How To Climb 45°+ Slopes: Video Tutorial
#1
Posted 02 July 2013 - 06:40 PM
#2
Posted 02 July 2013 - 06:42 PM
Edited by FupDup, 02 July 2013 - 06:42 PM.
#4
Posted 02 July 2013 - 06:49 PM
#5
Posted 02 July 2013 - 07:08 PM
They should have just created deceleration based on angles, not "derp 45+ degrees? Lol 0kph"
#6
Posted 02 July 2013 - 07:13 PM
#7
Posted 02 July 2013 - 07:15 PM
Mechs are supposed to be able to climb up hills that are tall enough to hide them from LOS, using their arms/stumps/whatever. Zigzagging at an angle like that is sort of the same thing.
#8
Posted 02 July 2013 - 07:16 PM
#9
Posted 02 July 2013 - 07:36 PM
Back to the drawing board.
#10
Posted 02 July 2013 - 07:40 PM
RootBeerBaron, on 02 July 2013 - 07:13 PM, said:
Exactly! by taking the 45° slope at an angle you reduce the upward gradient to something that the mech can handle. If they wanted to make 45° slopes an absolute no go zone without the use of jump jets they should've made it so you cannot hold your footing and simply slide down any slope that is at 45° or greater.
#11
Posted 02 July 2013 - 07:41 PM
#12
Posted 02 July 2013 - 07:46 PM
Bromineberry, on 02 July 2013 - 07:41 PM, said:
I would think it would be situational, the slope might now give enough room for you to head up it on an angle. However this update was to completely stop mechs from traversing 45° slopes as stated in the command chair post:
http://mwomercs.com/...ement-behavior/
"In addition to the SlowdownAngle, there is a slope level which no mech can traverse. That angle is 45°. For an easy reference as to which slopes are 45°, take a look at Alpine Peaks. Where you see white snow, the slope is less than 45°. Where you see dark coloured rock, the slope is greater than 45°, and you cannot climb it."
But you can still climb 45° slopes using this method and therefore the update was only effective in slowing down mech movement and not actually stopping it.
#13
Posted 02 July 2013 - 07:58 PM
Since restricting movement was part of the goal, I wouldn't be surprised if they do follow it up at some point with a change to the incline at which a mech starts sliding down. Perhaps something just a bit above 45 degrees. So that you'd only be able to do this in a narrow window of borderline slope.
#14
Posted 02 July 2013 - 08:00 PM
DOH!
#15
Posted 02 July 2013 - 10:51 PM
Eh, why not. Don't think they read this anyway.
If the "pill" that each mech has also has its own "platform" to stand on in relation to the surface of the ground. Make it so that "platform" can't move in a vertical path alongside an incline that exceeds, lets say... 50-55, maybe 60 degrees tops. Assuming that incline exists in this situation. By doing so that "platform" hits a snag and becomes a no-move situation, removing the ability to move vertically up the hill as seen in that video.
Basically the platform is the mechanical means to say the legs can keep a mech balanced. The system tells it can't keep balance on that hill and stops you from moving in that direction. You can then only move downward from there, to avoid making it into a stuck situation.
All assuming that's part of what can be coded, of course. I mean, I assume so. You got that incline thing in, right?
Anyhow... in the meantime we've got hill climbers needing to realize how to see the terrain slopes to make it up the hill.
#16
Posted 02 July 2013 - 11:10 PM
FupDup, on 02 July 2013 - 06:42 PM, said:
Yep. It's exactly the same thing.
http://mwomercs.com/...-4-and-the-90s/
It blows my mind too because MW:O's amazing animation system that would allow 'mechs to walk over terrain that otherwise would just be eye candy and make the map all "corridor like" was one of MW:O's very biggest strengths.
They kicked it square in the balls.
The sad part is I made the mistake of being optimistic about this system. Like I've said before, I love the idea of lights taking hills and rough terrain faster than assaults. That's just not what is happening here, at all.
EDIT: I really want to know why every poll is going in favor of this train wreck of a system? I think a lot of people buy into whatever a system is "supposed to do" and defend it, even if it doesn't work. I've had people accuse me of not wanting a "sim" because think it's silly a 'mech can't step over something that's not even up to it's knee.
Edited by Victor Morson, 02 July 2013 - 11:12 PM.
#17
Posted 02 July 2013 - 11:15 PM
#18
Posted 02 July 2013 - 11:42 PM
Narcisoldier, on 02 July 2013 - 11:15 PM, said:
I was going to buy Project Phoenix this week.
I've said several times I'm buying it.
I'm now rethinking it. If they don't hotfix this crap because casual forum people who aren't even posting are voting for keeping it (purely on the IDEA of it, not how it actually works) - and a few that are posting are screaming how invisible walls = simulation.. I'm worried PGI might not correct this or even view it as working.
Given it literally nerfs mediums far worse than anything else (despite it's intent to buff them), simply because if you happen to turn onto a slight slope that's a TAD too high you'll find yourself limping at 30kp/h in under a second... something that isn't half as bad on an assault.. I've officially come to absolutely hate this implementation. At first I was just annoyed.
#19
Posted 02 July 2013 - 11:47 PM
Victor Morson, on 02 July 2013 - 11:42 PM, said:
I'm now rethinking it.
I've been holding out until they do something to make me happy. This system in a fully functional form might have done it, but this definitely isn't it. I need details on CW or balance fixes or something; I'm pretty worried with what this game is going to look like at launch.
I really want to buy the Phoenix 'mechs, but it's hard to justify when I can't tell whether or not I'll be around in October =/
#20
Posted 03 July 2013 - 12:03 AM
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users