DOM-RMC-Jupiter alpha testers wanted!

Tutorials and discussions about Mapping - Introduce your own ones!
ExpEM
Skilled
Posts: 172
Joined: Wed Nov 09, 2016 1:48 am

DOM-RMC-Jupiter alpha testers wanted!

Post by ExpEM » Sun Dec 13, 2020 2:23 am

Alpha testers, please have a run with DOM-Jupiter_J_1. And let me know what you think.

The 'Cyber Dragon' can't yet attack.
Net code is non existent, I have no means to test nor the knowledge of how to implement it. I don't play online so I don't care.

What I am looking for is this crash:
Image
Specifically, I want to know when/if you DON'T see it!

Technical Shit:
The 'Cyber Dragon' is 'driven' by a pawn called 'ExpEM_FlyerPawn',if the driver misses it's Destination or otherwise gets to a location where it shouldn't be (caused by lag or general buggery) then this crash will occur.
I don't have a fix for this crash without a complete refucktoring of the code (removing the reliance on the driver pawn) but lazy.

Known ways to cause the crash:
-If your PC is old and shit.
-If you take a screenshot at the wrong time.
-If you use slomo to speed up the game.
-If you use mutators that cause lag (NW3 or Psi Weapons Dreams etc).

Known ways to prevent the crash:
-Refuctoring the code.
-Reducing Airspeed.

If you experience this crash please open up the level in the Editor, open the script for ExpEM_FlyerPawn, find this:

Code: Select all

//=============================================================================
// Global Functions.
//=============================================================================

function PreBeginPlay()
{
//====================================
	//Shift into Properties.
//====================================
	AirSpeed = 350; //400
	AccelRate = 2*AirSpeed;
	RotationRate.Pitch = MaxInt;
	RotationRate.Roll = MaxInt;
	RotationRate.Roll = 0;
Lower Airspeed in lots of 25 (eg: AirSpeed = 350 ------> Airspeed = 325 etc) compile and save then test again. And keep going until you don't get the crash.

Useful Info:
Let me know if everything is stable (No crash).
If you get the crash, let me know what AirSpeed it is (Default is 350).
Let me know what your pc is running (OS, Processor, Ram etc).
Let me know what Mutators you were running.

More useful info:
If you have a fix without refuctoring the code entirely.
If you can help with net code. I don't care if it has it or not but I can respect that the community will want it.

EDIT: Forgot to add the DL...
DOM-Jupiter_J_1.zip
You do not have the required permissions to view the files attached to this post.

User avatar
SteadZ
Average
Posts: 44
Joined: Mon Mar 10, 2014 11:19 pm
Personal rank: A Person
Location: Scotland

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by SteadZ » Sun Dec 13, 2020 4:05 am

So I gave it a whirl for you, managed to get down to 0 crashes eventually :P

The final airspeed was 75, however I did notice that stability improved at 125 (I was able to take several screenshots and it only crashed when I switched slomo to 100) and even more at 100 (lasted many sshots and slomo switches). In the end I was able to play a full game taking sshots and switching slomo throughout. Everything above 125 crashed instantly upon taking a screenshot.

The PC I'm using is a piece-of-shit HP Pavilion G6 laptop, running Windows 7 with AMD E2-3000M APU processor and 4gb ram.

No mutators were running, I was using patch 469a though (no idea if this affects the test!).

Hope this was helpful to you, it's a fun little map to run the test on and I'm looking forward to seeing the CyberDragon add some spice to the combat :D

~SZ
"The problem with dreams is... you never know when you're gonna wake up"

Image

User avatar
EvilGrins
Godlike
Posts: 7876
Joined: Thu Jun 30, 2011 8:12 pm
Personal rank: God of Fudge
Location: Palo Alto, CA

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by EvilGrins » Sun Dec 13, 2020 5:10 am

I'm holding off on viewing this map until it's ready for the contest.
http://unreal-games.livejournal.com/
Image
medor wrote:Replace Skaarj with EvilGrins :mrgreen:
Smilies · viewtopic.php?f=8&t=13758

ExpEM
Skilled
Posts: 172
Joined: Wed Nov 09, 2016 1:48 am

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by ExpEM » Sun Dec 13, 2020 5:23 am

EvilGrins wrote:
Sun Dec 13, 2020 5:10 am
I'm holding off on viewing this map until it's ready for the contest.
Fair enough, still a bit to go. After I get it stable, I need to get it attacking and using the chat (like a bot) then I'll probably call it ready for the contest.
I'd like to do more in the 'eventually' bracket too like smoothing out wing movement, get collision working, have it be destructible with bots properly attacking back, etc.
May look at net code in the future too but I'll only be able to test on LAN.

