Standing Jumps (split from Minkata Shard Updates)

Discussions about the OpenUru.org Minkata test shard

Moderator: rarified

User avatar
D'Lanor
Member
Posts: 142
Joined: Tue Dec 23, 2008 11:23 pm

Re: Standing Jumps (split from Minkata Shard Updates)

Post by D'Lanor »

As expected this still doesn't do anything for me. No matter how I jump (running, walking, standing) or from what kind of slope, I never get that knee-bend landing. I did get it in the previous version but that must have been caused by the directional factor which has been fixed.
Christian Walther
Member
Posts: 317
Joined: Sat Dec 13, 2008 10:54 am

Re: Standing Jumps (split from Minkata Shard Updates)

Post by Christian Walther »

Thanks for testing, Treehugger and D'Lanor! You’re right, it’s confusing.
  • The processor speed/frame rate issue is still present. It can indeed be direction-dependent (it is for me on Relto, for example) because the frame rate depends on the complexity of the scene you have in view. The symptom is that your standing jump abruptly stops midway through (or you barely get off the ground at all), and you don’t get a landing animation at all. (We’ll be taking a stab at fixing this for some cases, but still not all, in the next testing round.)
  • The direction-dependence Annabelle found should be fixed now. You should never get a full-kneeling landing when jumping forward now, and never a small foot-shuffling one when landing straight down.
So, you need to distinguish between the wrong landing and no landing at all. It is still possible that you get no landing animation at all (and I think this is what at least D'Lanor is seeing), but it should not happen anymore that you get the wrong one of the two possible landing animations, which are knee-bend for landing straight down (e.g. after standing jump), and foot-shuffle for landing with some forward velocity (e.g. after a running or walking jump).

Also, this is all for jumps on a horizontal plane. If you jump up or down a slope, you also don’t get a landing animation in some cases (I forgot what the exact criterion is), but that is intended.
User avatar
Annabelle
Member
Posts: 118
Joined: Thu Dec 22, 2011 4:40 am

Re: Standing Jumps (split from Minkata Shard Updates)

Post by Annabelle »

Christian wrote
•The direction-dependence Annabelle found should be fixed now. You should never get a full-kneeling landing when jumping forward now, and never a small foot-shuffling one when landing straight down.
I confirm

This is what I've tested earlier today playing my "Wheel of Prediction" marker quest in Library Courtyard.

ADDED after edition: Right, I tested the stand jumps in all directions too, we get the full kneeling on 360 degrees. (Library Courtyard)

We land correctly. Note that this courtyard is flat so I didn't see what's the effect on upward or downward planes yet.
Annabelle-Sophie KI 4247 & Annabelle KI 5152 both members of DRC (2) :) & the lovable Grr KI 106414 sole member of Grr's Hood
User avatar
Annabelle
Member
Posts: 118
Joined: Thu Dec 22, 2011 4:40 am

Unlocking the framerates

Post by Annabelle »

Hi,

I'm coming back with this subject: "Unlocking the framerates" because I want to push it in the forefront and not let it go unsolved.

Maybe you will say I'm repeating but it's crucial to understand this fact.

Having the framerates locked on a low number (dating from 2003 videocards capabilities) is just good to do one thing: segregate those who "had the bad idea" (sarcastic here) to buy a new computer in the past 4-5 years. There's nothing enjoyable having a new computer and playing on MOULa (by extend on Minkata). I'm the living proof. My new computer is used for 99% of my apps. I merely use my old computer for anything else other than playing on MOULa and Minkata shards.

Every day that I'm talking with a skydiver or a skydiving afficionado, I have to repeat to them: "Sorry! You bought a new PC, you have to live with the consequences of your choices..." Which statement is totally absurd! It should be the exact opposite.

On Gehn & TOC MOUL, we see that no matter how old or how fast or how new your PC is, the differences (yes there are some, I know, I have 2 PCs) are so minors that everyone can enjoy the game fully.

I know your priority was to implement the natural landing/kneeling to MOULa shard. You have done it on the built 1.912. I think it should be time to work on a more beneficial change (for everyone, that matter doesn't just concern Skydivers & Afficionados): unlocking the framerates dated from the prehistorical times (aka 2003).

Thanks for considering my plea (...again ;) )
Annabelle-Sophie KI 4247 & Annabelle KI 5152 both members of DRC (2) :) & the lovable Grr KI 106414 sole member of Grr's Hood
User avatar
Lyrositor
Member
Posts: 156
Joined: Sun Feb 05, 2012 10:58 pm
Contact:

Re: Standing Jumps (split from Minkata Shard Updates)

Post by Lyrositor »

If I remember correctly, this unlocked framerate lead to the Relto-plunge bug. :? I don't remember whether this was fully fixed or not.
Lyrositor
Explorer #16601888
To D'ni, or not to D'ni. There is no question.
Image
MustardJeep
Member
Posts: 54
Joined: Sat Dec 27, 2008 8:15 pm

Re: Standing Jumps (split from Minkata Shard Updates)

Post by MustardJeep »

@Annabelle

This is actually one of those things I am currently looking at for a future project on the Plasma engine. While people as mentioned above have had some success unlocking the frame rate updating things to current hardware is a non-trivial undertaking. Frankly this would need one person at expert level on c++, DirectX, physx, and ati\nvidia video cards to do a proper job.

