Current siege version

[][X][~FLuKE~][X][]

Minigun = Rapid fire ZP sniper rifle - needs to have the shit nerfed out of it

lag compensator is still doing the opposite of what it should be doing.

50-70% Shots still not registering from sniper, in fact I think it is worse since the last update.

Server crashing.

Seems it gets gradually worse each release.


I am curious as to why the server never gets reverted back to the best working version until the problems have been smoothed out, forcing players put up with problems until the next release is wrong and frustrating for them and is exactly what has happened over recent weeks, loyal traffic should come first, mod development second and keep them as happy as you can with solid versions, not ones riddled with issues or game changing changes.

Higor, from one developer to another.....you should stop forcing people to use bad versions/Releases, you will lose support rapidly and without support you are basically lining yourself up to be flamed and blamed. You are the developer so you have the control, do you really want people to think your coding skills are poor? because that is what is happening because you are forcing them to use broken versions. Not insulting you at all, I am giving you a reality check from personal experience of what happens if you keep releasing buggy versions with many faults and things added/removed without checking what the people would prefer.

Not liking additions/removals/changes is a normal process of development, crashing, bugs and serious game inhibiting problems is not. The development of siege is being portrayed very badly and seems to be getting worse, soon all you will end up doing for months is fixing bugs/problems along side losing the support of the siege community.

this feels like deja vu.

[|uK|kenneth]

1 guy in the server when i joined today gj.

[SAM]

So. Siege ultimate 21 then?

[Sheepy]

why doesn't siege use newnet?

[|uK|B|aZe//.]

bring back zeroping

zp211 like fluke said which fixes the double shot with sniper problem. also bring back the siege version first used in the last mlsg season that just finished before the season even started

was either v3 or v4 im not sure now cause theres so many versions

[|uK|fleecey]

Chamberly why do you dislike the post? you think the siege version that is on right now is good?

[Higor]

I leave for 1 day...
And none of the 'small' changes was supposed to even affect any critical server components, besides fixing imagedropping in a safer and less expensive way.
I really doubt the Siege package is doing this but I'll profile the server to see if I can reproduce this.

EDIT: I see 23 players and 8 spectators.
Kenneth, would you kindly STFU?

--- Updated ---

SAM

r=65
cli=28
act=2.8~4.0
net=16~23

The server's CPU can't handle the replication code, something is horribly hogging the Server's resources.

--- Updated ---

SAM

I will reduce tickrate from 65 to 50 until I can find out what's going on.

--- Updated ---

r=50
cli=28
act=2.8~4.0
net=12~19

Tickrate 50 profile, 300 net channels open which is excessive.
I believe we've got lots of bAlwaysRelevant actors with a NetUpdateFrequency of 100 getting replicated to all players.
0011 isn't creating any new actors, I will take a look at this NexGen later.
If it's the NUF value, it could be tweaked without replacing the package.

--- Updated ---

Had my profilers print a client actor list:
Actors 0-119 are the map's static actors (Command 4 way).
Actor 120 is the local player pawn.
Actor 121 and on are local and replicated actors.
Saved move actors are data placeholders created on the client.

Everything else (except items and weapons) will be subject to analysis.
EDIT: Filtered list.


