Join us at the Oolite Anniversary Party -- London, 7th July 2024, 1pm
More details in this thread.

[Solved] wormholes in interstellar space not quite fixed yet

For test results, bug reports, announcements of new builds etc.

Moderators: winston, another_commander, Getafix

User avatar
Zieman
---- E L I T E ----
---- E L I T E ----
Posts: 680
Joined: Tue Sep 01, 2009 11:55 pm
Location: in maZe

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Zieman »

Switeck wrote:
I can explicitly say I cannot duplicate that. :(
What's more I'm not sure what you meant by "the equipment was still primed / in use"? :?

The act of priming/firing the equipment using the activate equipment key (typically N) takes probably less than 1/20th of a second. After that moment, it's not in a "just started"/primed state. It's simply running/in use.
I meant that I had primed the equipment, and had not cancelled the priming by pressing 'Shift + n' again, thus hitting just 'n' would begin the wormhole sequence.
Switeck wrote:
I'm completely unable to both prime the equipment AND hit pause AND hit F2 all within the 1/100th to 1/20th of a second window.
No need for that. :)
If you get into a wormhole loop, just reduce your ship's speed to 0, you then stop in front of the next wormhole and can fly away or do whatever you wish (e.g. pause & restart ;) ).
...and keep it under lightspeed!

Friendliest Meteor Police that side of Riedquat

[EliteWiki] Far Arm ships
[EliteWiki] Z-ships
[EliteWiki] Baakili Far Trader
[EliteWiki] Tin of SPAM
Switeck
---- E L I T E ----
---- E L I T E ----
Posts: 2411
Joined: Mon May 31, 2010 11:11 pm

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Switeck »

This in particular caught my eye:
Zieman wrote:

Code: Select all

00:21:49.484 [script.javaScript.exception.unexpectedType]: ***** JavaScript exception (owt2o_jumper 0.42.2 - for Oolite 1.75 and later): TypeError: System.infoForSystem(galaxyNumber, system.ID).routeToSystem(System.infoForSystem(galaxyNumber, player.ship.targetSystem), "OPTIMIZED_BY_TIME") is null
00:21:49.484 [script.javaScript.exception.unexpectedType]:       ../AddOns/OneWayTicket2Oresrati v0.43 STRIPPED.oxp/Scripts/owt2o_autoJumper.js, line 34.
00:23:29.812 [script.load.world.listAll]: Loaded 73 world scripts:
after first try got completed (2 last jumps and beginning statement of script loading after I destroyed my ship by flying it to the side of Coriolis included)
I chopped my original OXP up -- leaving only the "troublesome" part, but apparently left out quite a few critical checks to prevent disastrous results...like trying to hyperspace to an unreachable system from the same unreachable system through "regular" systems elsewhere in the galaxy. The unreachable system cannot be used as the player's target destination with the remaining OXP code...it's supposed to be a "Surprise, you're screwed!" destination instead of an intentional one set by the player. :twisted: That's probably what's generating the "null" value on line 34.

I can release an updated crash-prone-only-for-me OXP that acts in a more predictable fashion if you want. It corrects a couple errors that apparently are completely unrelated to my crashes.

In talking with someone on IRC, crashing the ship into the station seems a regular problem for them when using these OXPs. I'm at a loss to explain why.
Zieman wrote:
I meant that I had primed the equipment, and had not cancelled the priming by pressing 'Shift + n' again, thus hitting just 'n' would begin the wormhole sequence.
...
If you get into a wormhole loop, just reduce your ship's speed to 0, you then stop in front of the next wormhole and can fly away or do whatever you wish (e.g. pause & restart ;) ).
I've paused during the middle of the jump chains many times, but never managed to pause while in the middle of a wormhole tunnel animation. I do admit I've not tried at TAF 16x -- the game is likely to crash before I succeed.

Interesting you found a way out of a wormhole loop, since I forgot to mention it. It helps to also re-activate the equipment if in interstellar space, since this will make the equipment think it's unable to begin and stop the wormhole chain from just continuing from its stuck point after your next hyperspace jump.


I THINK I FOUND THE TRIGGER FOR MY CRASH!


It was pointed out by Svengali in IRC that the wormhole exit tunnel animation seems a little "off" for the Boa 2. It steps down right before the end of the animation. Seems even my hacked ship has a wormhole animation problem...it skips frames.

In retesting, I discovered my Boa 2 never crashed when I tested the OXP. I'd previously been using the hacked super-ship. So I thought my hacked super-ship was buggy. So I started removing the hacked lines in my hacked ship...it still continued to crash. I've been testing this so long that I'm doing the test pattern entirely by reflex/muscle memory...and that's a bad thing because I've been doing the same stupid pet trick over-and-over again! Because I didn't want the Thargoids to "interfere" with my test, I was doing the test while cloaked. I gave the Boa 2 a cloaking device, retested with it...and it crashed as well!

PLEASE retest my OXP using a ship that can cloak! Cloak BEFORE activating the equipment. Hopefully, your ship has enough energy to complete the test! (Or just use the ship I supplied with the savegames.)
User avatar
Smivs
Retired Assassin
Retired Assassin
Posts: 8408
Joined: Tue Feb 09, 2010 11:31 am
Location: Lost in space
Contact:

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Smivs »

Switeck wrote:
I'd previously been using the hacked super-ship. So I thought my hacked super-ship was buggy. So I started removing the hacked lines in my hacked ship...
I love this board :D :D :D
Commander Smivs, the friendliest Gourd this side of Riedquat.
User avatar
Kaks
Quite Grand Sub-Admiral
Quite Grand Sub-Admiral
Posts: 3009
Joined: Mon Jan 21, 2008 11:41 pm
Location: The Big Smoke

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Kaks »

Heck, yeah! :)

Anywho, I'm knee deep in js reset 'loveliness': I think I've more or less figured out a fix, but the testing needed is ... testing, of course!

I'll see if I get the new, invisibly improved crashiness crashes once I emerge from that round of tests!
Hey, free OXPs: farsun v1.05 & tty v0.5! :0)
User avatar
Zieman
---- E L I T E ----
---- E L I T E ----
Posts: 680
Joined: Tue Sep 01, 2009 11:55 pm
Location: in maZe

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Zieman »

Switeck wrote:
PLEASE retest my OXP using a ship that can cloak! Cloak BEFORE activating the equipment. Hopefully, your ship has enough energy to complete the test! (Or just use the ship I supplied with the savegames.)
My ridiculous test-ship can stay cloaked indefinitely while constantly using ECM :D.

And yes. A quick test showed that starting the sequence when you're cloaked leads to crashes. Oolite crashed 2 times out of 2 using trunk 4660 and only installed OXPs were my ridiculous test-ship and the troublesome OXP.

Latest.log is clear, only information I get out of it (and Previous.log) is that the first time the crash occurred after jump 16 and the second time after jump 27.

I don't know if the Windows error report is useful at all, but here it is:

Code: Select all

<?xml version="1.0" encoding="UTF-16"?>
<DATABASE>
<EXE NAME="oolite.exe" FILTER="GRABMI_FILTER_PRIVACY">
    <MATCHING_FILE NAME="gnustep-base-1_20.dll" SIZE="2805248" CHECKSUM="0x93E8E479" MODULE_TYPE="WIN32" PE_CHECKSUM="0x2B2320" LINKER_VERSION="0x10000" LINK_DATE="02/16/2011 09:28:35" UPTO_LINK_DATE="02/16/2011 09:28:35" />
    <MATCHING_FILE NAME="js32ECMAv5.dll" SIZE="2393600" CHECKSUM="0x655A509" MODULE_TYPE="WIN32" PE_CHECKSUM="0x25244B" LINKER_VERSION="0x10000" LINK_DATE="03/10/2011 17:24:24" UPTO_LINK_DATE="03/10/2011 17:24:24" />
    <MATCHING_FILE NAME="libespeak.dll" SIZE="200704" CHECKSUM="0x128A8625" MODULE_TYPE="WIN32" PE_CHECKSUM="0x3140B" LINKER_VERSION="0x10000" LINK_DATE="05/02/2010 15:22:54" UPTO_LINK_DATE="05/02/2010 15:22:54" />
    <MATCHING_FILE NAME="libiconv-2.dll" SIZE="922112" CHECKSUM="0x4DBD2CBA" MODULE_TYPE="WIN32" PE_CHECKSUM="0xE3D7C" LINKER_VERSION="0x10000" LINK_DATE="12/10/2006 18:13:13" UPTO_LINK_DATE="12/10/2006 18:13:13" />
    <MATCHING_FILE NAME="libpng14.dll" SIZE="293888" CHECKSUM="0x20F8400E" MODULE_TYPE="WIN32" PE_CHECKSUM="0x53240" LINKER_VERSION="0x10000" LINK_DATE="04/12/2011 20:40:21" UPTO_LINK_DATE="04/12/2011 20:40:21" />
    <MATCHING_FILE NAME="libportaudio-2.dll" SIZE="70144" CHECKSUM="0x2E89C0B" MODULE_TYPE="WIN32" PE_CHECKSUM="0x2028E" LINKER_VERSION="0x10000" LINK_DATE="03/17/2009 17:54:48" UPTO_LINK_DATE="03/17/2009 17:54:48" />
    <MATCHING_FILE NAME="libxml2-2.dll" SIZE="1082908" CHECKSUM="0xC2F1AA72" MODULE_TYPE="WIN32" PE_CHECKSUM="0x109AD2" LINKER_VERSION="0x10000" LINK_DATE="04/08/2006 11:57:22" UPTO_LINK_DATE="04/08/2006 11:57:22" />
    <MATCHING_FILE NAME="objc-1.dll" SIZE="64512" CHECKSUM="0x259CBC90" MODULE_TYPE="WIN32" PE_CHECKSUM="0x19C41" LINKER_VERSION="0x10000" LINK_DATE="02/12/2011 01:08:46" UPTO_LINK_DATE="02/12/2011 01:08:46" />
    <MATCHING_FILE NAME="oolite.exe" SIZE="2639360" CHECKSUM="0x9D2BC578" MODULE_TYPE="WIN32" PE_CHECKSUM="0x292BC9" LINKER_VERSION="0x10000" LINK_DATE="11/22/2011 03:52:59" UPTO_LINK_DATE="11/22/2011 03:52:59" />
    <MATCHING_FILE NAME="pthreadGC.dll" SIZE="58131" CHECKSUM="0xBD09D74F" MODULE_TYPE="WIN32" PE_CHECKSUM="0x1BA2B" LINKER_VERSION="0x10000" LINK_DATE="10/11/2004 12:23:32" UPTO_LINK_DATE="10/11/2004 12:23:32" />
    <MATCHING_FILE NAME="SDL.dll" SIZE="320000" CHECKSUM="0xCDF0963B" BIN_FILE_VERSION="1.2.13.0" BIN_PRODUCT_VERSION="1.2.13.0" PRODUCT_VERSION="1, 2, 13, 0" FILE_DESCRIPTION="SDL" COMPANY_NAME="" PRODUCT_NAME="Simple DirectMedia Layer" FILE_VERSION="1, 2, 13, 0" ORIGINAL_FILENAME="SDL.dll" INTERNAL_NAME="SDL" LEGAL_COPYRIGHT="Copyright © 2007 Sam Lantinga" VERFILEDATEHI="0x0" VERFILEDATELO="0x0" VERFILEOS="0x40004" VERFILETYPE="0x2" MODULE_TYPE="WIN32" PE_CHECKSUM="0x58E63" LINKER_VERSION="0x10000" UPTO_BIN_FILE_VERSION="1.2.13.0" UPTO_BIN_PRODUCT_VERSION="1.2.13.0" LINK_DATE="03/07/2009 21:06:45" UPTO_LINK_DATE="03/07/2009 21:06:45" VER_LANGUAGE="English (United States) [0x409]" />