User avatar
EvilGrins
Godlike
Posts: 7876
Joined: Thu Jun 30, 2011 8:12 pm
Personal rank: God of Fudge
Location: Palo Alto, CA

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by EvilGrins » Sun Dec 13, 2020 6:32 am

Well, the contest ends in just under 3 weeks... so take your time but don't take too much time.
http://unreal-games.livejournal.com/
Image
medor wrote:Replace Skaarj with EvilGrins :mrgreen:
Smilies · viewtopic.php?f=8&t=13758

User avatar
OjitroC
Godlike
Posts: 2311
Joined: Sat Sep 12, 2015 8:46 pm

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by OjitroC » Sun Dec 13, 2020 11:30 am

I've done a quick test in 3 UT versions - not playing a match but just standing on the spot where the body armour is.

436 with OpenGL - the CyberDragon barely moves, the body just flexes before the map crashes - without taking a screenshot.

451 with OpenGL and 469b with D3D9 - the CyberDragon moves OK - crashes on taking a screenshot. This you already know.

The interesting thing is that in 436 and 451 the CyberDragon circles from left to right whilst in 469b it circles from right to left.

Since SteadZ has noted that the map doesn't crash with AirSpeed at 75 I'll try this later and see what happens.

Running UT on an iMac (3.06Ghz Intel processor with 4GB ram) in a wineskin wrapper with a version of Windows that seems to be equivalent to XP.

------------------------------------- UPDATE --------------------------------------
Quick test in 3 UT versions - AirSpeed at 75 - not playing a match but moving around.

436 with OpenGL - same result as above, almost no movement and an almost immediate crash (no mutator likely to cause a lag).

451 with OpenGL - took multiple screenshots, no crash - CyberDragon moves OK.

469b with D3D9 - took multiple screenshots, no crash - CyberDragon moves OK. This time it circled from left to right.

I didn't test slowmo - have never used it and TBH have no idea how to do so.

User avatar
SteadZ
Average
Posts: 44
Joined: Mon Mar 10, 2014 11:19 pm
Personal rank: A Person
Location: Scotland

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by SteadZ » Sun Dec 13, 2020 1:21 pm

Interesting, seems using 469 rather than 436 does have an impact after all.. I should maybe have noted also in my report that I was using OpenGL.

Slowmo is a console command: slowmo 1 = normal game speed; slowmo 0.5 = half speed; 2 = twice as fast etc. I set it to 100 during my tests to really strain the system lol

~SZ
"The problem with dreams is... you never know when you're gonna wake up"

Image

User avatar
OjitroC
Godlike
Posts: 2311
Joined: Sat Sep 12, 2015 8:46 pm

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by OjitroC » Sun Dec 13, 2020 2:43 pm

SteadZ wrote:
Sun Dec 13, 2020 1:21 pm
Slowmo is a console command: slowmo 1 = normal game speed; slowmo 0.5 = half speed; 2 = twice as fast etc. I set it to 100 during my tests to really strain the system lol
Thanks - don't know if I should try that or not - can't imagine what would happen with slomo 100 on my system :P

------------------------------------------------------------------------------------------------------------------------------

A couple more tests with AirSpeed 75:-
451 with D3D9 - same result as 436 tests - CyberDragon barely moves - crash soon after starting map and without taking screenshots - crash log
Spoiler
Show
ScriptLog: Infinite line rejected!
XC_UCC: === Last line repeats 9297 times.
ScriptLog: ---------------------------------------------------
ScriptLog: Run away While loop in UpdateBody()
ScriptLog: Segment = 1
ScriptLog: Segment = 1
ScriptLog: vScale = 52.465523
ScriptLog: Location = -0.000000,2304.000000,-960.000000
ScriptLog: Location = -0.000000,2304.000000,-960.000000
ScriptLog: Destination = 2270.079346,393.897461,-960.000000
ScriptLog: Distance is 2966.774658
Critical: appError called:
Critical: ExpEM_FlyerPawn DM-Jupiter_J_1.ExpEM_FlyerPawn0 (Function DM-Jupiter_J_1.ExpEM_FlyerPawn.UpdateBody:0253) Assertion failed, line 349
Critical: Windows GetLastError: Success. (0)
Exit: Executing UObject::StaticShutdownAfterError
Exit: Executing UWindowsClient::ShutdownAfterError
Exit: UGalaxyAudioSubsystem::ShutdownAfterError
Log: DirectDraw End Mode
Exit: UD3D9RenderDevice::ShutdownAfterError
Critical: FFrame::Serialize
Critical: UObject::ProcessEvent
Critical: (ExpEM_FlyerPawn DM-Jupiter_J_1.ExpEM_FlyerPawn0, Function DM-Jupiter_J_1.ExpEM_FlyerPawn.Tick)
Critical: AActor::Tick
Critical: TickAllActors
Critical: ULevel::Tick
Critical: (NetMode=0)
Critical: TickLevel
Critical: UGameEngine::Tick
Critical: UXC_GameEngine::Tick
Critical: UpdateWorld
Critical: XC_MainLoop
Exit: Exiting.
469b with OpenGL - works fine, multiple screenshots - flying into vacuum zone etc