ScriptLog: Actor 120: bbCHSpectator0
ScriptLog: Actor 122: NexgenClient7
ScriptLog: Actor 123: GeneralConfig6
ScriptLog: Actor 124: ServerConfig6
ScriptLog: Actor 125: NexgenLang6
ScriptLog: Actor 126: NexgenHUD6
ScriptLog: Actor 127: TMale11
ScriptLog: Actor 128: PlayerShadow147
ScriptLog: Actor 129: TMale47
ScriptLog: Actor 130: PlayerShadow148
ScriptLog: Actor 131: TMale12
ScriptLog: Actor 132: PlayerShadow149
ScriptLog: Actor 133: TFemale48
ScriptLog: Actor 134: PlayerShadow150
ScriptLog: Actor 135: TMale48
ScriptLog: Actor 136: PlayerShadow151
ScriptLog: Actor 137: TFemale40
ScriptLog: Actor 138: PlayerShadow152
ScriptLog: Actor 139: TFemale41
ScriptLog: Actor 140: PlayerShadow153
ScriptLog: Actor 141: TFemale49
ScriptLog: Actor 142: PlayerShadow154
ScriptLog: Actor 143: TMale49
ScriptLog: Actor 144: PlayerShadow155
ScriptLog: Actor 145: TMale50
ScriptLog: Actor 146: PlayerShadow156
ScriptLog: Actor 147: TFemale50
ScriptLog: Actor 148: PlayerShadow157
ScriptLog: Actor 149: TMale51
ScriptLog: Actor 150: PlayerShadow158
ScriptLog: Actor 151: TFemale51
ScriptLog: Actor 152: PlayerShadow160
ScriptLog: Actor 153: TFemale42
ScriptLog: Actor 154: PlayerShadow162
ScriptLog: Actor 155: TMale52
ScriptLog: Actor 156: PlayerShadow163
ScriptLog: Actor 157: TMale13
ScriptLog: Actor 158: PlayerShadow164
ScriptLog: Actor 159: TFemale52
ScriptLog: Actor 160: PlayerShadow165
ScriptLog: Actor 161: TFemale43
ScriptLog: Actor 162: PlayerShadow166
ScriptLog: Actor 163: TFemale44
ScriptLog: Actor 164: PlayerShadow167
ScriptLog: Actor 165: sgBaseCore0
ScriptLog: Actor 166: WildcardsMeshFX0
ScriptLog: Actor 167: sgBaseCore1
ScriptLog: Actor 168: WildcardsMeshFX1
ScriptLog: Actor 169: SAClient1
ScriptLog: Actor 170: sgPlayerData1
ScriptLog: Actor 171: sgPlayerData3
ScriptLog: Actor 172: SoundsHUDNotify0
ScriptLog: Actor 173: sgPlayerData4
ScriptLog: Actor 174: sgPlayerData5
ScriptLog: Actor 175: sgPlayerData6
ScriptLog: Actor 176: sgPlayerData7
ScriptLog: Actor 177: sgSpecInv0
ScriptLog: Actor 178: sgPlayerData8
ScriptLog: Actor 179: sgPlayerData9
ScriptLog: Actor 180: sgPlayerData10
ScriptLog: Actor 181: sgPlayerData11
ScriptLog: Actor 182: sgPlayerData12
ScriptLog: Actor 183: sgPlayerData13
ScriptLog: Actor 184: sgPlayerData14
ScriptLog: Actor 185: sgPlayerData15
ScriptLog: Actor 186: sa0
ScriptLog: Actor 187: sgPlayerData16
ScriptLog: Actor 188: sgPlayerData17
ScriptLog: Actor 189: sgPlayerData18
ScriptLog: Actor 190: sgPlayerData19
ScriptLog: Actor 191: sgPlayerData20
ScriptLog: Actor 192: sgGameReplicationInfo0
ScriptLog: Actor 193: NexgenABMClient1
ScriptLog: Actor 196: sgBaseCore2
ScriptLog: Actor 197: WildcardsMeshFX9
ScriptLog: Actor 204: NXPClient1
ScriptLog: Actor 205: NXPLang1
ScriptLog: Actor 206: TeamInfo24
ScriptLog: Actor 210: NexgenClientCore7
ScriptLog: Actor 215: ConsoleChecker0
ScriptLog: Actor 219: sgPRI0
ScriptLog: Actor 220: sgClient0
ScriptLog: Actor 221: sgPRI1
ScriptLog: Actor 222: SavedMove303
ScriptLog: Actor 223: FV_ImpactMacroActor13
ScriptLog: Actor 224: sgHUD0
ScriptLog: Actor 225: FontInfo24
ScriptLog: Actor 226: ServerInfoTeam6
ScriptLog: Actor 227: FontInfo25
ScriptLog: Actor 228: SoundsHUDHelper0
ScriptLog: Actor 229: sgPRI2
ScriptLog: Actor 232: sgPRI3
ScriptLog: Actor 233: PlayerReplicationInfo14
ScriptLog: Actor 257: TeamInfo25
ScriptLog: Actor 268: SAReplicationInfo1
ScriptLog: Actor 272: NexgenConfigExt6
ScriptLog: Actor 274: TeamInfo26
ScriptLog: Actor 287: NexgenGameInfo6
ScriptLog: Actor 292: TeamInfo27
ScriptLog: Actor 356: sgPRI5
ScriptLog: Actor 362: sgPRI8
ScriptLog: Actor 363: sgPRI9
ScriptLog: Actor 364: sgPRI10
ScriptLog: Actor 365: sgPRI11
ScriptLog: Actor 368: sgPRI12
