Pages: 1 2 3 [4] 5
  Print  
Author Topic: 0.8.0 roadmap  (Read 80092 times)
kr3am
Nub


Cakes 1
Posts: 48


Kitty at work


« Reply #75 on: June 06, 2008, 06:54:43 AM »

i have a feeling that most of the players out there, actually see the game as fromhell had intended it to be. i for one like the game the way that it is and i haven't messed around with any setting other than these

/rate 25000
/snaps 30
/cl_maxpackets 30
/cl_packetdup 1
/cg_smoothClients 1
/cg_truelightning 1
/cg_fov 112
/com_maxfps 125
/cg_lagometer
/cg_showfps
/cg_drawtimer
/com_hunkmegs 128
/com_soundmegs 16
/com_maxfps 125

because for me it would change it for the worst. i couldn't stand to look at it with vertex and all the settings to make it easier to see in the shadows etc ugh. i like eye candy and so do most non pro people, we play for fun yeah. i have learned to tell when others are using these things and worse some people can actually see through the walls etc, these people are a joke, where is the skill in that lol, and you know that even though i am no pro i can beat them 1 on 1 most times coz they are not used to playing skillfully.

So fromhell, please for our sakes don't let this non issue get you upset you are doing a great job stay true to yourself and thank you for a splendid game.

A special server or two for hardcore would probably help the situation they could do what the fcuk they like together lol. But seriously most of us see the game way you intended and have no desire to dick with it.

There aint no point in winning if ur a lame ass cheater an u all know that!
Logged

~~((-_-))~~ Coz Its a bittersweet symphony ~~((-_-))~~
andrewj
Member


Cakes 24
Posts: 581



« Reply #76 on: June 06, 2008, 08:48:09 AM »

i couldn't stand to look at it with vertex and all the settings to make it easier to see in the shadows etc ugh.
So kream you never use RAZ3R's brightskin pack? Not even on CE1M7 where your player model suddenly becomes Grism instead the girl you normally use? :-)
Logged
fromhell
Administrator
GET A LIFE!
**********

Cakes 35
Posts: 14519



WWW
« Reply #77 on: June 06, 2008, 11:19:39 AM »

Well all this 'professional' visual aids are no different than the rather illicit use in performance-enhancing drugs in sports (in REAL LIFE)

Winners don't do drugs
Logged

asking when OA3 will be done won't get OA3 done.
Progress of OA3 currently occurs behind closed doors alone

I do not provide technical support either.

new code development on github
jackoverfull
Member


Cakes 14
Posts: 385


Member


WWW
« Reply #78 on: June 06, 2008, 11:20:10 AM »

i couldn't stand to look at it with vertex and all the settings to make it easier to see in the shadows etc ugh.
So kream you never use RAZ3R's brightskin pack? Not even on CE1M7 where your player model suddenly becomes Grism instead the girl you normally use? :-)

the question was not directed to me, but i haven't. and i will not.
Logged
kr3am
Nub


Cakes 1
Posts: 48


Kitty at work


« Reply #79 on: June 06, 2008, 11:37:08 AM »

No Andrew i don't use it and i have no idea why my model becomes Grism instead of Sly, i never even knew that.
Maybe you should talk to me more often than u do lol.
im new to this, some maps are too dark admittedly but i just think, hell the next map will be different OA_DM3 is way too dark for me, so some sort of compromise should be arranged i guess.
i feel that setting the game up to suit a persons eyesight, monitor, etc is a valid thing and should not be considered a cheat but i have to fall in with fromhell here when taken to the extreme it is detrimental to the game.
Actually im quite lazy when it comes to config stuff and i just drag the brightness slider up a bit to compensate coz i really just want to get in there and frag about, so watch out im coming after u with that saw that i owe u buddy, lol.. ;-)
Logged

~~((-_-))~~ Coz Its a bittersweet symphony ~~((-_-))~~
fromhell
Administrator
GET A LIFE!
**********

