To complicate the discussion even further: > I did have one question about licensing
Licencing is probably the main issue at play for any non-US data. My experience is that even what is widely considered as "publicly available imagery" is in fact subject to strict licencing restrictions. To put into perspective licencing restrictions might require to widthdraw immediately public access on demand (example: in case on conflict) and maintain a list of those who have been granted data access. In practical terms in MadMappers' case data control is the main restriction we have to comply with: following our involvement with NASA WorldWind since 2004 ( http://www.space.gov.za/worldwind/aerial_screenshot.html ) we have obtained web distribution licence for most of the data we have tiled but subject to certain restrictions. A typical example: for Johanensburg 15cm/px colour aerials we have permission to make 2.5m/px imagery available in tile format (but not as wms) to the greater public but then we do provide the full resolution feed to companies who have paid the a licence fee to the City of Johannesburg. An example of licence restrictions applicable to particular datasets appears at the bottome of the page: http://www.space.gov.za/worldwind/data.php or http://worldwindcentral.com/wiki/Add-on:ZoomIt%21 Licencing issues can therefore become extremely complex and often almost impossible for a central OAM server to handle. I am a believer of locally based servers and communities: with some effort they can obtain from local authorities (conditional) permissions and be able to give confidence to those authorities that they will be adhered to. As I have previously mentioned in an earlier post: MadMappers is willing to tile and serve any African imagery in compliance with any licence restriction. However we will refrain to serve any data which is provided without permission in writing (and that is a major component of our data archives). ciao Maurizio ----- Original Message ----- From: Brian Russo To: Maurizio Cc: Schuyler Erle ; talk Sent: Wednesday, November 11, 2009 11:02 AM Subject: Re: [OAM-talk] a dramatically simplified technical proposal If I'm not mistaken the 256^2 PNG restriction is because that is what the WMS-C standard specifies (A flaw in the standard IMO - most of our tiles are JPG). (Totally not trying to derail this main discussion) Overall I think your revised standard looks great though I admit I haven't pored over it in detail. I did have one question about licensing. "Each layer should be marked with the license of its source, including (at a minimum) descriptive text for the license, plus flags for public domain, attribution, non-commercial, and sharealike licensing. " I'm skeptical on the real utility of OAM if these sort of restrictions are put in place. My preference would be for including only public domain imagery. Anything else and I think we may just end up with a collage of differently licensed imagery and you have to jump through hoops to figure out what imagery is licensed under what etc.. Not to be dramatic, but I'd say it diminishes my interest in OAM as a large part of my interest is rooted in the problem that right now releasability/licensing issues is really the main draw for me. I.e. we can use the basemap and/or release products knowing that other people can use the basemap for "whatever". For HADR applications in particular there is often sensitive information included in products that are created. If I understand share-alike correctly then this means our derived product using OAM would need to be licensed under that. Bluntly, that just won't work. An example could be two countries that are hostile to each other and let's say you take OAM data and overlay airfield or similar sensitive data on it. You would then give a map like this to something like ICRC, USAID, etc with the caveat that the information cannot be disclosed widely, etc. You can't use sharealike data in that case, unless I'm misunderstanding the license. Even attribution can be a problem when you consider things like mobile devices where real estate is precious and the tools may not be built with this in mind. Priority for adding attribution display capability so we can use OAM is.. pretty much nonexistent. Anyway, I'm really not trying to be a naysayer or overly dramatic - feel free to point out holes in my logic if I'm misunderstanding the CC licenses, I just think such a licensing/releasability issue could fundamentally undermine the main point of OAM - which (in my eyes) is that it's a big cool skin of imagery that you can do whatever you want with. Again, I understand my usage case does not match everyone's. - bri On Tue, Nov 10, 2009 at 10:28 PM, Maurizio <[email protected]> wrote: Thanks for the outline. Image tiles should be 256x256 pixels in 8- or 24-bit compressed PNG format Being a tile person my main concern is that tile format is being too tightly restricted leaving out the vast majority of image tile formats currently in use: Google Earth / WorldWind: geographic 512px jpg Google Maps / YM / and VE: mercator 256px jpg In actual fact I think most existing OAM tiles would not qualify under the proposed format. While the variety of tile formats might seem superflous they are there for precise reasons. One of them is that internet access outside the US and Europe is both extremely slow and extremely expensive: those of you who were at FOSS4g 2008 in Cape Town last year have have experienced this first hand. Image quality is very poor as png 8bit and 24bit pngs are huge and should only be used for no-data areas. In past posts I had noted that one of OAM main aims was to make image data accessible for disaster-relief: in Africa we have lot of that. Accessibility is the key: my personal opinion is that any tile which can be viewed in OpenLayer should be considered as suitable. As madmappers.com (an African non profit initiative aimed at GIS data accessibility) we are prepare to TILE, HOST and SERVE any African data which can be distributed publicly. In actual fact madmappers could immediately start dishing out, under the OAM umbrella, a few hundreds of thousands kmsq of already tiled imagery for various African countries. But then that should probably be a separate post. ciao Maurizio ----- Original Message ----- From: "Schuyler Erle" <[email protected]> To: "talk" <[email protected]> Sent: Tuesday, November 10, 2009 9:30 PM Subject: [OAM-talk] a dramatically simplified technical proposal Hello, friends. In response to feedback both on and off list, I have rewritten the previous draft technical proposal into a much, much simpler implementation plan that reduces the initial complexity of the storage network, and describes the primary use cases: http://wiki.openaerialmap.org/Technical_Proposal I hope that this new proposal highlights the importance of the layer catalog. I have moved all of the lovely-but-possibly impractical P2P stuff here: http://wiki.openaerialmap.org/P2P_Network_Proposal Please, please, please have a look at the new technical proposal draft and send comments to the list (or just edit the wiki page). I will be attending Random Hacks of Kindness in Mountain View on Friday, and I would really like to help coordinate interested developers in hashing out the first cut of the catalog server and the cache node configuration. Please feel free to contact me if you'll be there and interested in getting involved. Thanks! SDE _______________________________________________ talk mailing list [email protected] http://openaerialmap.org/mailman/listinfo/talk_openaerialmap.org _______________________________________________ talk mailing list [email protected] http://openaerialmap.org/mailman/listinfo/talk_openaerialmap.org
_______________________________________________ talk mailing list [email protected] http://openaerialmap.org/mailman/listinfo/talk_openaerialmap.org