ScriptLog: Actor 371: sgPRI13
ScriptLog: Actor 372: sgPRI14
ScriptLog: Actor 373: sgPRI15
ScriptLog: Actor 374: PlayerReplicationInfo15
ScriptLog: Actor 379: sgPRI16
ScriptLog: Actor 380: sgPRI17
ScriptLog: Actor 381: sgPRI18
ScriptLog: Actor 382: sgPRI20
ScriptLog: Actor 383: sgPRI21
ScriptLog: Actor 388: sgPRI23
ScriptLog: Actor 390: sgPRI24
ScriptLog: Actor 391: sgPRI25
ScriptLog: Actor 392: sgPRI26
ScriptLog: Actor 393: NexgenSharedDataSyncManager8
ScriptLog: Actor 394: NXPHUD1
ScriptLog: Actor 395: NexgenSharedDataSyncManager9
ScriptLog: Actor 396: NexgenPlayerInfo120
ScriptLog: Actor 397: NexgenPlayerInfo121
ScriptLog: Actor 398: NexgenPlayerInfo122
ScriptLog: Actor 399: NexgenPlayerInfo123
ScriptLog: Actor 400: NexgenPlayerInfo124
ScriptLog: Actor 401: NexgenPlayerInfo126
ScriptLog: Actor 402: NexgenPlayerInfo127
ScriptLog: Actor 403: NexgenPlayerInfo128
ScriptLog: Actor 404: NexgenPlayerInfo129
ScriptLog: Actor 405: NexgenPlayerInfo130
ScriptLog: Actor 406: NexgenPlayerInfo131
ScriptLog: Actor 407: NexgenPlayerInfo132
ScriptLog: Actor 408: NexgenPlayerInfo135
ScriptLog: Actor 409: NexgenPlayerInfo136
ScriptLog: Actor 410: NexgenPlayerInfo137
ScriptLog: Actor 411: NexgenPlayerInfo138
ScriptLog: Actor 412: NexgenPlayerInfo139
ScriptLog: Actor 413: NexgenPlayerInfo140
ScriptLog: Actor 414: NexgenPlayerInfo141
ScriptLog: Actor 415: NexgenPlayerInfo142
ScriptLog: Actor 416: NexgenPlayerInfo143
ScriptLog: Actor 417: NexgenPlayerInfo145
ScriptLog: Actor 418: SALang1
ScriptLog: Actor 419: NXPConfigDC1
ScriptLog: Actor 420: NexgenABMConfigDC1
ScriptLog: Actor 421: NexgenABMBanListDC1
ScriptLog: Actor 422: NXPMapListDC1
ScriptLog: Actor 423: sgScore0
ScriptLog: Actor 424: FontInfo26
ScriptLog: Actor 428: sgPlayerData22
ScriptLog: Actor 429: sgPRI30
ScriptLog: Actor 431: TMale53
ScriptLog: Actor 432: PlayerShadow169
ScriptLog: Actor 433: NexgenPlayerInfo150
ScriptLog: Actor 435: sgPRI31
ScriptLog: Actor 437: NexgenPlayerInfo151
ScriptLog: Actor 438: TMale55
ScriptLog: Actor 439: PlayerShadow171
ScriptLog: Actor 440: sgPlayerData24
ScriptLog: Actor 442: sgPRI33
ScriptLog: Actor 443: NexgenPlayerInfo152
ScriptLog: Actor 444: sgPRI34
ScriptLog: Actor 447: NexgenPlayerInfo153
ScriptLog: Actor 541: sgPlayerData26
ScriptLog: Actor 543: TMale56
ScriptLog: Actor 544: PlayerShadow173
ScriptLog: Actor 545: sgPRI36
ScriptLog: Actor 546: NexgenPlayerInfo154
ScriptLog: Actor 547: zp_SniperRifle160
ScriptLog: Actor 548: sgBuildingCH0
ScriptLog: Actor 549: UT_Eightball552
ScriptLog: Actor 552: zp_SniperRifle168
ScriptLog: Actor 553: UT_Eightball558
ScriptLog: Actor 554: UT_Eightball562
ScriptLog: Actor 555: sgConstructor149
ScriptLog: Actor 556: zp_SniperRifle172
ScriptLog: Actor 557: sgEnforcer243
ScriptLog: Actor 558: sgEnforcer245
ScriptLog: Actor 559: UT_Eightball565
ScriptLog: Actor 560: ut_biorifle124
ScriptLog: Actor 561: zp_SniperRifle173
ScriptLog: Actor 562: sgEnforcer248
ScriptLog: Actor 563: UT_Eightball567
ScriptLog: Actor 564: UT_Eightball568
ScriptLog: Actor 565: zp_SniperRifle176
ScriptLog: Actor 566: zp_ShockRifle116
ScriptLog: Actor 567: sgMinigun77
ScriptLog: Actor 568: sgNukeLauncher23
ScriptLog: Actor 569: sgEnforcer249
ScriptLog: Actor 570: UT_FlakCannon157
ScriptLog: Actor 571: zp_SniperRifle177
ScriptLog: Actor 572: zp_ShockRifle117
ScriptLog: Actor 573: zp_SniperRifle178
ScriptLog: Actor 574: sgMinigun78
ScriptLog: Actor 575: zp_ShockRifle118
ScriptLog: Actor 576: ripper103
ScriptLog: Actor 577: UT_FlakCannon158
ScriptLog: Actor 578: zp_SniperRifle179
ScriptLog: Actor 579: sgMinigun79
ScriptLog: Actor 580: sgEnforcer250
ScriptLog: Actor 581: UT_Eightball571
ScriptLog: Actor 582: sgPulseGun40