Cakes 35
Posts: 14519



WWW
« Reply #80 on: June 06, 2008, 11:52:09 AM »

Chances are the server doesn't have Sly and runs 0.7.6 / 0.7.7. Pure servers would Grism you up

Sly would make a return if someone really did some good looking polish on the skin. Until then, she's retired.
Logged

asking when OA3 will be done won't get OA3 done.
Progress of OA3 currently occurs behind closed doors alone

I do not provide technical support either.

new code development on github
andrewj
Member


Cakes 24
Posts: 581



« Reply #81 on: June 07, 2008, 02:49:17 AM »

No Andrew i don't use it and i have no idea why my model becomes Grism instead of Sly, i never even knew that.
Ah ok, I apologise for jumping to the wrong conclusion.

Quote
so watch out im coming after u with that saw that i owe u buddy, lol.. ;-)
Come on you've pummelled me heaps of times now, you don't need to be sour kream Smiley
Logged
CFQ
---
Member


Cakes 0
Posts: 173


---


« Reply #82 on: June 07, 2008, 06:02:58 AM »

----
« Last Edit: August 13, 2008, 06:42:39 PM by CFQ » Logged

I blanked my posts because I don't like this forum anymore! BAWWWWW
sago007
Posts a lot
*

Cakes 62
Posts: 1664


Open Arena Developer


WWW
« Reply #83 on: June 07, 2008, 06:33:58 AM »

Damn , if u have a DSL connection you cannot set cl_maxpackets to 30 , 'cause u warp and other players have difficult to hit you.

That can be considered an accusation. Do you have any proof of this?

cl_maxpackets 30 is the default value and should be enough. Unless the server has a high sv_fps enabling cg_smoothclients should prevent warping.

Not that I am that experienced in Quake network technology. I use http://ucguides.savagehelp.com/Quake3/connection.html as source.

EDIT:
0.7.6+ does no longer have cg_smoothclients because the unlagged code has an improved way that can compensate for higher packet loss with far better prediction. It might still be relevant in mods.
« Last Edit: June 07, 2008, 06:42:20 AM by sago007 » Logged

There are nothing offending in my posts.
CFQ
---
Member


Cakes 0
Posts: 173


---


« Reply #84 on: June 07, 2008, 11:32:11 AM »

---
« Last Edit: August 13, 2008, 06:42:58 PM by CFQ » Logged

I blanked my posts because I don't like this forum anymore! BAWWWWW
Mr. Oho
Half-Nub


Cakes 0
Posts: 55

I will press the button!


« Reply #85 on: June 07, 2008, 10:57:43 PM »

This is an age old discussion.

Could you possibly post a demo of this? Maybe with /players output so its possible to see what rate/packet settings these guys are at? Some scoreboard to get a feeling for their overall pings wouldnt hurt either Wink
« Last Edit: June 07, 2008, 11:07:47 PM by Mr. Oho » Logged
CFQ
---
Member


Cakes 0
Posts: 173


---


« Reply #86 on: June 08, 2008, 12:11:16 PM »

----
« Last Edit: August 13, 2008, 06:43:16 PM by CFQ » Logged

I blanked my posts because I don't like this forum anymore! BAWWWWW
fufinha
stop making alt accounts and self-termination
Member


Cakes 7
Posts: 584


retired


« Reply #87 on: June 08, 2008, 02:55:49 PM »

Damn , if u have a DSL connection you cannot set cl_maxpackets to 30 , 'cause u warp and other players have difficult to hit you.

That can be considered an accusation. Do you have any proof of this?

