Author Topic: Possible end to polybugs? [old post]  (Read 3259 times)

0 Members and 1 Guest are viewing this topic.

Offline p0ppin

  • Camper
  • ***
  • Posts: 303
  • Soldat Mapmaker
Possible end to polybugs? [old post]
« on: March 15, 2010, 10:34:06 pm »
Bringing up this old post of mine because I believe that this issue requires some attention.
...
Quote
Hello there, before I start this tyrade I'm warning that I'm making a few assumptions as to how Soldat works in the area of polygon design and their interaction with the player.  If anyone knows for sure that my assumptions are wrong, please feel free to slap me upside the head and lock this topic :P.  This will be a long, nerdy explanation full of physics jargon.  That aside, here we go :D.

We all know most of the objects in soldat are directed by some sort of physics (bullets accelerating as they fall for example), so my first assumption is going to be, to counteract the force of gravity pushing down on a standing gostek, there is an equal amount of Nomal force pushing UP on said gostek.  This must be true, or else gravity would otherwise drive our poor players into the ground through the map :(.

Case 1)  Player lands in the center of a single, "normal type" polygon.  As a result, the force of gravity - X, is cancelled by Normal force of the same magnitude.  The player remains stationary after landing.

My next assumption arises in case 2:

Case 2)  Player lands in the center of 2 overlapping "Normal type" polygons.  Since there are now 2 polygons, I assume the Normal force is now twice as large.  As a result, the force of gravity is not only negated, it is overpowered - causing a momentary upward force on the player, something we've come to call the "polybounce" or "polybug".

At this point you may ask, "But p0ppin!  I've been polybounced on a single polygon also!  How do you explain THAT?"

Oh ho I also have come up with an explanation for this too.  As I and some of you may have come to notice, the majority of these "single-polygon" bounces happen on thin polygons and also on the very edge of a polygon (where the 2 sides meet).  Which brings me to my third assumption about polygons.

Case 3[see picture below])  Player lands on the very edge of a single, semi-thin polygon.  One may think that since he is coming into contact with a single edge, there should be no upward push.  Nonetheless, the player is bounced (sometimes rocketed) upwards.  My only explanation for this is that at the thinnest part of the polygon, the players gostek comes into contact with 2 edges - the edge he lands on, and the edge beneath the surface he lands on.

The good thing about these problems (if I am correct in my assumptions) is that it can be easily fixed.  By putting a limit on the amount of normal force acting on a player, there would be no more polybugs like the cases I've described above.  Thank you for actually finishing my long, boring essay on a well-known Soldat bug.  *confetti and balloons* :P

I posted this a while ago and it never really got any closure, mostly because supporters of climbing thought it would ruin countless maps that feature the "bouncer" polygons.  This is a worry I am happy to see vanquished due to the updated polygon types.  There was another argument that this would change the gameplay too much from what we have grown used to.  I believe soldat has been held back far too long by these kinds of conservative views.  For God's sake, how far gone are some of it's players that they've actually resorted to DEFENDING a bug to preserve how the game is played? Far too long have polybugs been a part of this game - and I think its about time that we looked for a reasonable solution to them...  or maybe you'd like to keep pretending that soldat is fine the way it is.
For signatures, you are allowed only one image in your signature which may not be wider and taller than 300 and 125 pixels, and may not be over 20kB in file size. No BMPs are allowed.

Offline Horve

  • Flagrunner
  • ****
  • Posts: 692
  • Vig
Re: Possible end to polybugs? [old post]
« Reply #1 on: March 16, 2010, 12:05:31 am »
However, if normal force is over 69 (intentionally), proceed with bouncing?
Otherwise, if you want to completely eradicate bouncing and teleportation, you can imagine what I'd suggest you did with yourself.
« Last Edit: March 16, 2010, 02:11:21 am by Horve »

Offline Bistoufly

  • Soldat Beta Team
  • Flagrunner
  • ******
  • Posts: 782
Re: Possible end to polybugs? [old post]
« Reply #2 on: March 16, 2010, 01:56:35 am »
@pOppin.

I'm a climber. In fact I pass most of my soldat time in climb maps. ( Btw Thank you for the nice maps you made.)

And although it would break some of the climb maps I like to play, I totally agree with you.

