Re: [Flightgear-devel] Next world scenery build
Martin Spott wrote: fails in the end. After the 'standard' shapefile-based scenery has proven to work we can start tackling issues like the Great Lakes shorelines and such. The SRTM water body dataset seems to give some nice improvements on this front. It does need smoothing though, since all the points are the SRTM "pole" locations - so when viewed close up it's just made up a lots of perpendicular lines. Jon ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Stefan Seifert wrote: > Martin Spott wrote: >> You can zoom in and see, if the curves match your expectation and so >> on. O.k., you probably should not submit your very first try but the >> second one might be a good guess. If the result in FlightGear looks >> much worse, then blame Curt :-) >> > But how do I know what's wrong about the first try? I simply redid my favourite airport from scratch a second time, profiting from the expience I gained the first time and because I had the desire to make it nearly perfect, I redid it a third time before submitting :-) Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Martin Spott wrote: Then, why don't you simply post a link to your work and ask someone else to have a look at it. I didn't feel offended by your decision not to try TaxiDraw, I simply can't follow your argument. That's actually an interesting idea. Didn't occur to me. So if someone wants to see my first steps in TaxiDraw, here it is... Nine LOAB.dat Description: MPEG movie ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Martin Spott wrote: You can zoom in and see, if the curves match your expectation and so on. O.k., you probably should not submit your very first try but the second one might be a good guess. If the result in FlightGear looks much worse, then blame Curt :-) But how do I know what's wrong about the first try? There are only two ways and like you said submitting it right away is probably not the best idea, which leaves only generating the scenery, which may be too difficult for some. Nine ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Stefan Seifert wrote: > I just wanted to point out, that the reason for so few people (if it > really are few) use it yet is, that it may be just too difficult and/or > time consuming to start using it. I actually took some hours to port > TaxiDraw to wxGTK-2.6 so I could finally compile it. I took an easier route and used wxX11-2.4 for TaxiDraw :-) O.k., I admit that I had a very urgent desire to get a certain airport accepted into the airport database that there was almost no room for investigation if I like TaxiDraw or not > I normally try to get something into the best shape possible before > submitting my work. Using TaxiDraw for the first time and not knowing if > I did it anything nearly correct, I just didn't feel comfortable > submitting it. Then, why don't you simply post a link to your work and ask someone else to have a look at it. I didn't feel offended by your decision not to try TaxiDraw, I simply can't follow your argument. Cheers, Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Karsten Krispin wrote: > Am Dienstag, 20. Dezember 2005 22:05 schrieb Martin Spott: > just submit your >> airport and you'll see how it looks after the next scenery update. > [...] Do you contribude C++-Code to CVS without > testing it for functionality or even for syntac correctness, you don't, > right? > > And in particular this is the same with Taxiways: > I don't believe that what I see in TaxiDraw is exactly what I get in FGFS. To > become clear: Taxiways without centerlines, because the taxiway is to big; > wrong overlapping curve-tiles (immitating a curve with many small rects, you > know..) and so on. To my experience the 'preview' in TaxiDraw gives a pretty good guess what you have to expect. This one for example: http://document.ihg.uni-duisburg.de/bitmap/FGFS/LCHM-screen05.jpg You can zoom in and see, if the curves match your expectation and so on. O.k., you probably should not submit your very first try but the second one might be a good guess. If the result in FlightGear looks much worse, then blame Curt :-) Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Am Dienstag, 20. Dezember 2005 22:05 schrieb Martin Spott: just submit your > airport and you'll see how it looks after the next scenery update. Hi Martin, but that's the point where Stefan and I and probably many other don't agree with. When I produce something, I want to see the result before pulluting the FGFS-Enviroment with broken stuff. Do you contribude C++-Code to CVS without testing it for functionality or even for syntac correctness, you don't, right? And in particular this is the same with Taxiways: I don't believe that what I see in TaxiDraw is exactly what I get in FGFS. To become clear: Taxiways without centerlines, because the taxiway is to big; wrong overlapping curve-tiles (immitating a curve with many small rects, you know..) and so on. Karsten ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Martin Spott wrote: I usually don't judge a piece of software after just one single use - I wonder how you managed to stick to FlightGear as it definitely has some rough edges for the first-time user. Having at least a second try with TaxiDraw does not only get you into routine, you probably also have the chance to explore additional features. _And_, you don't have to build the scenery - just submit your airport and you'll see how it looks after the next scenery update. I'm sorry, I did not want to offend anyone and I for sure did not judge TaxiDraw. If I had to, I'd say it's a great tool, as it allowed even me to somehow model that airport quite easily. I just wanted to point out, that the reason for so few people (if it really are few) use it yet is, that it may be just too difficult and/or time consuming to start using it. I actually took some hours to port TaxiDraw to wxGTK-2.6 so I could finally compile it. Never submitted the patch though, because someone else obviously did the same and without CVS history it was nearly impossible to merge the changes. I'm for sure no one that gives up too early because of a few rough edges. I normally try to get something into the best shape possible before submitting my work. Using TaxiDraw for the first time and not knowing if I