Aerosoft official retail partner for Microsoft Flight Simulator !! 
Click here for more information

Jump to content
Sign in to follow this  
B21

CumulusX! feedback

Recommended Posts

Peter - let us know what / how much feedback you want on CumulusX!

I've downloaded it and been flying the Appalacian ridges in the NE USA that I know first-hand, and am very very impressed with the ridge lift CumulusX produces. I've got some immediate comment although I'd guess you're already familiar with the points:

CumulusX! could be auto-started in the FSX startup file, and so would automatically connect to FSX, wait for the SimStart event, and start producing lift, rather than having its own window with the buttons as now, and exit when it detects fsx closing.

Simconnect can put items on the FSX menu - when you mentioned the menu before I thought that's where it was rather than on the CumulusX window - I think the FSX menu options would be similar. It *might* be worth having two independent start-stop menu (and config) options for ridge lift and thermals. Having a separate window at run-time isn't ideal I think.

I think it's worth including some .slo files in the package just for a try-out, as it took me *much* more time to find the .dat files and convert to .slo than it did to install CumulusX! in the first place. My suggestion is the .slo files default to a slo directory beneath cumulusx, rather than in cumulusx as now. I think when people try flying the test area they'll be convinced enough to search for .dat files and convert them themselves (or we'll get the FSX terrain method working...)

Great work on the vario's - I did something similar so I admire your work - I recon they're worth putting in their own 'gauges' folder and just updating panel.cfg in the DG. There could be a whole topic just on designing glider instruments. My instrument is here: http://carrier.csi.cam.ac.uk/forsterlewis/...fsx/simobjects/ which is clever because it does a total energy calculation but dumb because I couldn't work out how to get the audio to follow the TE needle...

I had a look to see how often the lift actually changed, and would suggest calculating the lift once-per-second instead of every frame would be fine.

CumulusX! is an outstanding piece of work - well done & thanks.

Ian

Share this post


Link to post
Share on other sites
My suggestion is the .slo files default to a slo directory beneath cumulusx, rather than in cumulusx as now. I think when people try flying the test area they'll be convinced enough to search for .dat files and convert them themselves (or we'll get the FSX terrain method working...)

Ian

Just a suggestion; I keep my .slo files under CCS2004\Slope Data\ , so I can use them in either FS2004 with CCS2004 as in FSX with CumulusX.

Bert

Share this post


Link to post
Share on other sites

Just a suggestion; I keep my .slo files under CCS2004\Slope Data\ , so I can use them in either FS2004 with CCS2004 as in FSX with CumulusX.

Bert

I too believe it would be to CumulusX!'s benefit to provide sample .slo file(s) to get a person to try it with the greatest of ease. But as far as a location to put them, maybe just put them in a folder to let people add the .slo file(s) to whichever folder they want. I don't have any idea as to how much space in the download it would take up offhand though. A person would also have to define the world location of where the slope file(s) is well for an inexperienced pilot, so as not to discourage them if they slipe out of the slope area unexpectedly.

sf4JC

Share this post


Link to post
Share on other sites

Hi all,

feedback in any sense is highly welcome. I admit that a few .slo-files would help a lot, and I have put an extra archive with four tiles on my webpage.

In addition, I have also provided a new tool for the creation and examination of the slope-data base which may help a bit.

regards,

Peter

Share this post


Link to post
Share on other sites

My own experience as a new user of CumulusX! v 9.11 without older version usage and with experience with CCS2004 from FS9:

- after first cumulus startup continued with "Quick Start in AutoThermal Mode"

- after connect "You should now see a number of thermal clouds around you." but clouds nowhere .-( .-)

- after few minutes with lot of experimentation i found, that thermal vizualization must be ON in weather FSX settings .-)

I thing, that it will be good to take this in installation steps.. and add step add scenery to scenery library .-) for totally unfamiliar people .-)

Thats all .-) I know, this is only banality and it is normal at first state of development, that documentation isn´t finish and perfect, but due this a lot of "less-sedulity" people take condor becouse as thay say it is RTF (by click ready to fly) .-) .-( For complexity of installation a lot of people in our comunity never try CCS .-( I wrote help in our language but it dosn´t help too .-( -)

Sněhulák .-)

Share this post


Link to post
Share on other sites

Hi Peter,

I just joined in a multiplayer session. I checked the simprobe and the autothermal checkbox. And the autothermal colorbox went orange and no autothermals were generated.