[|uK|kenneth]

dude why the fuck are you raging its true. there was 1 guy in the server when i joined today.

[Higor]

118 NPLoader actors being constantly polled for variable changes, for 28 clients on every frame?
SAM, I could use a server changelog.


ScriptLog: Actor 121: NPLReplicationInfo611
ScriptLog: Actor 194: NPLReplicationInfo612
ScriptLog: Actor 195: NPLReplicationInfo613
ScriptLog: Actor 198: NPLReplicationInfo614
ScriptLog: Actor 199: NPLReplicationInfo615
ScriptLog: Actor 200: NPLReplicationInfo616
ScriptLog: Actor 201: NPLReplicationInfo617
ScriptLog: Actor 202: NPLReplicationInfo618
ScriptLog: Actor 203: NPLReplicationInfo619
ScriptLog: Actor 207: NPLReplicationInfo620
ScriptLog: Actor 208: NPLReplicationInfo621
ScriptLog: Actor 209: NPLReplicationInfo622
ScriptLog: Actor 211: NPLReplicationInfo623
ScriptLog: Actor 212: NPLReplicationInfo624
ScriptLog: Actor 213: NPLReplicationInfo625
ScriptLog: Actor 214: NPLReplicationInfo626
ScriptLog: Actor 216: NPLReplicationInfo627
ScriptLog: Actor 217: NPLReplicationInfo628
ScriptLog: Actor 218: NPLReplicationInfo629
ScriptLog: Actor 230: NPLReplicationInfo630
ScriptLog: Actor 231: NPLReplicationInfo631
ScriptLog: Actor 234: NPLReplicationInfo632
ScriptLog: Actor 236: NPLReplicationInfo633
ScriptLog: Actor 237: NPLReplicationInfo634
ScriptLog: Actor 240: NPLReplicationInfo635
ScriptLog: Actor 241: NPLReplicationInfo636
ScriptLog: Actor 245: NPLReplicationInfo637
ScriptLog: Actor 246: NPLReplicationInfo638
ScriptLog: Actor 247: NPLReplicationInfo639
ScriptLog: Actor 248: NPLReplicationInfo640
ScriptLog: Actor 249: NPLReplicationInfo641
ScriptLog: Actor 250: NPLReplicationInfo642
ScriptLog: Actor 251: NPLReplicationInfo643
ScriptLog: Actor 252: NPLReplicationInfo644
ScriptLog: Actor 253: NPLReplicationInfo645
ScriptLog: Actor 254: NPLReplicationInfo646
ScriptLog: Actor 255: NPLReplicationInfo647
ScriptLog: Actor 256: NPLReplicationInfo648
ScriptLog: Actor 258: NPLReplicationInfo649
ScriptLog: Actor 259: NPLReplicationInfo650
ScriptLog: Actor 260: NPLReplicationInfo651
ScriptLog: Actor 261: NPLReplicationInfo652
ScriptLog: Actor 262: NPLReplicationInfo653
ScriptLog: Actor 263: NPLReplicationInfo654
ScriptLog: Actor 264: NPLReplicationInfo655
ScriptLog: Actor 265: NPLReplicationInfo656
ScriptLog: Actor 266: NPLReplicationInfo657
ScriptLog: Actor 267: NPLReplicationInfo658
ScriptLog: Actor 269: NPLReplicationInfo659
ScriptLog: Actor 270: NPLReplicationInfo660
ScriptLog: Actor 271: NPLReplicationInfo661
ScriptLog: Actor 273: NPLReplicationInfo662
ScriptLog: Actor 275: NPLReplicationInfo663
ScriptLog: Actor 276: NPLReplicationInfo664
ScriptLog: Actor 277: NPLReplicationInfo665
ScriptLog: Actor 278: NPLReplicationInfo666
ScriptLog: Actor 279: NPLReplicationInfo667
ScriptLog: Actor 280: NPLReplicationInfo668
ScriptLog: Actor 281: NPLReplicationInfo669
ScriptLog: Actor 282: NPLReplicationInfo670
ScriptLog: Actor 283: NPLReplicationInfo671
ScriptLog: Actor 284: NPLReplicationInfo672
ScriptLog: Actor 285: NPLReplicationInfo673
ScriptLog: Actor 286: NPLReplicationInfo674
ScriptLog: Actor 288: NPLReplicationInfo675
ScriptLog: Actor 289: NPLReplicationInfo676
ScriptLog: Actor 290: NPLReplicationInfo677
ScriptLog: Actor 291: NPLReplicationInfo678
ScriptLog: Actor 293: NPLReplicationInfo679
ScriptLog: Actor 294: NPLReplicationInfo680
ScriptLog: Actor 295: NPLReplicationInfo681
ScriptLog: Actor 296: NPLReplicationInfo682
ScriptLog: Actor 297: NPLReplicationInfo683
ScriptLog: Actor 298: NPLReplicationInfo684
ScriptLog: Actor 312: NPLReplicationInfo685
ScriptLog: Actor 313: NPLReplicationInfo686
ScriptLog: Actor 314: NPLReplicationInfo687
ScriptLog: Actor 315: NPLReplicationInfo688
ScriptLog: Actor 316: NPLReplicationInfo689
ScriptLog: Actor 317: NPLReplicationInfo690
ScriptLog: Actor 318: NPLReplicationInfo691
ScriptLog: Actor 319: NPLReplicationInfo692
ScriptLog: Actor 320: NPLReplicationInfo693
ScriptLog: Actor 321: NPLReplicationInfo694
ScriptLog: Actor 322: NPLReplicationInfo695
ScriptLog: Actor 323: NPLReplicationInfo696
ScriptLog: Actor 324: NPLReplicationInfo697
ScriptLog: Actor 325: NPLReplicationInfo698
ScriptLog: Actor 326: NPLReplicationInfo699
ScriptLog: Actor 330: NPLReplicationInfo700
ScriptLog: Actor 357: NPLReplicationInfo701
ScriptLog: Actor 358: NPLReplicationInfo702
ScriptLog: Actor 359: NPLReplicationInfo703
ScriptLog: Actor 360: NPLReplicationInfo704
ScriptLog: Actor 361: NPLReplicationInfo705
ScriptLog: Actor 366: NPLReplicationInfo706
ScriptLog: Actor 367: NPLReplicationInfo707
ScriptLog: Actor 369: NPLReplicationInfo708
ScriptLog: Actor 370: NPLReplicationInfo709
ScriptLog: Actor 375: NPLReplicationInfo710
ScriptLog: Actor 376: NPLReplicationInfo711
ScriptLog: Actor 377: NPLReplicationInfo712
ScriptLog: Actor 378: NPLReplicationInfo713
ScriptLog: Actor 384: NPLReplicationInfo714
ScriptLog: Actor 385: NPLReplicationInfo715
ScriptLog: Actor 386: NPLReplicationInfo716
ScriptLog: Actor 387: NPLReplicationInfo717
ScriptLog: Actor 389: NPLReplicationInfo718
ScriptLog: Actor 425: NPLReplicationInfo719
ScriptLog: Actor 426: NPLReplicationInfo720
ScriptLog: Actor 427: NPLReplicationInfo721
ScriptLog: Actor 430: NPLReplicationInfo722
ScriptLog: Actor 434: NPLReplicationInfo723
ScriptLog: Actor 436: NPLReplicationInfo724
ScriptLog: Actor 441: NPLReplicationInfo725
ScriptLog: Actor 445: NPLReplicationInfo726
ScriptLog: Actor 446: NPLReplicationInfo727
ScriptLog: Actor 542: NPLReplicationInfo728