So to summarise - at AirSpeed 75
469b runs fine in D3D9/OpenGL
451 only runs with OpenGL
436 crashes

-------------------------------- UPDATE----------------------------------
Tried it again in 469b with the AirSpeed at 100 - crashed when taking multiple screenshots.

ExpEM
Skilled
Posts: 172
Joined: Wed Nov 09, 2016 1:48 am

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by ExpEM » Sun Dec 13, 2020 8:16 pm

EvilGrins wrote:
Sun Dec 13, 2020 6:32 am
Well, the contest ends in just under 3 weeks... so take your time but don't take too much time.
3 weeks isn't alot of time between the kids finishing school for the year, work going into last minute panic mode and mandatory xmas gatherings. :lol2:
OjitroC wrote:
Sun Dec 13, 2020 11:30 am
The interesting thing is that in 436 and 451 the CyberDragon circles from left to right whilst in 469b it circles from right to left.
The direction is randomly determined at level start.
OjitroC wrote:
Sun Dec 13, 2020 2:43 pm
451 with D3D9 - same result as 436 tests - CyberDragon barely moves - crash soon after starting map and without taking screenshots - crash log
Spoiler
Show
ScriptLog: Infinite line rejected!
XC_UCC: === Last line repeats 9297 times.
ScriptLog: ---------------------------------------------------
ScriptLog: Run away While loop in UpdateBody()
ScriptLog: Segment = 1
ScriptLog: Segment = 1
ScriptLog: vScale = 52.465523
ScriptLog: Location = -0.000000,2304.000000,-960.000000
ScriptLog: Location = -0.000000,2304.000000,-960.000000
ScriptLog: Destination = 2270.079346,393.897461,-960.000000
ScriptLog: Distance is 2966.774658
Critical: appError called:
Critical: ExpEM_FlyerPawn DM-Jupiter_J_1.ExpEM_FlyerPawn0 (Function DM-Jupiter_J_1.ExpEM_FlyerPawn.UpdateBody:0253) Assertion failed, line 349
Critical: Windows GetLastError: Success. (0)
Exit: Executing UObject::StaticShutdownAfterError
Exit: Executing UWindowsClient::ShutdownAfterError
Exit: UGalaxyAudioSubsystem::ShutdownAfterError
Log: DirectDraw End Mode
Exit: UD3D9RenderDevice::ShutdownAfterError
Critical: FFrame::Serialize
Critical: UObject::ProcessEvent
Critical: (ExpEM_FlyerPawn DM-Jupiter_J_1.ExpEM_FlyerPawn0, Function DM-Jupiter_J_1.ExpEM_FlyerPawn.Tick)
Critical: AActor::Tick
Critical: TickAllActors
Critical: ULevel::Tick
Critical: (NetMode=0)
Critical: TickLevel
Critical: UGameEngine::Tick
Critical: UXC_GameEngine::Tick
Critical: UpdateWorld
Critical: XC_MainLoop
Exit: Exiting.
That crash is really, really wrong... Did you change anything other than the Airspeed?
I haven't tested XC Engine so that could be effecting it too.

--------------------------------------------------------------------------------------------------------------------------
Thank you all for your testing. Refuctoring seems to be the way to go forward.
I'll keep updating the spoilers thread with news.

User avatar
OjitroC
Godlike
Posts: 2311
Joined: Sat Sep 12, 2015 8:46 pm

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by OjitroC » Sun Dec 13, 2020 8:36 pm

ExpEM wrote:
Sun Dec 13, 2020 8:16 pm
That crash is really, really wrong... Did you change anything other than the Airspeed?
I haven't tested XC Engine so that could be effecting it too.
No didn't change anything else. I only run XC Engine in 451 and then it's only a fairly early version (as I can't run the runtime libraries required for later versions). The crashes in 436 are identical in the log (except, of course, for the lines to do with XC Engine, everything else is the same).