I haven't dared muck with the render pipeline yet (no server to test with) but just reviewing it.... Just updating the out of date stuff almost means for sure a rewrite of the whole thing to make better use of cpu\video card\newer directx\newer physx. Its one of those things that would benefit everyone but be a major headache to whom ever does it and annoy everyone else because it will take a lot longer then originally figured.

On the plus side this is the project that will upgrade our plasma to eoa or beyond.
Ehren
Member
Posts: 36
Joined: Sun Feb 05, 2012 7:20 am

Re: Standing Jumps (split from Minkata Shard Updates)

Post by Ehren »

I'd agree with the unlocking framerate thing if it weren't for all the problems it causes in MOUL... Shards with unlocked frames, get the Relto dive bug, along with other camera problems. Also, sometimes, it can make it impossible to get to the Sphere 4 areas and Engineer Hut of Ahnonay without panic linking right away. As far as I know, it has something to do with the framerate change causing things to load up in the wrong order, so you hit panic regions and camera regions before the game moves your avatar to the correct position in space for where you are linking to in an age.
Deledrius
Member
Posts: 99
Joined: Sun Dec 28, 2008 6:29 pm

Re: Standing Jumps (split from Minkata Shard Updates)

Post by Deledrius »

Basically, it's a bug we can fix that's hiding bugs we may not. :(
MustardJeep
Member
Posts: 54
Joined: Sat Dec 27, 2008 8:15 pm

Re: Standing Jumps (split from Minkata Shard Updates)

Post by MustardJeep »

I wouldn't say the frame lock is a bug hiding bugs we may not fix, it's more of a leg cast letting you at least walk even if you want to run.

my current favorite bit of insanity is.

Code: Select all

	// There's a bug in NVidia drivers on Windows 2000 for GeForce1-4 (all flavors, including MX).
	// When the amount allocated into managed memory approaches the on board memory size, the performance
	// severely degrades, no matter how little is actually in use in the current rendering. So say all
	// our d3d textures are created into managed memory at age load. Also say you are
	// consistently viewing only 5Mb of managed materials (texture + vertex buffer). So as
	// you walk through the age, the new textures you see get loaded on demand into video memory.
	// Once you've seen enough to fill the on board memory, your frame rate starts falling and 
	// continues to fall as more textures get loaded. So either the memory manager is not letting
	// go of LRU textures, or fragmentation is so horrible as to make the manager useless.
	// So on these boards and with this OS, we keep track of how much managed memory we've seen,
	// and when it reaches a threshhold, we flush managed memory with an EvictManagedResources() call.
	// There's an unfortunate glitch, and then the frame rate is fine again.
	// So if we need this workaround, we set fManagedCutoff to 1 here, and then once we have our
	// D3D device, we query for the amount of memory and set the threshhold for flushing memory
	// based on that.
There are a whole bunch of other stuff in the restrict caps, but it sums up the issues nicely. That a flag is still being set in the runtime render system "fixing" a error for a eight generation old card series when used with a three generation old operating system. :lol: It's things like this that are why I agree with Annabelle the graphics issues can and should be fixed, the level of knowledge however that is evident in the hardware level code........ The best we can hope for is a very generic interface for current hardware rather then the very detailed interface currently there for the old hardware.

:)
User avatar
Annabelle
Member
Posts: 118
Joined: Thu Dec 22, 2011 4:40 am

Re: Standing Jumps (split from Minkata Shard Updates)

Post by Annabelle »

@ MustardJeep

If I got you right... unless we move everything to a whole new engine (like Plasma), we won't see any further changes on physX engine more than what TOC MOUL & Gehn Shards already implemented (with some mitigation on the success of all the new way to handle the jumping stuffs).

Well ok.

I don't want to have 2 new servers looking like Gehn & TOC MOUL personally (with all the respect to respective teams working behind these shards). It was once funny to go everywhere...now it's plenty pityful. A new comers playing MOULa would go and say ... What (... I let you imagine the remaining... ;) )

File closed then, put that in basket 13 :lol: :lol:
Annabelle-Sophie KI 4247 & Annabelle KI 5152 both members of DRC (2) :) & the lovable Grr KI 106414 sole member of Grr's Hood
Post Reply

Return to “OpenUru.org Minkata Test Shard”