It's fact. 30 is recommended for modems. rate at 25000? Well that definately does not make any sense. He probably loses  an extra bit of ping and believes less ping = better. If you want a faster network performance set rate at 15000  and set packets at 60. If you ping increases a little and your rate is stable you might find it better if you forget about what your ping is. You send the server more info to calculate your position more accurately then it also applies to the position of the enemy. This is more an issue with rail players as you need accurate hits. I force that setting to a minimum of 60. A very good player will be able to tell if your using 30 max_packets.
Logged
kr3am
Nub


Cakes 1
Posts: 48


Kitty at work


« Reply #88 on: June 09, 2008, 01:55:41 AM »

Ok guys, now i am really confused.
What is this warping thing, i have never seen anyone warping.
Pls could someone tell me what to set /cl_maxpackets ?
i have a crap ping all the time and even on the Australian server which is closest it is rarely below 90. Sad

Andrew, you are forgiven Smiley

Chances are the server doesn't have Sly and runs 0.7.6 / 0.7.7. Pure servers would Grism you up
Sly would make a return if someone really did some good looking polish on the skin. Until then, she's retired.
How would i go about doing that, as i have a term break coming up and could do it then.
Logged

~~((-_-))~~ Coz Its a bittersweet symphony ~~((-_-))~~
Mr. Oho
Half-Nub


Cakes 0
Posts: 55

I will press the button!


« Reply #89 on: June 09, 2008, 07:18:24 AM »

Ok guys, now i am really confused.

You arent the only one. In short cl_maxpackets seems (at least from my _short_ look over the source so take what i say with a grain of salt Tongue) to control the number of input updates packets per second  the client sends to the server. Low values save a bit of network traffic by stuffing more updates into 1 single packet and thereby saving overhead of additional packet headers. So the real questions (that im far to lazy to investigate... someone else wanna waste time? Tongue) is how will the server handle a couple of input updates arriving at once.

I have experimented a bit with it in ET (yeah that might not be representative for Quake3 because of maybe different antilag code) and was unable get any decent jerkyness. At least thats what the guys watching me said. Anyways theres a server setting in ETPro that limits the maximum way a player can warp so there has to be a problem with this somewhere. Im pretty sure bani would implement this if it was nonsense. If its related to netsettings i dont know... Thats why i was kinda interested in seeing it in a demo.
Logged
CFQ
---
Member


Cakes 0
Posts: 173


---


« Reply #90 on: June 09, 2008, 09:07:04 AM »

----
« Last Edit: August 13, 2008, 06:58:28 PM by CFQ » Logged

I blanked my posts because I don't like this forum anymore! BAWWWWW
pulchr
Member


Cakes 34
Posts: 626



WWW
« Reply #91 on: June 09, 2008, 09:41:24 AM »

when rtcw was new there was talk about players using 'warp scripts'. by changing different network settings by keybinds they supposedly got very difficult to hit, when they wanted to. i'm really not sure if it worked or not. after some patches the network settings required a "/restart" to activate new settings, this must've made 'warp scripts' pretty useless as you're basically a sitting duck while your computer restarts the level with the new settings.

note that i'm not entirely sure if that was the reason for the need to restart or if it actually worked, but i know there was some talk about these scripts.
Logged
CFQ
---
Member


Cakes 0
Posts: 173


---


« Reply #92 on: June 09, 2008, 10:15:31 AM »

-----
« Last Edit: August 13, 2008, 07:42:40 PM by CFQ » Logged

I blanked my posts because I don't like this forum anymore! BAWWWWW
Mr. Oho
Half-Nub


Cakes 0
Posts: 55

I will press the button!


« Reply #93 on: June 09, 2008, 10:23:10 AM »

First of all , fix these 2 points :

Even if was in charge of stuff here (i am NOT an OpenArena developer) i couldnt fix something i cant reproduce. Im simply trying to get a clear picture.

The problems is NOT how many packets the client send to server , but how it has influence in the server and in the other clients to recalculate NEW PLAYER POSITION WITH LESS DATA . With less data server is not able to calculate accurate position . Older DELTAS on the server and on all other clients become invalid in a while and client's prediction code has to work harder than before at least TO TRY predicting new position. But positions of other players are still valid in the client of the player that warps.