[TDNConvict]

agreed server has been aweful for a long time now.. if you put last mlsg seasons version on all the binds will be fucked up again but at least it will be working better.... I still don't understand why you don't just go off the pug version....

[Higor]

NPL actors discarded.
It's something else.

EDIT:
I'll take a look at NexGen.

EDIT 2:
IGUK server has 20 NPL actors, and Siege has 150.
I don't know, something doesn't seem right about this.

EDIT 3:
NexGen actors aren't resource hungry at all.
I will try altering the player pawns update rates for very short amounts of time and observe any changes in the CPU usage.

[Higor]

Map: BlackRiver2a, 5 minutes into the game.

r=50
cli=19
act=1.1~1.8
net=4.0~8.5

15 players, 4 spectators.

[|uK|fleecey]

Map: BlackRiver2a, 5 minutes into the game.

r=50
cli=19
act=1.1~1.8
net=4.0~8.5

15 players, 4 spectators.

When I use to play siege in the mornings, it was always full server.. now when you joined, its not even people on or its 1-2 people on like kenneth said.. we are losing traffic.

I'm not saying this as a hateful thing, just saying as it is...

yes, we do have lots of people at times...but the version is not good, just listen to the people that trying to help you to realize it

[Higor]

It's an incremental issue, something's slowly crippling the server over match time.
CPU timings are great during the first minutes of the game, but as the game gets closer to overtime it starts getting a lot slower.
I've also noticed a huge increase in the actor processing times, even greater than RC12's unoptimized values as it gets over 30 minutes into overtime, and there were only 75 NPL actors during last test (Baths 4way).

I did not add anything that causes this behaviour, I do write my changes in the journal so I and everyone else can double check.

My bet's still on NexGen, which btw is spamming my log with warnings everytime a player leaves.

EDIT:
Adrian with 150 ping... we're all doomed.

[][X][~FLuKE~][X][]

Still dancing around the fact that the last few versions have had nothing but a negative response and is causing problems from the majority.

if one of my RX's got that reaction I would change it right back and then fix the issue, you keep dancing around peoples opinions.

fleece and kenneth are both right, the server traffic has changed and traffic has dropped, mainly due to rage quitting over the sniper not working correctly and the minigun, aka rapid fire ZP Sniper.

You may have changed a lot in your eyes Higor but dislike it or not, the people haven't seen any dramatic changes, they can only see what is on the server and that is lag, sniper not working (like blaze said, ZP211). for me and not sure who else but buildings flicker most of the time.


The whole point of this thread is as soon as a bad bug is detected, then get it off the server after each release, then put back the previous one that kept players happy and off your back.

cant escape the reality that there are huge problems Higor, like it or not.

[Crsk]

It's an incremental issue, something's slowly crippling the server over match time.
CPU timings are great during the first minutes of the game, but as the game gets closer to overtime it starts getting a lot slower.
I've also noticed a huge increase in the actor processing times, even greater than RC12's unoptimized values as it gets over 30 minutes into overtime, and there were only 75 NPL actors during last test (Baths 4way).

I did not add anything that causes this behaviour, I do write my changes in the journal so I and everyone else can double check.

My bet's still on NexGen, which btw is spamming my log with warnings everytime a player leaves.

There is no problem in configure the player mouse properties under tick function? I mean about the possible lag induced by doing so, by the re-doing thing. maybe it's helping to kinda fuck up people? you know, AntiAimCheat... hey, and hi :D

[|uK|B|aZe//.]

servers not even managing to give over 30 tick

gg siege

['Zac]

why doesn't siege use newnet?

its going to be so hard for TimTim to create a newnet siege, its constantly getting updated. Assualt doesnt even have newnet yet

[Higor]

5 hour game (Coret 4 way), at the end the profile reads this:

r=50 (nominal tickrate)
cli=28 (all clients)
act=44.2 (engine in milliseconds)
net=106.2 (net in milliseconds)
pv/c=238
rep/c=119
rpc/c=3

The server barely handled 8 fps.
We need tests in different configurations ASAP.

[Higor]

With such a slow tickrate and processing times, not even player net rates would cause this.

- I disabled the image drop fixer for a couple of minutes and that didn't solve anything.
- I don't think it's overheating, otherwise the problem would persist between map changes.
- Actor count didn't skyrocket, it increased as players joined (around 40 location markers per player), but still, these location markers exist since RC17 and are recycled when players leave and re-enter, not to mention they're not replicated at all.
- This issue is directly affecting the ZP weapons as time passes.

This server would be able to handle the AI of 150 BotZ playing a single CTF map with 5 flags on each base and yet this literally slams it down to 10 fps.

[SAM]

There's obviously issues with the version. No major server side changes. BT igus combo all on the same box and no complaints.

Time to troubleshoot further. I would suggest first step revert to zp and disable LC. If that doesn't work rollback to RC21 which seem to be the last stable version. There is nothing wrong with server otherwise we would have reports on combo BT and us ig.

Tomorrow after gym about 8pm gmt we will revert to zp. Higor would appreciate if you could let me know the values and whether it's possible to use zp still. I have been a bit out of touch lately.

[terminator]

Server was always slowing down after we played long time on it. Really out of idea what could be causing it.
Used to think that it could be due to increasing numbers of fortifications but even when u kill a big part of it , the lag persists.

Personally i've faced some zp sniper issues when it didn't register my shots. Also don't like the fact that minigun is overpowered.
Other than that it's pretty ok.

About losing traffic... i am really not tracking it but i don't remember that server was full during morning (european time).
Used to be few players.. as it was today when i joined but never full.

[Higor]

No major server side changes
Yet the entire NexGen was changed.


There is nothing wrong with server otherwise we would have reports on combo BT and us ig.
12 slots, less in-and-out player traffic, 15 minute games max.
Moot point here SAM.


There's obviously issues with the version.
I will create a .PATCH file to diff both 0010 and 0011, and you'll see if those changes really have the potential to cause such issues.

I'm really getting tired of this.

[Higor]

The hell with it, have fun pinpointing this.
I'm out.

[SAM]

I don't understand the issue with rollback ~ to see whether it's the version or not. People have been complaining before the nexgen upgrade. But I'd help to put your mind at ease can also rollback that too.

[][X][~FLuKE~][X][]

You need a thicker skin if you are going to develop a mod that is popular like this Higor and your ignorance towards peoples opinions will come back and bite you in the ass if you dont sit up and listen, I had my fair share of rage attacks on the RX releases over the years, you need to learn to read them less personal but still take in the feedback.

I would love to say you are doing an awesome job Higor but weighing up the pro's and cons of siege progression and I have to say there has been way more negative than positive, its fact Higor, not an insult.

also please stop with the patch after patch after patch man, fix the issue so it is hardcoded instead, I am getting sick of downloading constant patching and new releases and I am sure others will agree, if I am going to download over and over at least give something visible that we can see, each version seems to either be nothing noticeable change wise or something has been added/removed that immediately pisses a lot of players off.

concentrate less on doing things the people really wont like and update the look of siege builds first, give the people something they can see at least, IMO you are working on the totally wrong thing to start off, visual first which keeps the players occupied while you work on functions etc.

I am always brutally honest Higor and I dont kiss ass in case I offend a developer who then childishly rage quits the development of that mod, online and Real Life so dont take this as a rage or attack on anything you have done in a personal way, I have had more than my fair share of posts like this when I was just starting out on RX, I am simply pointing out facts that need to be addressed before you lose the support that you have right now.

Public Server traffic is dropping more (especially euro players through the day), this needs to be reverted to a good version before more people leave, roll back to a decent working version.

[Feralidragon]

As a professional programmer in a fast growing project, I must point out that a code update and server problems do not correlate with each other tons and tons of times.
It happens quite frequently me putting stuff into production, from small things to fundamental structure changes, and then the servers starting to have problems, and once we check if it's the new code or not, many many times it's a coincidence and the servers simply started to have more load or something else completely unrelated caused said problems.

My boss requested me to rollback my code in some occasions, but when I saw that there was absolutely no possible correlation between the changes, I explained it to him and it turned out I was right. In the past 2 months, there was only 1 single time where my code did cause problems, and I rolled back immediately because I saw that the changes I did could indeed cause a little bit of extra load, it just happened to cause a lot more load than I first expected.

Having that said, you guys should look at things objectively and not with circumstantial assumptions, specially considering that the server is running plenty of other unstable mods, and updates of such mods were also been installed afaik.
I think Higor should indeed show a diff between the version which worked properly and the current one, and let us the rest of developers see and check if there's something that could indeed influence any of the problems of the server. Any developer that knows the engine to some extent will be able to tell if such changes have any potential to cause problems.

Hardcoding something goes completely against proper coding practices, and if Higor started to do that you would end up with a lot more problems and what developers call: spaghetti code and "magic statements/numbers", causing the code to be actually a lot more unstable.
Furthermore patches exist with the intent to actually try to fix the problems. In many of nowadays games players complain instead on the lack of such patches as they have to deal with glitches for months. Excessive patching however also detract players from actually playing since they have to download everything again, and once the patch is too frequent the developer should think in restructuring the whole thing in a way that patches are as friendly as possible (division by modules for example, common APIs, etc). In UT this kind of restructuring is rather hard due to the fact that you have to rename the packages to avoid version mismatches in the future, but if properly thought out and developed, it's possible.
But then again, if these patches happen is a good sign things are getting looked into and developed, and in contrast with not having them at all, I think having them is preferable.

Visuals before functionality is what leads great games to a premature death. Many many games lack visual upgrades but receive fixes constantly and new functionality, and they are still pretty much alive, and some of them are almost as old as UT itself. Furthermore, Siege players complain about every single thing that changes, visuals and eye candy included.
At the first sign of lag, even if the change is a red texture being just a bit more red, people jump in and start to blame it out of sheer ignorance how this stuff really works, and even when explained, many of them still keep this almost religious belief that they're right and the developer is wrong.

All this to say that you should give the benefit of the doubt, and be tolerant, something I never saw coming from you guys in this whole year.

Perhaps a "staging" server would be best to test the versions first before they go to public though, something like a public dev-test server where everyone is able to join in, specially beta-testers, that way some problems may be detected sooner, however keep in mind that other times they just happen in production, given that as everything else in life, shit just happens, there's no way around it.
However, I guess as of now it may be to late to think in alternatives. I warned you guys before about this to eventually happen, you didn't listen.

[TimTim]

I must point out that a code update and server problems do not correlate with each other tons and tons of times.
It happens quite frequently me putting stuff into production, from small things to fundamental structure changes, and then the servers starting to have problems, and once we check if it's the new code or not, many many times it's a coincidence and the servers simply started to have more load or something else completely unrelated caused said problems.

My boss requested me to rollback my code in some occasions, but when I saw that there was absolutely no possible correlation between the changes, I explained it to him and it turned out I was right.
Yeah, this happens way too often. People are way too quick to blame developers/code for almost every little issue. Obviously, it does happen with any project under active development, but not as much as people would like to think. I guess it is easiest for people on the outside to point at developers since the developers are the only real tangible part of the stack visible at the surface. It's also incredibly easy for people to make demands and complain and say, "Just fix it!" when they have no idea how difficult some things can be. Some issues can be clearly attributed to code and development, but there are many other factors usually beyond our control that can cause problems.

[Feralidragon]

And to prove further what I and TimTim just said, coincidentally I got this just in:
me - <sends email to sys admin> frontend X is timing out;
<receives email from sys admin asking if I put anything into production>
me - Yes I did, a major one, however none of the functionalities which frontend X holds were even touched
<sys admin checks>
admin - oh yeah, backend Y was dead for some reason, just got it back up, test to confirm
me - yep, all working again and fast as it should


This sort of thing happens 2 to 3 times a week (I mean, some server in the many we have having some problem, and getting asked if I put anything into production).
It's kinda standard to ask it of course, but the cases where the code is the culprit is very rare, at least in my case (but I guess it's the same for most as well).

[][X][~FLuKE~][X][]

the main point of developing is so there is changes made that the user themselves can see if it is such a well known mod like this, once you captivate that mindset then "under the hood" so to speak kind of dev can be done without people being on your back.

my point isn't blaming the mod for the crashing or server instability, hell I have had loads of those kind of blames over the years so IMO it is normal to get that, its the sheer volume of changes and files added with little or no examples of what has been changed for the player, too many changes, too many upgrades in such a small amount of time and majority votes etc. are being ignored.

on the dev side, I can surely understand why developers do not bother with UT99 any more, a lot of the players left are brutal and love to take without giving, That's my reason for me no longer developing the RocketX mod, even more so now after VM have mutilated the RocketX.

Developers need to be humble, you have a skill, dont be arrogant with it.

[Feralidragon]

Although I agree that developers need to be humble, humbleness itself has proven to be abused by the player base very often, sometimes even by the admins, so sometimes a developer has almost the "right" to be a bit "arrogant" at times, if you know what I mean, otherwise he will simply be stomped over and over and over.

The skill needed to code doesn't compare to the actual work of research, trial and error and actual solid non-stop programming work a developer has to undertake. As you obviously know, it's not like a developer just sits down, scratches his balls and code gets magically done and everything gets working, so I would expect you to also understand that.
From this thread alone, I can see Higor trying to debug and pinpoint the root of the problems *heavily* and in an organized matter, so while he does that he gets the kind of input we're all seeing here, it's an overwhelming lack of respect.

If players and even the admin doesn't like the way Higor is running things, establish rules, rollback if you want to and set up a public server for him to test things on (a staging server), give him active support, ask him to share his code as well to check what was modified, stuff like that, stuff that actually needs to happen once something gets used a lot and to guarantee minimal service outages and problems.
If he doesn't accept doing it under those rules, then you got a point, other than that, a developer needs to see things moving and being tested to know what's right and what isn't. He's going as far as to track tickrates, actor count, etc, etc.

Complaining and saying that versions got progressively worse alone won't help, even because the only basis for such claims is his frequent announcements (and patch downloads), there's nothing more to correlate them (unless you have specific Siege bugs to report) and there's a ton of other constantly changing variables in the server for everything to go wrong, starting with the fact that Siege by itself is a long lasting game, and UT wasn't made for long lasting games as you may have noticed given the way UT is built (tons of unused properties in actors that have to be stored in memory for each spawn + the way UT garbage collects things).

So even before anything "visual" is made, the root has to be taken care off and it isn't easy in this particular gametype, specially with all the other mods involved and which affect the overall status of the game and the servers, and which may or may not have a direct correlation with both the problems and improvements in the gameplay experience.

[Shotman]

We were pretty much asking to roll back the versions to see if it will fix the problem. If it didn't then it was something server service related. Even Sam would roll back the Nexgen to see if it would fix it. Higor refused to do so, as a result the public server lost traffic.

[Feralidragon]

Sorry if this is a dumb question, but who's the admin and in charge of the servers? Or better yet, the Siege server in specific? My guess is that is Sam, not Higor?
So if everyone asked for a rollback, and Higor refused it, couldn't Sam simply do it anyway if he felt that he should do it? I mean, Sam is no noob at this sort of stuff, he's one of the most experienced admins around, he may not know UScript and stuff like Higor does, but it's not like Higor is the one holding the doe for the pizza.

I think that all of this resumes to a lack of organization between the parties. You guys have a thousand rules for the server, but apparently almost none for where it actually matters. I don't know exactly what is going on internally, so I may as well saying quite inaccurate things (I am aware of that), but from what I see from the outside, every single one of these issues would be fixed if everyone did their part properly, such as:
1 - Sam setting up a staging public server to test new versions before they go to public (not a private one, a public one). It would always be populated once the main server was full.
2 - Higor making his debugging in such server in very similar conditions while Sam tried to at least stabilize the server to a secure previous point (not only of Siege, but every upgraded, added and changed mod in the server, including SO ones from Windows updates or so);
3 - Higor posting a diff with the changes and let the other developers take a look and confirm his claims (not only to help his case, but as a means to check if there was something he missed);
4 - Every single player posting here (including admins) going to play to the staging server often (perhaps 30% of their play time) to test things up and ensure it's proper for the transfer for public.

You do these 4 things, and you eliminate a ton of problems, the tension between players and developer and admins and players again, and you get a constantly working public server while at the same time allowing the developer to take his time to figure things out and potentially helping him getting there faster.

So once again, every party needs to act, complaints and saying it got worse won't lead anywhere. You guys should take this once in a lifetime chance that Sam is rebuilding the servers to think this through and do things properly from now on, a fresh start if you will. Otherwise this kind of shit will keep happening over and over and over again for as long this community exists.
This is my opinion for what it's worth, but fact is what warned about months ago is happening now.

[SAM]

I warned everybody about this also but everyone kept on moaning and moaning. To be honest. I did say to higor if he decide to stop at any point I won't question it. I wouldn't put up with any of it.

I will post more later on this once the servers are all back up.