NoComment

Well if that's the case, a set of sophisticated filter rules based on OSM data and random placement can entirely replace the AutoGen world, because more up to date OSM data can be used, a bigger variety of objects can be linked into the areas they fit and blank/little covered spots can be filled up . If it is faster to generate individual countries instead of continents a batch functionality would probably really helpful. So if you want europa it internally just does all the countries individually. Or if all states of a country are processed faster than the country itself, some kind of a recursive loop .

Those config rules seem really straight forward but really flexible at the same time . If I understand your code snippets correctly by using <filter> one can exactly specify certain criteria. Can you use two (or multiple) filters like one for longitudes and latitudes and one for land type (rual, city, etc.) to make a even more precise selection as well? Also I wondered how long does the program take to run? Is it as slow as osm2xp or is it faster? And whats the bottleneck when running it? I believe the random placement feature could improve the plausiblity and the awesomeness of the entire scenery a lot . Sometimes the OSM data is just not sufficient enough (which is now displayed as empty space which is of course not the reality) or nobody cares to record the data to OSM because it is too tedious to record it (like every single tree). Random placement would fill these empty spots (just like alpilotx tree placement alongside of roads) and could create even more detail like cattle on fields or the expamples I mentioned above.

Yeah, shortly after I asked whether you'd release it as open source, I figured OSM2XP is open source as well, didn't know that before :huh: . I think the main reasons why people decide to work on open source projects is on the one hand that they want to work with the software as well (and therefore want to improve its quality) and on the other hand that it is relativly easy to get started coding. To be honest I didn't browse through the osm2xp code, but this projects sounds like currently it is at stage where the amount of code is not too big... So if I understand you correctly it would be possible to assign a bunch of objects from a scenery library to a certain OSM tag (and maybe only in a certain area) and the resulting scenery would use those objects in this area. Or to take this concept one step further, theoretically I could map an entire scenery library like OpenSceneryX to OSM tags and your scripts would build a scenery with many, many different objects which fit to the local area? I also like your idea of creating a highly configurable program and being able to add scripts into the pipleline B) . I guess this will enable the user to create truely plausible scenerys, as one could write scripts to do cool stuff, just like alpilotx did when he randomly added trees along the streets and farmbuildings in rual areas (take a look at http://www.alpilotx.net/downloads/x-plane-10-tree-lines-and-farms/ and his before and after screenshots here: https://plus.google.com/photos/101666907909842492197/albums/5765374892383381569?banner=pwa). So one could write scripts to place reeds at the shore of lakes, put traffic signs at intersections (if no OSM data is avaiable), place shrubs and grasses all over the world, put farm animals on the meadows, replace trees with smaller plants in higher altitudes, etc. Basically the idea is just to add "random noise" objects if OSM data is insufficent to make the scenery even more plausible. Sorry for the many questions and ideas, I am just a little bit excited about this project .

Tony, first of all I just want to say your works looks really awesome ! I wondered what kind of objekts does your script link into the generated scenery, just the X-Plane Standard ones or also objects from librarys like OpenSceneryX? Using objects from OpenSceneryX could make the impression much more realistic since most american buildings don't really fit into Europe or even Asia . Also it would be really cool if you could tell us what your plans with this tool are? Are you thinking of earning money with it (by selling generated scenerys) or do you consider publishing it as open source on a platform like github (which would enable other coders to send in patchs to support other tags in OSM for example)? The later one would be really awesome and I think this tool could become a very powerful scenery generator to make x plane even more awesome .