Offline Fryer

  • Camper
  • ***
  • Posts: 261
  • Game dev
Re: Possible end to polybugs? [old post]
« Reply #3 on: March 20, 2010, 02:24:07 am »
Okay here is my idea about how I think it should be done:
Every frame there should be a line intersection test from the soldiers previous position to the current position. It should then get the position and normal direction of the first intersection point and go back to that position and add a countering force from the normal. If there are many intersection points together with the first one, all of them should add a force so that intentional bouncing could still be possible. Any other laws in the world of Soldat physics could be left as they are. (Yay, all problems solved!) ;)
...PC vs Mac is like AK47 vs M4A1...
<DutchFlame`> i once heard running runescape in the background gave you a speedboost
<Mr> yes, it allocates more electrons, so there are more electrons available for Soldat -> they are streched less and it is more fluent

Soldat PolyWorks 1.5.0.13 - AimMode - Aim practise gamemode script - Fryer's SoldatStream Mod

Offline L[0ne]R

  • Soldat Beta Team
  • Rainbow Warrior
  • ******
  • Posts: 2079
  • need a life. looking for donors.
Re: Possible end to polybugs? [old post]
« Reply #4 on: March 20, 2010, 03:46:14 pm »
Bouncing and teleportation should be as features, not as bugs. Also, bouncy polys are already a work-in-progress feature for 1.5.1, so hopefully it'll successfully replace the buggy way of making them.

Teleports might be gone, but they are not required to make interesting climb maps (after all, climbing is based on your own skills of moving around the map, not teleporting). Perhaps they too will be re-added as a properly-working feature someday.

It's true that polybugs can sometimes be fun and useful, but it's nothing compared to their bad effects on mapping and gameplay. Keeping polybugs just for the sake of a few "features" is simply not worth it. Soldat and game modes can still live on without them.

-----

By the way, have anyone noticed that there already is part of the "fix" done? Polygon edge graphics only appear on the outside parts of the polygons and don't appear on the overlapping parts or the "doesn't collide" polygons. Maybe that algorithm of edge rendering could be used for collision detection as well?

Offline Fryer

  • Camper
  • ***
  • Posts: 261
  • Game dev
Re: Possible end to polybugs? [old post]
« Reply #5 on: March 21, 2010, 10:57:38 am »
By the way, have anyone noticed that there already is part of the "fix" done? Polygon edge graphics only appear on the outside parts of the polygons and don't appear on the overlapping parts or the "doesn't collide" polygons. Maybe that algorithm of edge rendering could be used for collision detection as well?

Well... Edges actually do appear on all edges. It's just that edges are drawn before the polygons, so that the polygons cover them (you can see proof of this by using half-transparent polygons). :-\
Also, that wouldn't fix the case where there is just one very thin polygon.
...PC vs Mac is like AK47 vs M4A1...
<DutchFlame`> i once heard running runescape in the background gave you a speedboost
<Mr> yes, it allocates more electrons, so there are more electrons available for Soldat -> they are streched less and it is more fluent

Soldat PolyWorks 1.5.0.13 - AimMode - Aim practise gamemode script - Fryer's SoldatStream Mod

Offline VirtualTT

  • Veteran
  • *****
  • Posts: 1026
Re: Possible end to polybugs? [old post]
« Reply #6 on: March 22, 2010, 01:11:09 pm »
Edges appear on the polygon edge only if it is at least 60% not covered by overlapping polygon. However this glitches sometimes... Same for edges on the semi-transparent polygons.

Offline Neosano

  • Camper
  • ***
  • Posts: 253
  • IIAWAK!
Re: Possible end to polybugs? [old post]
« Reply #7 on: March 24, 2010, 09:52:26 am »
Edges appear on the polygon edge only if it is at least 60% not covered by overlapping polygon. However this glitches sometimes... Same for edges on the semi-transparent polygons.
And how is this related?



Retards! FIX THIS BUG!
KAWAAAAAAAIIIIIIIIII

Offline bhs

  • Major(1)
  • Posts: 43
    • My Artdump
Re: Possible end to polybugs? [old post]
« Reply #8 on: March 24, 2010, 10:18:11 am »
This is definitely an important issue for the next version of Soldat. It's really holding Soldat back, and I've seen it even determine matches outcome to a random bounce before.