From my understanding the server takes the recieved input (i would assume it will just take last input recieved) and generates sv_fps frames from it. This is what gets send to to clients then. So maybe antilag code (im totaly ignorant here sry) gets confused by unusual movement patterns caused by using out of date input on the server? :S

sv_ lanforcePackets "1" maybe could be useful but take a look at qcommon.h :

<code>
...

#define   PACKET_BACKUP   32   // number of old messages that must be kept on client and
                     // server for delta comrpession and ping estimation
#define   PACKET_MASK      (PACKET_BACKUP-1)

#define   MAX_PACKET_USERCMDS      32      // max number of usercmd_t in a packet

#define   PORT_ANY         -1

#define   MAX_RELIABLE_COMMANDS   64         // max string commands buffered for restransmit

...
</code>

Im sorry but i dont get what that should tell me. If i was to decide on it id go for a limit on warp range as this seems to be a tested cure - whatever its caused by in the end.

A demo is not useful 'cause it's executed in a complete different context ( not live )

To my best knowledge a demo is a dump of network traffic. So id think it would pretty much reproduce the same enviroment.

when rtcw was new there was talk about players using 'warp scripts'. by changing different network settings by keybinds they supposedly got very difficult to hit, when they wanted to. i'm really not sure if it worked or not. after some patches the network settings required a "/restart" to activate new settings, this must've made 'warp scripts' pretty useless as you're basically a sitting duck while your computer restarts the level with the new settings.

note that i'm not entirely sure if that was the reason for the need to restart or if it actually worked, but i know there was some talk about these scripts.

Yeah thats why i said its an age old discussion. I dont deny that warping players are a problem (if bani thinks so i wont argue with it Tongue). I just havent realy noticed it and wasnt able to reproduce it by playing with client settings (at least on ET) so it took its place in the ranks of "high pingers cause server lag" and "timenudge has effects that are noticeable for other players" for me Wink
Logged
Mr. Oho
Half-Nub


Cakes 0
Posts: 55

I will press the button!


« Reply #94 on: June 09, 2008, 10:30:09 AM »

If I try to set cl_maxpackets from 100 to 30 , I clearly can see lower tx data rate in the monitor

Unless it ignores packet headers its not that suprising as maxpackets is designed to save overhead data. So 70 less packets = 560bytes less overhead.

Edit: Yeah i know i know should have put this in the last post but i was to late when i realized. Sorry :/

Btw: I totaly agree those who can afford it should use a high setting. It can only benefit them. I just dont feel the default is such a bad thing.
« Last Edit: June 09, 2008, 10:34:43 AM by Mr. Oho » Logged
fufinha
stop making alt accounts and self-termination
Member


Cakes 7
Posts: 584


retired


« Reply #95 on: June 09, 2008, 01:04:30 PM »


i have a crap ping all the time and even on the Australian server which is closest it is rarely below 90. Sad

Andrew, you are forgiven Smiley

Chances are the server doesn't have Sly and runs 0.7.6 / 0.7.7. Pure servers would Grism you up
Sly would make a return if someone really did some good looking polish on the skin. Until then, she's retired.
How would i go about doing that, as i have a term break coming up and could do it then.

heheh I'm sorry I didn't know you were in Australia Smiley Those settings might not be so bad after all if your playing on Europe servers. You are right about one thing. Vertex is a non issue. I play for fun, all the best players I've known use lightmap. It is the only way you can adjust your levels perfectly to suit, and also because you *can* better performance if you decide to pass on the eye candy.  Players will use floor shadows as guides and learn where where people spawn and make a long jump to a ledge. Anyone can adjust their screen to many different variants with lightmap. If you were to dig out configs of good players you see how much they differ, some look horrible but it obviously suits them. The ONLY reason in my own opinion people would use vertex is simply for performance and then they can increase detail etc because it's not using the cpu resource of the effects.

