Topic: Bug Reports For v0.3.1 alpha

[Morgul]

Well, it's that time again... we need to start listing all the bugs we can find, and squishing them till they go away. Here's the list so far:

The Bug List:[/u]

(Reported by Piotr Obrzut)
o Spaceship sometimes disapears (mostly after flying in darkness after hitting an asteroid)
It's quite possible that the camera can't keep up with the mesh for some reason... the mesh is moving too fast, and in one frame, the mesh is moving much rather than the camera. (perhaps we should take a look at how we're updating the camera position)[/i]

(Reported by Piotr Obrzut)
o Sometimes shooting doesn't stop after a mouse click
This is because we sometimes don't recieve a button up event from the mouse. (Instead it's sending a button click event) This is a CS problem, and we'll try to figure it out soon.[/i]

o Sometimes setting anything but the default video mode crashes the game.
We have no idea why this is. It works fine for all CS/CEL applications, but ours.[/i]

o Entities are never truly deleted.
This bug has been biting our asses for a while now. More work will be done on it.... eventually.[/i]

More to come, I assure you!

--Chris

[tuxp3]

- fullscreen doesn't seem to work (on linux atleast)
- speedometer is hard to control, its not a gradual linear adjustment, its jsut needs tweaking, and the limits seem to not work (also i noticed controls were loose, if any control at all)
- lighting is weird when going directly towards "sun" if the wrong angle is used, then sun will disappear along with the "sheen/reflection" making it appear dark for small periods of time (its a special angle that one can get, but not keep)

Thanks, Andrew

this is my first bug report for you all, but im sure more will come.. if you want me to keep the more technical bugs seperate from the more 'playablity'-related ones, i can, but right now the first is a major one..

[whitelynx]

- fullscreen: we noticed this too... It was noticed at the same time as the video-mode bug. (not sure why morgul didn't include it)
- speed: Actually, it _is_ linear, and there shouldn't be any limiting... In space, there's no limits to how fast you can go except the speed of light. :-P (and that will be passable in this game as well, using subspace) The controls are only "loose" because you're doing everything with reactionary thrust... you're only pushing against your own weight, since there's no air or ground to push on.
- lighting: This seems to be video card-dependent. I've run into this before on morgul's laptop, but it seems to work fine on both of our desktops. What video card/driver version are you using?

Dave

PS - If/when we move to the new forum system, I'm going to see if we can add some fields to the user profiles so people can put in information about their system.

[tuxp3]

- speed: Actually, it _is_ linear, and there shouldn't be any limiting... In space, there's no limits to how fast you can go except the speed of light. :-P (and that will be passable in this game as well, using subspace) The controls are only "loose" because you're doing everything with reactionary thrust... you're only pushing against your own weight, since there's no air or ground to push on.

 The only problem with that is if i hold down the up or down, it can go to values that are insane, if i try to bring it back to "neutral" this becomes a problem timing it right, and the direction of the speed is not defined which is irriating.. perhaps color would help, or instead of speed, we could have thrust, which determines the speed... Just my 2 cents on that

- lighting: This seems to be video card-dependent. I've run into this before on morgul's laptop, but it seems to work fine on both of our desktops. What video card/driver version are you using?

 my gfx card is nvidia geforce 2 32mb, and i'm using nvidia's offical drivers just a version behind because they dropped support for my card in the newer versions

[whitelynx]

The only problem with that is if i hold down the up or down, it can go to values that are insane, if i try to bring it back to "neutral" this becomes a problem timing it right, and the direction of the speed is not defined which is irriating.. perhaps color would help, or instead of speed, we could have thrust, which determines the speed... Just my 2 cents on that

1. Bringing the ship to "neutral" is as easy as releasing all the keys, since then there will be 0 thrust. If you want to stop the ship, try holding C. (read the README... it has the controls)
2. thrust instead of speed - When using the keys, the thrust would either be full or nothing. Thrust doesn't directly determine speed, it determines acceleration... so if you showed thrust, you could be moving at 324235 m/s, and if you released the keys, you'd still have 0 thrust. That's why we put numbers on the displays... since speed can go over the max of the display, that allows you to still see it. (although we probably should raise the max of the display.)

my gfx card is nvidia geforce 2 32mb, and i'm using nvidia's offical drivers just a version behind because they dropped support for my card in the newer versions

Yeah, the other card we saw this behaviour on was also an older nvidia... I'm not too sure what to say about how to fix it, unless you want to try hacking CS a bit...

Dave

[Rosencrantz]

Possible bug: High energy interactions between the ship and the flying utility tool   seem to not conserve energy properly.

Things seem to work out at lower energies, and at higher energies when the tool is stationary. But if the tool is rotating and spinning then it seems to gain and lose energy in unexpected ways after interactions.

Idea: Add some sort of command that prints various energies to stderr or to a file. Total energy, linear and rotational energies about x. y, and z. Also, do that for both the players ship and the targeted ship.  In addition to those energies have a running net on each axis for what has been added by the user.

