data elsewhere:

http://earthobservatory.nasa.gov/Newsroom/BlueMarble/ 
BlueMarble_monthlies.html
http://visibleearth.nasa.gov/view_detail.php?id=7100

This page indicates that they have 15m resolution data:
http://onearth.jpl.nasa.gov/


So we should consider taking one of these routes. Either hooking into  
a NASA server or hosting the data ourselves. If we don't need a full  
highspeed tileable client (since you typically load an image then  
spend an hour tracing it) I suggest we pursue the route of hooking  
into NASA's servers.

- Josh



On Mar 29, 2007, at 8:56 AM, SteveC wrote:

> Joshua Marinacci wrote:
>> I think this is all the more reason to focus on using NASA imagery  
>> which has no such restrictions.
>
> Great, what are you going to work on? Where do they have recent  
> high res data for?
>
>
>> - Josh
>> On Mar 29, 2007, at 8:43 AM, SteveC wrote:
>>> Frederik Ramm wrote:
>>>> But gifts can also do damage. On the one hand, we risk becoming   
>>>> influenced (or should I even sy corrupted) in or thinking by the  
>>>> very  "intellectual property/licensing/corporate attitude" that  
>>>> was one  reason to start OSM. Call it political decision-making  
>>>> or whatever -  the fact that their data is there and they have  
>>>> nothing against our  tracing OSM data from the satellite/aerial  
>>>> imagery, but they want to  tell us exactly that one way of doing  
>>>> it is ok and other way is not.
>>>
>>> What makes you think yahoo care about any of this? Their hands  
>>> are tied by the agreement(s) made with their imagery providers.  
>>> It's not like Yahoo own this stuff and can do what they like.
>>>
>>>> Look at it like this: There's litte Fred in his computerized  
>>>> hut. One  evening, existing OSM data plus Yahoo imagery goes in;  
>>>> the next  morning, a lot of OSM data comes out. That's the  
>>>> basics. Nothing else  comes out, no traces remain. Now why in  
>>>> the world do Yahoo want to  tell Fred that he may do his work  
>>>> using the applet, but he may not do  it using other means?
>>>
>>> Because they have contracts to honour?
>>>
>>>> And we're starting to jump through loops. We're starting to  
>>>> speculate  what Yahoo's reasons might be and how we might "work  
>>>> with them" - how  we might tune our technical processes so that,  
>>>> while technically less  than optimal (or more complex than  
>>>> necessary), they will fit  soneone's license model.
>>>
>>> No, I really don't think "we" are doing these things, like  
>>> working with them. There's a bunch of pontificating about the  
>>> moral wrongs and dark purposes, but only one guy, mikel, actually  
>>> going to all the effort.
>>>
>>> Really, your time is much better spent coding, suggesting a  
>>> better way, getting better imagery...
>>>
>>>> We're even starting to believe that the JOSM plugin might have   
>>>> violated their terms of use by storing a temporary file, while  
>>>> at the  same time taking it for granted that browsers may make  
>>>> any number of  temporary files accessible for eternity just  
>>>> because that is "normal".
>>>
>>> It's not a matter of belief, it's a fact that Yahoo! feel that it  
>>> goes too far. So what? So we find a way to make it work. I  
>>> couldn't give a monkeys about whether their reasoning is  
>>> technically sound. It's just not their fault, they're protecting  
>>> their investment and contractual obligations as would any sane  
>>> person. I'm sure that if they owned the images outright they'd  
>>> happily let us have them. As it turns out they don't.
>>>
>>> They have an agreement on using this stuff through their API in  
>>> principle and spirit. We ran up against that, we can still use  
>>> things like potlatch. If we can't make it happen in JOSM then  
>>> c'est la vie. So make potlatch do what you want, or go and spend  
>>> the time to build a 'better' relationship with someone else. We  
>>> really don't need to throw the baby out with the bathwater.
>>>
>>>> And there's another danger with the gift of Yahoo imagery as  
>>>> well: I  have heard people in my area say "don't bother mapping  
>>>> that, I'm sure  we will soon have better Yahoo imagery for our  
>>>> region and then wee'll  trace it from there". So the potential  
>>>> later availability of  satellite images is actually keeping  
>>>> people from making an effort to  get the data now. (Not Yahoo's  
>>>> fault, of course. But if we weren't
>>>
>>> What people? I'm not held back. Have you met any of these people?  
>>> I've met tons of people who are held back due to a slow server,  
>>> technical knowledge or what have you. Nobody's said to me they  
>>> can't be bothered to map in case Yahoo decide to release aerial  
>>> data... they'd have to go out _anyway_ to get the street names.
>>>
>>>> involved with them, then the chances of having satellite images  
>>>> for  the region soon would be much more remote, motivating some  
>>>> people to  tackle it themselves instead of waiting for better  
>>>> days.)
>>>>
>>>> For a third time: Satellite images are valuable and I would like  
>>>> to  have them. But even though Lars has used drastic words when  
>>>> he spoke  of queues on Soviet streets... I, too, find it  
>>>> tempting to walk away  in this case.
>>>
>>> Walk away from something you didn't set up, because it can't be  
>>> done with your favourite editor as it stands? Please.
>>>
>>> have fun,
>>>
>>> SteveC steve at asklater.com http://www.asklater.com/steve/
>>>
>>> _______________________________________________
>>> talk mailing list
>>> talk at openstreetmap.org <mailto:talk at openstreetmap.org>
>>> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
>> - Blasting forth in three part harmony!
>
> -- 
> have fun,
>
> SteveC steve at asklater.com http://www.asklater.com/steve/

- Blasting forth in three part harmony!


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20070329/310f94bb/attachment.html>


More information about the talk mailing list