Is this one of your easter eggs? I never saw that before. (When I left the multiplayer mode and opened a free flight, the autothermal box went green again.)

Bert

Share this post


Link to post
Share on other sites
- after few minutes with lot of experimentation i found, that thermal vizualization must be ON in weather FSX settings .-)

Sorry for that, but this is the problem with quick-releas of beta versions. Documentation usually doesn't catch up. I will catch up your recommendations. I appreciate your effort to help your fellows along, because that is perhaps one of the most critical things in freeware development, that you are not able to offer proffessional support to the end-users.

And the autothermal colorbox went orange and no autothermals were generated.

That's another thing of those. Finally, the color coding of the thermal indicators is:

[grey] no information on status available or disabled or inactive

[red] error, e.g. script file missing

[orange] too early, thermals will start later (=dusk)

[blue] too late, thermals finished for today

regards,

Peter

Share this post


Link to post
Share on other sites

In the multi-user session Dirk and I had the other day I discovered that the host (me in this case) sets the season, time and the weather conditions in FSX and when in session theoptions to change these settings are greyed and can not be changed by the sessionmembers, the host included.

I mention this because I saw Peter mention the developement of script files for multi-player sessions to get everybody to have the same conditions.

Maybe those script-files are not necessary after all, since all sessionmembers automatically have to fly in the same conditions. And the charme of CX! auto-thermals is that it comes close to the real-life situation where every themal has its own unpredictable life-cycle and "being at the right place at the right time" is the challenge for every pilot.

Bert

Share this post


Link to post
Share on other sites
In the multi-user session Dirk and I had the other day I discovered that the host (me in this case) sets the season, time and the weather conditions in FSX and when in session theoptions to change these settings are greyed and can not be changed by the sessionmembers, the host included.

I mention this because I saw Peter mention the developement of script files for multi-player sessions to get everybody to have the same conditions.

Maybe those script-files are not necessary after all, since all sessionmembers automatically have to fly in the same conditions. And the charme of CX! auto-thermals is that it comes close to the real-life situation where every themal has its own unpredictable life-cycle and "being at the right place at the right time" is the challenge for every pilot.

Bert

Hi Bert,

you are perfectly right bert. If the weather settings really all came from the host-session... but are we really sure of that? I must admit that I have my doubts after all. True: Wind direction and wind speed were definitively dictated by your (the host's) settings. But please remember that (IIRC) we were both running CX! and SimProbe with equivalent settings. And after the time passing (two days without any FSX-session at all) I'm not sure if our auto-thermals were at same position and strength at one time or if we just thought it were this way because of two randomly set up thermals on our two computers coincidentally were near each other. We thought we had same conditions but at last I cannot tell if it would be reproducable.

We should really cross-check this before Peter starts to run in the wrong direction. Take care of Peter, he's the only real "Funky" we have. Me personally I don't want to see - nor hear - him, tearing his hair out because of a problem he never would have had if...

Bert: Just give me a few more hours to set my fsx up (FSUIPC now installed, just a few more installations to be done...)

Best regards

Dirk.

Share this post


Link to post
Share on other sites

Yes of course you are right Dirk. Weather settings can be a fixed setting by the host. But with CX! as it is right now the CX!-configuration can be manipulated by every user individually. And there is the random thermal thing also.