With rate/maxpackets etc, it's pretty much the same. It can be very complex when playing with delag and playing on servers with pmove_accurate (noghost cpma osp etc). There are rounding calculations and if you feed the server with less packets it will do the work for you. But if you actually play very accurately then you will get better timing and able to move around more flexable. There are also different variations of delag and typically the more you increase the delag, the more restrictive your movement will be. delag version2 works very well with pmove_accurate and I can easily play people in the US who have 250 - 300 ping and not notice any difference. Everything is fair. With rail games I tend to put splash damage up a little on the rail gun so that anyone using low maxpackets wont get any advantage.

So in your case, if you have a stable rate / route then I would try rate at 10000 and cl_maxpackets at 60. I have my snaps at 40 and a I prefer to have a server with sv_fps 30. Most servers have 20 fps so I guess your snaps is just right.

I agree, I do not think it matters so much and if it bothers admins they can reduce the maxrate or increase the packets.

It might have been an issue in the past but not so much now. I do remember one player telling some other guy to increase packets from 30. It's the first time I've ever heard anyone complain about it. His rate was 25000 and was a good player but he good reason because his isp restricted his upstream far more than what he could recieve.

Just to complicate thing further, alot of times you may not be playing with the settings you think you are. I remember making most cvars unrestrictive so that people with high performance machines and network to take advantage of it (I personally think it makes little diff) and then people complained about performance issues. Obviously because their setttings were too high and they previously used the servers defaults. I just put the setttings back, they had slightly less ping, and they were happy Smiley I also don't see the point of having your rate at 25000 when the server might have a max value of 10000. If your network can easily cope with it then I guess it no loss, but it certainly affects me.
 



Logged
kr3am
Nub


Cakes 1
Posts: 48


Kitty at work


« Reply #96 on: June 10, 2008, 05:01:43 AM »

missbehaving, big thanks for clarifying this. Smiley
Logged

~~((-_-))~~ Coz Its a bittersweet symphony ~~((-_-))~~
CFQ
---
Member


Cakes 0
Posts: 173


---


« Reply #97 on: June 10, 2008, 08:04:02 PM »

----
« Last Edit: August 13, 2008, 08:13:49 PM by CFQ » Logged

I blanked my posts because I don't like this forum anymore! BAWWWWW
Mr. Oho
Half-Nub


Cakes 0
Posts: 55

I will press the button!


« Reply #98 on: June 11, 2008, 10:31:11 AM »

So I could open a demo with wireshark ( a network packet inspector ) and read it ? ..... ROFTL

I think a demo is a commands dump and when anyone execute it , it is completely executed with no miss data , so some effect probably _ARE NOT_ the same as seen in a live game Wink

ok
Logged
fufinha
stop making alt accounts and self-termination
Member


Cakes 7
Posts: 584


retired


« Reply #99 on: June 11, 2008, 04:37:00 PM »

Something I noticed last night, I had fps at 30, must have set it by mistake thinking it was the server.  I didn't really notice it at first but when I looked at my ping it was at an average of 25 it's usually about 60. If players have bad ping then they might be better using 85 which is default.  I know optimum momvement is achieved at 125 but the difference will hardly have any affect in game.  As long as it's stable, it's fine.  You'll certainly be able to move around much better than fps which drops and changes often. This is because the calculations of movement will be wrong and you'll just slow down instantly. I'm not sure but if you had 250 fps and that drops to 220 and back up again is far worse than a stable 85. I think the 85 fps player will be running around faster than the other player

Also, you usually have to use g_sync* 1 clients to record a demo. You notice the lag since your synced to the actual server  itself. I wonder if this would be a good way to test your network settings and compare them as your more likely to feel how the server responds to what data your sending. Anyone tried this?

Logged
Pages: 1 2 3 [4] 5
  Print  
 
Jump to: