Kyro 2 support

by Mythor on 07/28/2002 18:18, 10 messages, last message: 10/16/2002 01:40, 1278 views, last view: 04/21/2024 20:24

i have KYRO II 64MB

the game works fine (except the Rocked launcher doesnt 'launch' the rocked, it instantly kills me)

however in 'edit' mode the problems arise

as you can see here
http://members.home.nl/hofstra/bug.jpg
and here
http://members.home.nl/hofstra/bug2.jpg

the grid isnt displayed correctly
and there is no 'selection' square

i've tested it on a GF2 MX, there it worked perfect.

every other function in the editor did work (as far as i've tested it)

but the current display / nav bug is quite iritating :)

are there more Kyro 2 users with the same problem ?

   Board Index   

#1: re: kyroII support

by buzzbomb on 07/28/2002 23:10

Not only does the rocket launcher detonate in the player's hand, but none of the other weapons fire to any useful range either, i.e. they don't do any damage to enemies. I can plug away at any enemy in a DMSP map, and no damage will be done to it.

reply to this message

#2: The problem is...

by SleepwalkR on 07/28/2002 23:28

that the drivers of your Kyro 2 claim to fully support OpenGL, which they actually don't. It's no Cube bug, but a driver bug.

reply to this message

#3: Re: The problem is...

by Mythor on 07/30/2002 00:46, refers to #2

buzzbomb, yep indeed no damage at all.


SleepwalkR:

please explain how the the collision detection be a opengl thing ?

reply to this message

#4: Re: The problem is...

by SleepwalkR on 07/30/2002 01:27, refers to #3

Ask Aardappel, he will enlighten you. I'm not down with this stuff. You can also search the forum for an answer.

reply to this message

#5: Re: The problem is...

by Aardappel on 07/30/2002 12:00, refers to #3

cube uses an opengl function to find out depth of the pixel under the crosshair (and thus the position in the 3d world), which is used for shooting etc.

This is a function quake engines etc don't use, so generally sloppy driver implementations do incorrectly.

reply to this message

#6: ..

by Dying'Danny on 08/07/2002 20:53

It could be because of the Hidden Surface Removal function of the Kyro 2.

reply to this message

#7: Kyro II and Cube

by rayle on 10/15/2002 10:30

Actually, the problem you\\\'re experiencing isn\\\'t caused by \\\"sloppy\\\" drivers, as Aard was so glad to (incorrectly) point out. The problem is due to the very way that the Kyro renders. The Kyro normally does not maintain a Z-buffer at all, because unlike traditional renderers, the tile-based method which PowerVR has always used does not care \\\"where\\\" a given bit of world geometry is along the Z-axis.

There should be an option in your PowerVR driver panel to \\\"Force External Z-Buffer\\\" or something to that effect. Please try that and e-mail me (azazelle at gt.rr.com, in case you couldn\\\'t figure it out) to tell me if my hypothesis is correct.

reply to this message

#8: Re: Kyro II and Cube

by Aardappel on 10/15/2002 10:35, refers to #7

it is not incorrect at all. regardless of how a card implements depth testing and such, if a card says it implements "OpenGL", it is up to the card to correctly implement the drivers to give correct results with its hardware. Not much cube can do about hardware implementation details. The drivers are definitely the problem.

reply to this message

#9: ugh

by rayle on 10/16/2002 00:20

It's not a "problem" because it's a feature of their hardware. A misfeature, perhaps, but the "problem" stems from the fact that the Kyro architecture is so different from any other renderer out there. In any event, PowerVR works around this most of the time by automatically detecting programs and enabling or disabling certain driver features to give the best mix of compatibility and performance. In reality, all drivers do this, but because the Kyro is so radically different from most renderers, the things it enables and disables are much more pertinent to overall compatibility.

In consclusion - yes, it's PowerVR's fault, but no, it's not a "problem" per se, merely a minor incompatibility that is (thankfully) solved by a simple modification to the default driver settings.

reply to this message

#10: Different issue

by spentron@PP on 10/16/2002 01:40

I was noticing a rather nasty thing that a non-repeating type texture (black on one edge and red on the other) seems to "wrap" so I get little dashes of the wrong color on the opposite edge of the texture (and comes very close to defeating the whole purpose of what I was attempting). Texture is 32 pix square, video is TNT card. Any guesses: Cube, OpenGL, specific card?

reply to this message

   Board Index