did it anything nearly correct, I just didn't feel comfortable submitting it. Also like I said, it's not too much fun to do something and having to wait for months before being able to try it out and see the results. Maybe I, or someone else finds the time somewhere to try to make the scenery generation part easier. So I hope you accept my apologies. Nine ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Paul Surgeon wrote: We're still stuck not being able to model airports properly. TaxiDraw is a great tool but I really don't like being limited to rectangular taxiway sections - they look awful. I started playing with the idea of modeling them as a 3D model in Blender and sticking that into the terrain instead. Yes it's a lot slower but you can get things spot on like the real thing. A few decent, well known airports would be nice in FG. I'm sure Curt would love to include a few dozen of them everytime he does a scenery rebuild. :P What would be really helpful though is a way to snap the vertices of the terrain in fgsd to the vertices of a placed model to eliminate seams or a way of converting a 3D model to one of those special binary model files that the airports are in. (*.btg.gz) In the long term an automatic airport slicer/cutter would be the best option. Pre-generate or generate an airport on the fly and cut and stitch it into the underlying terrain at run time. Hi Paul, We have had past discussions about making FG specific extensions to the X-Plane airport format and possibly maintaining the extra data ourselves. Curved taxiways is very high on that list, as well as some (yet to be determined) way to lay down arbitrary taxiway and hold short markings on top. That won't let you do every thing you want to do, but could be a big step in the right direction and make future airport building better and faster. (If you haven't noticed, I like to concentrate my effort on the db/algorithms side of life, rather than the one-off hand modeling side of life.) In terms of matching airport cutouts, you could provide an exact hole to the terragear scenery builder and it would honor that and leave that exact hole cut out of the scenery. There are some issues when a hole crosses tile boundaries, you might get some transformation/math errors so your points end up being up to a pixel off (from frame to frame) in the tiles that don't own the airport object. One idea to work around this problem is to build small skirts around the airport and the cutout. That hides the gaps pretty well if both sides have skirts. Regards, Curt. -- Curtis Olsonhttp://www.flightgear.org/~curt HumanFIRST Program http://www.humanfirst.umn.edu/ FlightGear Project http://www.flightgear.org Unique text:2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Hi, Paul Surgeon schrieb: We're still stuck not being able to model airports properly. TaxiDraw is a great tool but I really don't like being limited to rectangular taxiway sections - they look awful. Work is under way for major modifications to TaxiDraw, which also target different modelling possibilities. However, this is nothing which is done in a day or even seven days ;-) Regards, Ralf ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Stefan Seifert wrote: > Well, TaxiDraw is it's own story. Took me some hours to get it running > (due to wxGTK incompatibilities that are worked on in the CVS version). > Did one airport as good as I could but never submitted because actually > looking at the result would have taken me several more hours to play > with terragear and scenery generation, which I just did not have then. > > That's not really encouraging to do more. I usually don't judge a piece of software after just one single use - I wonder how you managed to stick to FlightGear as it definitely has some rough edges for the first-time user. Having at least a second try with TaxiDraw does not only get you into routine, you probably also have the chance to explore additional features. _And_, you don't have to build the scenery - just submit your airport and you'll see how it looks after the next scenery update. Cheers, Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Paul Surgeon wrote: We're still stuck not being able to model airports properly. TaxiDraw is a great tool but I really don't like being limited to rectangular taxiway sections - they look awful. NURBS surface primitives would be a great tool instead of being stuck to rectangles; of course, a further conversion to triangles would be necessary but that could be accomplished with a postprocessing algorithm, right before including the taxiways into the final airport file. Roberto ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
On Tuesday 20 December 2005 18:17, Dave Culp wrote: > I'm still itching to get the approach light structures working at KSFO, and > I think Curt's improvements will allow accurate modeling of unusual > airports, like KLGA, where both runways are extended out over the water on > "decks". We're still stuck not being able to model airports properly. TaxiDraw is a great tool but I really don't like being limited to rectangular taxiway sections - they look awful. I started playing with the idea of modeling them as a 3D model in Blender and sticking that into the terrain instead. Yes it's a lot slower but you can get things spot on like the real thing. A few decent, well known airports would be nice in FG. I'm sure Curt would love to include a few dozen of them everytime he does a scenery rebuild. :P What would be really helpful though is a way to snap the vertices of the terrain in fgsd to the vertices of a placed model to eliminate seams or a way of converting a 3D model to one of those special binary model files that the airports are in. (*.btg.gz) In the long term an automatic airport slicer/cutter would be the best option. Pre-generate or generate an airport on the fly and cut and stitch it into the underlying terrain at run time. Paul ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Martin Spott wrote: Dave Culp wrote: Adding to that, think of the Scenery Objects database. This database has been live for several months now and we're far from the situation where we have to queue submissions because we can't cope with the workload. The opposite is the case: I'm happy for every contribution. Think of TaxiDraw: This is a great tool for creating or improving airport layouts. It appears to me that the number of X-Plane users employing TaxiDraw for their airport layouts supersedes the number of FlightGear-TaxiDraw users significantly. Well, TaxiDraw is it's own story. Took me some hours to get it running (due to wxGTK incompatibilities that are worked on in the CVS version). Did one airport as good as I could but never submitted because actually looking at the result would have taken me several more hours to play with terragear and scenery generation, which I just did not have then. That's not really encouraging to do more. Nine ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Dave Culp wrote: > On Tuesday 20 December 2005 09:55 am, Martin Spott wrote: >> ... After the 'standard' shapefile-based scenery has >> proven to work we can start tackling issues like the Great Lakes >> shorelines and such. > > I think whoever is the gate keeper for scenery improvements is going to be a > VERY busy person! Well, you know, a really good plan (TM) per se includes guidelines on how to distribute the workload :-) I'm confident that when we announce our plan to the community, this plan is clear enough about the submission guidelines that the people involved will manage to handle the submissions that might show up. Adding to that, think of the Scenery Objects database. This database has been live for several months now and we're far from the situation where we have to queue submissions because we can't cope with the workload. The opposite is the case: I'm happy for every contribution. Think of TaxiDraw: This is a great tool for creating or improving airport layouts. It appears to me that the number of X-Plane users employing TaxiDraw for their airport layouts supersedes the number of FlightGear-TaxiDraw users significantly. Creating Scenery Objects or airport layouts as well as tweaking landcover data involves a certain amount of work for the contributor himself and I think this effect alone will prevent us from being covered with contributions for a while Cheers, Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
On Tuesday 20 December 2005 09:55 am, Martin Spott wrote: > ... After the 'standard' shapefile-based scenery has > proven to work we can start tackling issues like the Great Lakes > shorelines and such. I think whoever is the gate keeper for scenery improvements is going to be a VERY busy person! Once we have the tools to make permanent improvements to the scenery we'll have lots of people working on their favorite locales. Airplanes? This is about airplanes? ;) I'm still itching to get the approach light structures working at KSFO, and I think Curt's improvements will allow accurate modeling of unusual airports, like KLGA, where both runways are extended out over the water on "decks". Dave ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Vassilii Khachaturov wrote: >> This next world scenery build will include SRTM2 data. In the USA I > [snip] >> My goal is to have everything done (for this round) by Jan 1 of the new >> year. But I reserve the right to push that date back in case I run into >> any new glitches. > Thanks! Don't forget to take the rest on the seventh day of the world > creation :-) Hah, good point ! I'm in favour of Curt's current approach because it enables us to iron out those glitches that might surface during the shapefile-based scenery generation _before_ we start major changes alias improvements in the landcover nomenclatura. It's always dangerous to work on two different ends of such a complex building because you never know where the problems lie if something fails in the end. After the 'standard' shapefile-based scenery has proven to work we can start tackling issues like the Great Lakes shorelines and such. Best regards, Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
Vassilii Khachaturov wrote: Thanks! Don't forget to take the rest on the seventh day of the world creation :-) As I go through the process of world modeling, it become very clear that God is God and I am not even close! :-) It gives me a renewed appreciation for the immensity, complexity, beauty, and variety of our little Earth and it's surroundings. I don't feel so bad about the FG scenery shortcoming when I consider who I'm competing against. :-) Curt. -- Curtis Olsonhttp://www.flightgear.org/~curt HumanFIRST Program http://www.humanfirst.umn.edu/ FlightGear Project http://www.flightgear.org Unique text:2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Next world scenery build
> This next world scenery build will include SRTM2 data. In the USA I [snip] > My goal is to have everything done (for this round) by Jan 1 of the new > year. But I reserve the right to push that date back in case I run into > any new glitches. Thanks! Don't forget to take the rest on the seventh day of the world creation :-) Vassilii ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
[Flightgear-devel] Next world scenery build
In case anyone is wondering, I just wanted to let you all know that I am indeed working on the next world scenery rebuild. I am maintaining a little blurb on my home page with status updates and ETA's for those that are really keen on tracking my progress. This next world scenery build will include SRTM2 data. In the USA I have filled the srtm voids from the USGS DEM data, outside the usa, I simply interpolate across the voids because there are no other detailed data sources generally available. I have done quite a bit of (subtle) work on the airport generator code and there have been a few other bug fixes along the way as well. Right now I am processing the vmap0 -> shapefile conversion data. Now that our basic land use/land cover data is in shapefile format, the hope is that we will be able to use more common tools to fix and update the data for specific locations. This build will also have all the latest objects from Jon's object database. My goal is to have everything done (for this round) by Jan 1 of the new year. But I reserve the right to push that date back in case I run into any new glitches. Regards, Curt. -- Curtis Olsonhttp://www.flightgear.org/~curt HumanFIRST Program http://www.humanfirst.umn.edu/ FlightGear Project http://www.flightgear.org Unique text:2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d