So I would humbly want to withdraw my too quickly made suggestions. :blush:

Friends again, Peter? :cloud9:

Bert

Share this post


Link to post
Share on other sites

Maybe I missed something, but I cannot find a reason, why we shouldn't have stayed friends in between.

Anyhow, indeed, AutoThermals are definitely not multiplayer compatible at this time. The reason is the thermal creation algorithm which happens in a circular bubble around the current position of the user aircraft. Thermal creation goes on in pseudo-random way, until the predefined number of thermals per square km is obtained.

Everytime, a thermal deceases, or is discarded because the user aircraft moves one and the very thermal is no longer in that bubble, a new ones are created in the current bubble, until the required number is achieved again. If you are flying alone, you experience a cirtually unlimited extension of the thermal area.

The process is repeatable, but since the locations of the new thermals are relative to the user aircraft position, another session participant will have his thermal locations almost for sure at different positions. Still, the global thermal conditions (size, ceiling, strength, density) will be the same for both, as long as everybody has the same configuration (file) and the same weather. Even if you fly the same task once again, you will probably fly a different path, and the sequence of thermals will unavoidable change.

For that reason, multiplayer requires a script file for now, which makes life a bit more complicated.

Unfortunately, a multiplayer compatible, yet "random" algorithm is not straightforward, because you can tap easily in performance pitfalls. The "easiest" way would be to create some "AutoScriptFile" which containes all thermals for that day and region. The problem is, that you don't know exactly what your region will be, because it depends on your task and the position of the other participants. It turns out, that an area of 1x2 degrees at 45° lat will have approximately 20000 entries in that list, and that area is probably way too small for a challenging task. So I'm thinking about a sort of mapping strategy, which will allow to start with a limited "template area" in a say 1x1° tile, from which the thermals in all other tiles can be derived. It is easy to create the same small tile if the initial conditions are the same for each client.

best regards,

Peter

Share this post


Link to post
Share on other sites

Peter - if the thermals are currently pseudo-random from the user position, what would happen if they were pseudo-random from the bottom-left corner of the 1-degree x 1-degree box they're in, but you only accept the thermals within the existing range of the user aircraft ? Would that give two pilots in the same 1-degree x 1-degree box the same thermals without changing your code much and keep the thermal count low?

I.e. you'd need a function that gave pseudo-random lat/long values within the user bubble, based on an external seed, giving the kind of concept illustrated below:

Regards, Ian

post-16-1200999470.jpg

Share this post


Link to post
Share on other sites
Peter - if the thermals are currently pseudo-random from the user position, what would happen if they were pseudo-random from the bottom-left corner of the 1-degree x 1-degree box they're in, but you only accept the thermals within the existing range of the user aircraft ? Would that give two pilots in the same 1-degree x 1-degree box the same thermals without changing your code much and keep the thermal count low?

You would also have to realize timing of thermals from start to finish, so as to allow users to join a session at any given point, and have the same thermals. Is this possible? or, How close to these issues are you?

sf4JC

Share this post


Link to post
Share on other sites

Very close ....

Indeed, that's what I'm trying to figure out right now. As I already wrote before the problem would not exist, if there wouldn't be the sheer number of thermals possibly existing world-wide at a time.

Ian's suggestion is actually what I'm upon to, the nitty-critty problem in the background is, that I have to keep track of all thermals which are somewhere existing for all participating players. As a matter of fact, the pseudo-random-number generator surely creates a repeatable sequence for everyone, but everybody has to run through the entire sequence.

It means also, that even a thermal that is not relevant for the moment has to be considered, because later on you may enter the area where it exists, and then you have to remember that it is already there, rather than to create new one from scratch. Of course, as long as you are far away it is not needed to make them visible.

The overall procedure resembles more the scripted thermal approach where you have (large) list of thermals, covering the entire region and window of thermal activity. It is practically impossible to hold a scriptfile for the entire earth in memory, and even more, it would be totally inefficient to search it for active cells and so on.

