> Have you ever read the getstart.pdf? apparently not. I've read it once, a long while ago. But I don't feel bound by what it says, in my view the logic is that we implement what's reasonable, then change the documentation accordingly, not that we first have a documentation as god-given and only implement what is in there for all future to come. So the question to me is if z/Z for Advanced Weather is reasonable or not, not what the documentation says.
I do take the point that the documentation should be updated accordingly. > Don't we like to compare apples and oranges... Hmm, let's see, your > statement would be valid if: > a) The IAR 80 would be part of the core distribution. It isn't. So everything which is shipped with the core needs to contain warnings if it uses lots of resources? I don't think that's actually a design principle anywhere else. Neither Atmospheric Light Scattering not Advanced Weather are currently enabled by default. In order to run into the memory issues you're mentioning, you need to * switch Advanced Weather on * dial LOD bare to a high value * check 'realistic visibility' in the Advanced Weather dialog * move the max. visibility slider of Advanced Weather to a high value ( * install custom scenery, enable random vegetation, enable random buildings,...) I continue to argue that this can be taken as conscious intention of the user to render a high visibility scene, likewise that installing the IAR-80 (the b1900d,...) represents an intention to use a highly detailed airplane. I would argue that the user isn't stupid and knows that using better graphics usually costs more resources. > b) If parts of it would be a hidden requirement for the corect operation > of other aircraft, under the apparence of optional usage (it isn't), while > advanced weather is required to be on to get full advantage of the > atmospheric > scattering scheme, a thing that isn't specified anywhere (except in some > obscure forum post, or wiki page, for which you need to search really > hard and only when you know what you're looking for) Well, getting 'full advantage' is something other than 'being necessary' in my book. Atmospheric Light Scattering runs stable with Basic Weather, it just doesn't get all the goodies right. I am not happy about the state of affairs, but as I've stated a few times, I can't actually do anything about it. I'm not the maintainer of Basic Weather, I have a very vague notion where its internals are specified, and I'm not in a position to change them. I accept the point as valid, but not the responsibility as mine - I've offered again and again to help anyone who wants to drive Light Scattering properly from Basic Weather. By the way, the 'thing' is specified for instance in the release notes of 2.10, which doesn't really look like an attempt to hide it from the public to me. If you can come up with a suggestion where else it should be spelled out, let's do it. > c) it would completely override the operation of other core components > just because it can, and would do seemingly unexpected things for the > unaware > user (it doesn't). Well, it uses dds, which hasn't gone well with some folks either... You seem to have trouble accepting a simple point: z/Z is not 'a core component' of anything. It is a key binding to modify a weather parameter, and FG operates fine without it. Advanced Weather does not override it 'because it can' but because there is a good reason - it conceptually does not mix with a physically reasonable model of the visibility in the atmosphere. You may not want to model realistic modelling of the visibility, which is fine, then you're free not to use Advanced Weather. It's a choice you have. > As for the memory usage of the IAR80, you might be surprised if you > botherd to check. I don't need to - you kindly published the number in the forum at one point. :-) It's significantly more than the terrain mesh has according to your own numbers. > BTW, weren't you the one "crying" on quite a few forum > threads, some time ago, that aircraft need better 3d models/cockpits, the one > who > started rating them based on how they look... hmm... I'm saying that what you do is hypocrisy, i.e. you measure me by standard which you yourself don't adhere to. I think my own behaviour is quite consistent - I am arguing for more realistic cockpit designs, and I am arguing for more realistic environment and visuals. So please don't get this wrong, it's an important point: I think the IAR-80 is literally defining the standard of how a 5-star model can look like, and I would like to see many more. I'm not criticizing the way you did the IAR-80, I'm saying that if you think the way you did this is okay, you should also think the way Advanced Weather is implemented is okay, if you do not think Advanced Weather is implemented okay, you should change the way you handled the IAR-80. I believe in giving users a choice - those with high-end graphics cards should get the models/shaders/textures/environment settings to give them stunning visuals, but we should not throw away the alternative (legacy?) options for those who need high framerates or have weaker GPUs. But I do not believe that the high-end development should be restricted by what runs at the low end as long as it's optional. > I suppose it was just an attempt at a personal attack, but, unlike you, I > don't consider [more or less informed or documented] critique to my > work as a personal affront. No, the idea was simply to show you that you do not measure me by the same standards as you measure yourself. You yourself develop high-end visuals which don't run on some systems (see again e.g. the dds discussion), and you believe this is okay as long as it's optional. But you're not willing to accept me doing the same. As for personal attacks: "weren't you the one "crying" on quite a few forum threads" "The forums are still there if you feel the need for a choir of "oooh ", "aaahhh" and "woooow"s at your carefuly crafted screenshots" does sound a bit like that, doesn't it? > It also doesn't change the fact that visbility/fog has (had and will > have, in a sane setup) a dual purpose, that of being a _simulation_ (read > approximation) of an atmospehric phenomenon thus providing visual > hints, and of a resource management device, Yes, I know that. Which is why Advanced Weather has a slider to set max. visibility, so if you want to be sure that you don't run into trouble, you set the slider to 30 km. and the visibility will always be 30 km or less, but you get to keep the full modelling of lower visibility regions, or the vertical development of visibility and so on. That's why Advanced Weather also has the 'realistic visibility' checkbox to allow you to make a choice if you want to run the model in a less-demanding or in a fully-demanding mode. You're claiming that directly controlling visibility by z/Z is somehow the only way resource management can be implemented, but that is just not true. Advanced Weather has a resource management model which works along with the atmosphere modelling it does, not against it. It's also not true that this wouldn't be documented - Advanced Weather has a documentation where all this is spelled out. The only real point you make is that it isn't in getstarted.pdf, so the logical consequence would be to update the manual then, rather than to change the implementation, no? > As for the relation to this thread, the memory issues have been brought > up and I've expressed my concerns to how some related implementation might > affect those, if that's wrong, feel free to ignore me. Do you, or do you not agree that 20 (or 16) km terrain loaded regardless of the visibility is a sane value? Somehow, you haven't really answered the question, you're just expressing unspecified 'concerns' and attacking Advanced Weather. Note that I've been heavily using your own numbers for memory usage which you published in the forum where you argued that what's expensive about scenery is not vertex count but textures - do you no longer hold that view? * Thorsten ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_feb _______________________________________________ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