When polybugs are fixed most of the maps will need to be updated. This is definitely not a reason to leave polybugs in though, and if it is, it is out of pure indolence.

However, if normal force is over 69 (intentionally), proceed with bouncing?
Otherwise, if you want to completely eradicate bouncing and teleportation, you can imagine what I'd suggest you did with yourself.
Why would you want intentionaly implent bouncing? If it's for realism then you can say most people would die by the time they were to bounce. If it's for bouncing and teleports then you should look for a more bug free method(new polygon types).

Far too long have polybugs been a part of this game - and I think its about time that we looked for a reasonable solution to them
This should be enough to have polybugs fixed, there are no other reasonable arguments, and if Enesce can code Soldat Portal then he can obviously fix this little nuisance.

Offline L[0ne]R

  • Soldat Beta Team
  • Rainbow Warrior
  • ******
  • Posts: 2079
  • need a life. looking for donors.
Re: Possible end to polybugs? [old post]
« Reply #9 on: March 24, 2010, 03:30:58 pm »
When polybugs are fixed most of the maps will need to be updated.
Why? I think the older maps with overlapping polygon structure would work just fine. If not.. well, they'll need to be made to work fine, because overlapping polygons would still be useful in some situations, and it they don't work anymore - it would be just as worth fixing as polybugs are.

if Enesce can code Soldat Portal then he can obviously fix this little nuisance.
If he wants to..... :|

Offline Horve

  • Flagrunner
  • ****
  • Posts: 692
  • Vig
Re: Possible end to polybugs? [old post]
« Reply #10 on: March 24, 2010, 04:23:05 pm »
if this was fixed, all of the current maps that have this feature would have to be remade. This could make some people really aaaannngrrry. It should have been fixed a couple of years ago, when there weren't as many maps with intentional bouncing polygons as there are now.
« Last Edit: March 24, 2010, 04:25:11 pm by Horve »

Offline p0ppin

  • Camper
  • ***
  • Posts: 303
  • Soldat Mapmaker
Re: Possible end to polybugs? [old post]
« Reply #11 on: March 24, 2010, 06:03:55 pm »
I'm sorry but I spent a lot of time climbing, and most of the "intentional bouncing" maps are utter garbage.  The bouncing polybug is barely reliable at best.  If you really want to keep them, feel free to edit the maps yourself.  I can't even think of any traditional ctf maps that use bouncers... And what would you rather have, a bunch of old climbing maps or a future soldat in which mappers don't have to spend their time fixing polybugs, and focus on more important aspects of their maps?  The choice is obvious to me.
For signatures, you are allowed only one image in your signature which may not be wider and taller than 300 and 125 pixels, and may not be over 20kB in file size. No BMPs are allowed.

Offline Horve

  • Flagrunner
  • ****
  • Posts: 692
  • Vig
Re: Possible end to polybugs? [old post]
« Reply #12 on: March 25, 2010, 05:22:57 pm »
As long as it'd be done properly

Offline bhs

  • Major(1)
  • Posts: 43
    • My Artdump
Re: Possible end to polybugs? [old post]
« Reply #13 on: March 26, 2010, 01:34:16 am »
When polybugs are fixed most of the maps will need to be updated.
Why? I think the older maps with overlapping polygon structure would work just fine. If not.. well, they'll need to be made to work fine, because overlapping polygons would still be useful in some situations, and it they don't work anymore - it would be just as worth fixing as polybugs are.
Actually forget what I said about that, you're completely right.

if this was fixed, all of the current maps that have this feature would have to be remade. This could make some people really aaaannngrrry. It should have been fixed a couple of years ago, when there weren't as many maps with intentional bouncing polygons as there are now.
This definitely should have been fixed many years ago in earlier development. Soldat is still development and it should be taken care of now, before it gets out of hand.

Polygon bugs are limiting the map structure with complex & smooth shapes, look at ctf_Ruins and say it's not buggy. It would also help mappers and players, giving the mappers less work with overlapping polygons due to a faulty engine, not to mention the smoother gameplay wouldn't hurt.

if Enesce can code Soldat Portal then he can obviously fix this little nuisance.
If he wants to..... :|
I'm not going to attack Enesce, but he's really a foolish dev if he lets this go.

This seems like a mandatory problem and needs more attention from him and the community.