A couple of other things I noticed when running it in 469 at AirSpeed 75 (two 'games' of 20 minutes plus, each with a different renderer) :-
1 In D3D9 the CyberDragon gained height gradually until it was above the platform; with OpenGL it did not but remained low down - this could be intended?
2 When the CyberDragon is low down, the rocket launcher locks on (to the Engine.Pawn presumably) - once it gets to a certain height, it no longer locks on. Not sure but could be when it comes out of the vacuum zone.

ExpEM
Skilled
Posts: 172
Joined: Wed Nov 09, 2016 1:48 am

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by ExpEM » Sun Dec 13, 2020 9:09 pm

OjitroC wrote:
Sun Dec 13, 2020 8:36 pm
ExpEM wrote:
Sun Dec 13, 2020 8:16 pm
That crash is really, really wrong... Did you change anything other than the Airspeed?
I haven't tested XC Engine so that could be effecting it too.
No didn't change anything else. I only run XC Engine in 451 and then it's only a fairly early version (as I can't run the runtime libraries required for later versions). The crashes in 436 are identical in the log (except, of course, for the lines to do with XC Engine, everything else is the same).
Hmm okay, the interesting part in that was "This line repeats 9k times", that shouldn't be able to happen. :noidea
OjitroC wrote:
Sun Dec 13, 2020 8:36 pm
A couple of other things I noticed when running it in 469 at AirSpeed 75 (two 'games' of 20 minutes plus, each with a different renderer) :-
1 In D3D9 the CyberDragon gained height gradually until it was above the platform; with OpenGL it did not but remained low down - this could be intended?
2 When the CyberDragon is low down, the rocket launcher locks on (to the Engine.Pawn presumably) - once it gets to a certain height, it no longer locks on. Not sure but could be when it comes out of the vacuum zone.
The CyberDragon will circle the bottom of the tower until the ControlPoint is taken, then it will make its way up. If the point doesn't get captured it won't climb.
I haven't noticed anything out of the ordinary with the Rocket launcher locking on, ideally it wouldn't lock on at all until (when / if) I give it coollision and a death state. Added to the list of stuff for me to look at.

User avatar
OjitroC
Godlike
Posts: 2311
Joined: Sat Sep 12, 2015 8:46 pm

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by OjitroC » Sun Dec 13, 2020 9:42 pm

ExpEM wrote:
Sun Dec 13, 2020 9:09 pm
Hmm okay, the interesting part in that was "This line repeats 9k times", that shouldn't be able to happen. :noidea
Yes, the 436 log was 640KB after two minutes, probably with 9K repeating lines - not that unusual when something throws a repeated error. That's what I like about Higor's MiniLauncher/XC Engine as it truncates the log to unique lines, makes it much easier to read when testing things.

User avatar
sektor2111
Godlike
Posts: 5193
Joined: Sun May 09, 2010 6:15 pm
Location: On the roof.

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by sektor2111 » Sun Dec 13, 2020 11:06 pm

Interesting thoughts...
So if a map is crashing due to BAD codes spamming errors the machine is guilty because it's slow... :lol2: :sleep:.

Edit: Map named "DOM-Jupiter_J_1" did not show any issue for me. I tested it using my 440 for working. No crash no error just those with spectator jump stock trash (I gotta edit that some day...).
The Proof:
DOM-Jupiter_J1.7z
You do not have the required permissions to view the files attached to this post.
Last edited by sektor2111 on Sun Dec 13, 2020 11:45 pm, edited 1 time in total.

User avatar
OjitroC
Godlike
Posts: 2311
Joined: Sat Sep 12, 2015 8:46 pm

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by OjitroC » Sun Dec 13, 2020 11:36 pm

sektor2111 wrote:
Sun Dec 13, 2020 11:06 pm
Map named "DOM-Jupiter_J_1" did not show any issue for me. I tested it using my 440 for working.
That's intriguing - you had AirSpeed at 350 and took screenshots?

User avatar
sektor2111
Godlike
Posts: 5193
Joined: Sun May 09, 2010 6:15 pm
Location: On the roof.

Re: DOM-RMC-Jupiter alpha testers wanted!

Post by sektor2111 » Sun Dec 13, 2020 11:46 pm

I'm playing maps not taking screen-shots... I can even record images without using Engine resources...
But... I have 64 ScreenShots operated at 512×384 during a few seconds... and 6 Bots. The crash looks NORMAL, Engine breaks rendering and script fails whatever timing here, I can expect such crashes in other cases too. What's the deal then ? Making a map without game-play because of X screenshots or a map for playing game ?
When I changed resolution dragon direction was different.
That's all so far, I go to my other tasks...