1.3.386 - Still Bad Hitreg
#1
Posted 07 April 2015 - 04:46 PM
HSR still bad
Hitreg still bad
Hitboxes still bad
Lag still bad
Environmental collision boxes still bad
Phantom Damage still bad
Damage migration still bad
Enemy fire passing wide but registering damage on mech still bad
Network still bad
Major issues still unresolved
#2
Posted 07 April 2015 - 06:06 PM
ebea51, on 07 April 2015 - 04:46 PM, said:
HSR still bad
Hitreg still bad
Hitboxes still bad
Lag still bad
Environmental collision boxes still bad
Phantom Damage still bad
Damage migration still bad
Enemy fire passing wide but registering damage on mech still bad
Network still bad
Major issues still unresolved
Whats your ping? Because I'm just taking a guess that most of the problems on that list are a result of a high ping.
#3
Posted 07 April 2015 - 06:19 PM
But these problems shouldn't exist if Host State Rewind is doing its job of compensating for latency deficiencies as PGI have stated it is meant to... surely HSR is "Working as Intended"(C)?
#4
Posted 07 April 2015 - 08:12 PM
#5
Posted 07 April 2015 - 08:18 PM
#6
Posted 07 April 2015 - 10:05 PM
Postumus, on 07 April 2015 - 08:18 PM, said:
that's been happening a lot right now from what ive been seeing. ive shot an enforcer CT while it was cored dead red and I shot it as clear as it gets and hit reg indicated NOTHING. the Hit reg since this patch came out has just been NOT good.
#7
Posted 08 April 2015 - 02:47 AM
The only thing I've seen due to HSR other than trying to balance terrible ping with good ping, is that on occasion the servers delay the registration of damage until a later shot when there are spikes of a lot of damage being done in the course of a few seconds.
#8
Posted 08 April 2015 - 05:58 AM
ebea51, on 07 April 2015 - 04:46 PM, said:
HSR still bad
Hitreg still bad
Hitboxes still bad
Lag still bad
Environmental collision boxes still bad
Phantom Damage still bad
Damage migration still bad
Enemy fire passing wide but registering damage on mech still bad
Network still bad
Major issues still unresolved
I agree with all this. Got a ping of 91ms. Game is not fun anymore with this still going on.
#9
Posted 08 April 2015 - 07:20 AM
This is an on-and-off issue, not anything consistent. I recently replaced my router just to make sure it wasn't that. It wasn't. My connection bandwidth is fine and I don't generally see packetloss when I ping or tracert other domains, so I'm wondering if there are some issues with the server farm or the ISP where the MWO servers are hosted or perhaps something one hop away from there.
#10
Posted 08 April 2015 - 08:10 AM
But from time to time - i think when the servers are very busy - I also have very bad registration !
Last time I fired 6 times a pack of 3 x Large Lasers on a MystLynx - and only did light damage. Nothing was cored, everything ok.
#11
Posted 08 April 2015 - 12:09 PM
#12
Posted 08 April 2015 - 12:10 PM
#13
Posted 08 April 2015 - 01:05 PM
You need a much better netcode, i know this cost money, but this is something that has to be fixed, or the player base will never grow. I play since beta and the problem is still in the game, i hope you work hard on this problem cause it is crucial.
#14
Posted 08 April 2015 - 01:22 PM
#15
Posted 09 April 2015 - 02:52 PM
Soldner726, on 08 April 2015 - 01:22 PM, said:
This is the HSR in live action, it's trying to balance between the good pings and the ****** pings. Sometimes it saves up damage and unloads a crapton at once other times you are shooting where a mech is "supposed to be" instead of where the graphics engine is predicting it to be.
#16
Posted 09 April 2015 - 08:12 PM
#17
Posted 10 April 2015 - 06:59 AM
Postumus, on 07 April 2015 - 08:18 PM, said:
Just curious ... how do you decide that lasers begin to register a hit?
Do you judge by the reticle turning red?
You do realize that with an 86 ping ... the absolute minimum time you will have for hits to start registering is around 200ms (almost 1/4 of a second) simply due to your ping. Signals can't get to the server, get processed and get back to you any faster than that.
Half a second seems a bit much if that is how long it takes ... have you timed it? Or could it be closer to the 1/4 second you would expect from your ping?
Given 1/2 a second ... there could be packet loss between you and PGI ... every time a packet drops it has to be resent which will double the turn around time to about 1/2 a second. This would be inconsistent though ... sometimes it would be 1/2 a second and other times 1/4 second.
Another possibility would be server load ... if the server is taking a few hundred milliseconds to evaluate the effect of your shot via HSR calculations then you would also see about 1/2 second. However, it should not take that long ... and if it is taking that long then PGI might need to look into some server or HSR code optimization.
As for HSR and hitreg in general ... it gets worse the larger the pings are between the target and the firer since the server image of what is happening is less in sync with the two clients. A 250ms ping is 1/2 second round trip ... add to that the ping of the target and the 2 clients and server are juggling where the target and firer really are and exactly where the firer is aiming at the moment the trigger was pressed.
A 170kph mech is moving almost 50m/s. In the time it takes your client to register that the target has changed direction it could have moved about 12m (in that 1/4 second) ... which is at least twice its height and usually several times its width. In the meantime the server is also updating the target position from target data and trying to calculate where your client should be drawing the target in order to determine whether you hit or not. It is not an easy problem to fix. It is primarily due to the fact that MWO is server authoritative which significantly reduces hacking and cheating ... at the expense of more complex mech interaction code.
The same reason is responsible for lack of significant destructible terrain ... it is challenging to coordinate the collapse of a building among 24 clients so that the event happens simultaneously and does not give at least one player a momentary advantage (e.g. shooting through a collapsed building that has not collapsed on the target client yet).
#18
Posted 10 April 2015 - 07:34 AM
I would imagine with people with pings of 100+ shooting at people with pings 100+ would make it seem worse than it is because of an argument between clients on who shot what.
I've got a usual ping of 38ms to 60ms and don't really seem to have hit reg issues, I have aiming issues because its hard to shoot a light moving fast.
And it seems people do not realize they need to keep the laser on the same section of mech for full damage, which is hard to do with a moving target.
I will say though PPCs/ACs have such a small hit box you can get weird dodging if trying to shoot through the legs and arms.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users




