[Morgul]

Well, this is a major bug, and we're not sure why it's acting the way it is. In out physics engine, collision is handled completely, and acurately, however, for some reason all the map objects have a fixed center point. So, things only rotate... since the center position can't move. We're working on fixing this... (We've not figured out what we're forgetting to do to let it's position update.)

--Chris

[Rosencrantz]

Hmmm. Well the inaccuracy of the rotational dynamics could be a related the fixed center point problem. Anyway, if you want some examples of some of the reactions that I'm talking about start spinning some of the larger asteroids really fast. When they collide, they do so correctly, but occasionally do really weird things.

Side note: I just realized this also could be related to how the time sensitive the collision detection algorithm is. Which could cause some of these issues.

[Morgul]

Hmmm. Well the inaccuracy of the rotational dynamics could be a related the fixed center point problem. Anyway, if you want some examples of some of the reactions that I'm talking about start spinning some of the larger asteroids really fast. When they collide, they do so correctly, but occasionally do really weird things.

I bet all the things you're noticing are related to this problem. It's a pretty odd thing to have something free to rotate, but not translate...

Side note: I just realized this also could be related to how the time sensitive the collision detection algorithm is. Which could cause some of these issues.

THe collision detection system is not longer time sensitive... in the sense that it doesn't matter hot fast the program's running. (I don't know if that change got into the current release or not. I think it did.) It is time sensitive in the fact that it must take time into consideration when it does the calculations :-P ;-)

[woodspider]

ok i am getting : the installer you are trying to use is corrupted or incomplete. this could be result of damaged disk, a failed download or virus. you may want to contact the author of this installer to obtain a new copy. it may be possible to skip the check using the /ncrc command line switch(not recommended) well that is the error i am getting. any suggestions?

[Morgul]

That is because of some problems with our host. We'll have it fixed in a few days.

[Rosencrantz]

Here are some Graphical Glitches I found.

1) The lens flare and main light source (aka star) show through the body of the ship when they probably shouldn't.
2) The lens flare should only occur at small angles between the forward camera axis and star ship axis. Currently if the star is visible on screen you get lens flare. Kind odd looking, not sure what the actual physical range of angles should be, but small seems ok.

3) At large distances away from the star (definetly at around where the largest asteroid is about half the size of the disk of the star) if you turn far enough away from the star then the ship disappears. It looks like it's some lighting effects gone wonky. The angle that it disappears at is not set. If you get far enough away (a minute or two of just thrusting away) when the star leaves the screen the ship disappears).

[Morgul]

There has been a good deal of discussion about lensflare by the people who are working on the CrystalSpace Engine. Basically, the lensflare is a hackjob of code. For now, I'm thinking of removing it, since they are completely re-writing it. However, I will pass on yout comments to them, because all the feedback they can get is nice.

As far as the disapearing ship, I never saw that, until I went back and tried it. Yes, it cetainly does go away. I think  it's the lighitng system; but I'm not sure. Currently, in the development code we have the ship is completely unlit, and we don't know why... So, when we fix that, we'll fix the disapearing ship problem too! 

[PieceMaker]

I was playing a few minutes ago and I decided to crash into the asteroid, when I did, it sent me flying, and then, I couldn't see anything at all, I was shooting, turning, etc., but it still wouldn't show me. I'm not sure if that was supposed to happen so I thought I'd tell you guys.

[contingencyplan]

Yeah, I'm not quite sure why it does that (whitelynx may know better). But that's on the TODO list for the next release.

When that release will come, I do not know yet. But we're working on it, don't worry!

[Morgul]

It happens because the recoil of the collision send you flying off at such a high velocity that you leave the 'sector'. Basically, we only have one sector, which is rather bad. (The ship's position is based of the center of the sector it's in, and as all the CS majors know, there comes a point when any number datatype rolls over... when that happens the ship disapears.) It's not even a problem with the game... it's just a problem with the map. As soon as we work on it a little more (making it larger, adding sectors, etc.) the problem *should* go away. Now that I've said all this, whitelynx will post explaining why I'm completely wrong. Oh, well.

[Caenus]

Yeah, the speedometer has been mentioned before.  I tried the stuff mentioned (specifically holding C) and I could never get the ship to stop, it would decelerate to approximately 8000 units and then apparently alter directions because it would then jump right back up to 15000.  It's next to impossible to adjust the forward/reverse thrust to control your speed because after a minute or so of flying around, there is no point of reference.  All stars disappear.

[Morgul]

Yeah. We need to come up witha  good way of *faking* stars, or some other background reference. I want to make a dust particle cloud that sits in most of the sector, but with my luck that will slow everythingto a crawl. Dave and I are looking into what we can do... But we'd like some suggestions. (keep in mind this is all just map design stuff.. something I personally am not good with. I know what works, and what doesn't, but I don't know how to get from not working to woring. :P)

[Caenus]

The thing I do on CS maps is create a HUGE box with a sky texture skin.  Then I place the map itself within the box - instant sky. 

[Morgul]

Yes, a skybox. Problem is, I don't know how well it would work for space... but it might. We've talked about it... we'll have to give it a a shot.