So my initial idea was to have something like a template cell, which maps all over the world again and again. This could keep the thermals list short enough that it could be handled. Practically, this would mean a thermal cell has a number of clones around the world. If they are far enough apart, than it is not relevant for practical soaring.

The algorithm then would produce a list of thermals per solar day which is the same for all who have identical date.

Idea: I could create a sufficient count of random numbers first, based on a seed (say date), which is later on used to fill the area of interest, based on an absolute reference nearby, as the corner of the degree tile. This could be easily repeated again and again for every new area you may enter. The sequence of filling will be equal to all participants and could be easily modified with current conditions, e.g. latitude and season.

best regards,

Peter

Share this post


Link to post
Share on other sites
I have to keep track of all thermals which are somewhere existing for all participating players.

Peter, I think there is an efficiency difference between "keep track of" and being able to recreate the thermals in a consistent way for any lat/long/time within a given bubble, and for overlapping bubbles create the same thermals in the overlapping area.

So lets say you have a function

make_thermal(int k)
which given an integer k returns a pseudo-random structure for a thermal using that as a seed
{diameter, strength, cycle_length, cycle_time_offset}
which is enough for you to define your thermal parameters in a repeatable way given the same integer k. This just simplifies the following discussion an compresses a thermal into a single integer. Define a world grid such that a 3x3 set of boxes covers the 'bubble' you want to simulate around the user aircraft, e.g. a one-tenth degree grid so the reference to a cell is (x,y) where x is integer(longitude*10), y similar. A cell to the north of a given cell is (x,y+1) etc. The critical point here is you are defining a repeatable grid reference not dependent upon the position of the user aircraft within the cell. Create a pseudo-random function that produces a list of (x_delta,y_delta,k) tuples given the cell reference (x,y) as a seed, i.e.
struct Thermal {double x_delta, double y_delta, int k};
Thermal cell_thermals[100];
cell_thermals = make_thermal_list(x,y);

For the user aircraft, you calculate the current cell reference & calculate the cell_thermals for the current cell, and also the eight cells immediately around it. So lets say you have nine cell_thermals arrays: cell_thermals_NW, cell_thermals_N, cell_thermals_NE, cell_thermals_W, cell_thermals_user_aircraft, cell_thermals_E, SW, S, SE, i.e. one for each direction around the user aircraft. When the user aircraft flies into a new cell, you have to move the arrays around and calculate the new thermals in whichever direction the user is flying.

So if one user is in the cell (x,y) and another user is in the cell (x,y+2), they will both produce the same thermals for cell (x,y+1).

Ian

Share this post


Link to post
Share on other sites

Peter, I may have some bad news. I was hang gliding in the Dillingham, HI area and noticed something that I've noticed before, but since it was always in the same spot, underneath or pretty close to the FSX thermal, I just figured it was something to do with Hawaii scenery and never looked closely at it. But today, I had automatic thermals checked and I noticed a lot more along with the many thermals in the area and over the ocean along the north side of the ridge by the airport. The thing I was seeing is in fact and to my surprise, many of the cumulus01 textures in the ocean. It must have something to do with thermals because before the automatic thermals, and maybe even before CumulusX! but after fsxcloud update, I only saw the one place where I saw this image. It was under the only thermal in the area.

I will post a small video of what I'm talking about tomarrow, to give you an idea of what may be going on. It wasn't a good video though, but I think you'll get the picture.

sf4JC

Share this post


Link to post
Share on other sites

Yeah I generally see those floating in mid-air underneath some of the thermal clouds, maybe mid-way up from the ground to the cloud. They seem to be stationary, because I just fly past them. They're two-dimensional flat planes with that strange pattern.

It's never worried me though - just a minor glitch.

Ian

Share this post


Link to post
Share on other sites

Peter told me once he had built them in intentially for testing purposes.

Bert

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.
Sign in to follow this  

×
×
  • Create New...