    <MATCHING_FILE NAME="SDL_mixer.dll" SIZE="315392" CHECKSUM="0x9BF648D3" BIN_FILE_VERSION="1.2.6.0" BIN_PRODUCT_VERSION="1.2.6.0" PRODUCT_VERSION="1, 2, 6, 0" FILE_DESCRIPTION="SDL_mixer" COMPANY_NAME="" PRODUCT_NAME="Simple DirectMedia Layer" FILE_VERSION="1, 2, 6, 0" ORIGINAL_FILENAME="SDL_mixer.dll" INTERNAL_NAME="SDL_mixer" LEGAL_COPYRIGHT="Copyright © 2002 Sam Lantinga" VERFILEDATEHI="0x0" VERFILEDATELO="0x0" VERFILEOS="0x40004" VERFILETYPE="0x2" MODULE_TYPE="WIN32" PE_CHECKSUM="0x0" LINKER_VERSION="0x0" UPTO_BIN_FILE_VERSION="1.2.6.0" UPTO_BIN_PRODUCT_VERSION="1.2.6.0" LINK_DATE="11/15/2004 22:18:12" UPTO_LINK_DATE="11/15/2004 22:18:12" VER_LANGUAGE="English (United States) [0x409]" />
    <MATCHING_FILE NAME="zlib1.dll" SIZE="55808" CHECKSUM="0x3FA94B99" BIN_FILE_VERSION="1.2.1.0" BIN_PRODUCT_VERSION="1.2.1.0" PRODUCT_VERSION="1.2.1" FILE_DESCRIPTION="zlib data compression library" PRODUCT_NAME="zlib" FILE_VERSION="1.2.1" ORIGINAL_FILENAME="zlib1.dll" INTERNAL_NAME="zlib1.dll" LEGAL_COPYRIGHT="(C) 1995-2003 Jean-loup Gailly & Mark Adler" VERFILEDATEHI="0x0" VERFILEDATELO="0x0" VERFILEOS="0x10004" VERFILETYPE="0x2" MODULE_TYPE="WIN32" PE_CHECKSUM="0x17753" LINKER_VERSION="0x0" UPTO_BIN_FILE_VERSION="1.2.1.0" UPTO_BIN_PRODUCT_VERSION="1.2.1.0" LINK_DATE="11/18/2003 00:29:04" UPTO_LINK_DATE="11/18/2003 00:29:04" VER_LANGUAGE="English (United States) [0x409]" />
</EXE>
<EXE NAME="kernel32.dll" FILTER="GRABMI_FILTER_THISFILEONLY">
    <MATCHING_FILE NAME="kernel32.dll" SIZE="989696" CHECKSUM="0x2D998938" BIN_FILE_VERSION="5.1.2600.5781" BIN_PRODUCT_VERSION="5.1.2600.5781" PRODUCT_VERSION="5.1.2600.5781" FILE_DESCRIPTION="Windows NT BASE API Client DLL" COMPANY_NAME="Microsoft Corporation" PRODUCT_NAME="Microsoft® Windows® Operating System" FILE_VERSION="5.1.2600.5781 (xpsp_sp3_gdr.090321-1317)" ORIGINAL_FILENAME="kernel32" INTERNAL_NAME="kernel32" LEGAL_COPYRIGHT="© Microsoft Corporation. All rights reserved." VERFILEDATEHI="0x0" VERFILEDATELO="0x0" VERFILEOS="0x40004" VERFILETYPE="0x2" MODULE_TYPE="WIN32" PE_CHECKSUM="0xFE572" LINKER_VERSION="0x50001" UPTO_BIN_FILE_VERSION="5.1.2600.5781" UPTO_BIN_PRODUCT_VERSION="5.1.2600.5781" LINK_DATE="03/21/2009 14:06:58" UPTO_LINK_DATE="03/21/2009 14:06:58" VER_LANGUAGE="English (United States) [0x409]" />
</EXE>
</DATABASE>
stdout.txt & stderr.txt are both empty.
...and keep it under lightspeed!

Friendliest Meteor Police that side of Riedquat

[EliteWiki] Far Arm ships
[EliteWiki] Z-ships
[EliteWiki] Baakili Far Trader
[EliteWiki] Tin of SPAM
User avatar
Zieman
---- E L I T E ----
---- E L I T E ----
Posts: 680
Joined: Tue Sep 01, 2009 11:55 pm
Location: in maZe

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Zieman »

And in these pics (click to enlarge) is more stuff, not shown in plain text elsewhere than in the details screen of "Send Microsoft Error Report":
Image
Image
...and keep it under lightspeed!

Friendliest Meteor Police that side of Riedquat

[EliteWiki] Far Arm ships
[EliteWiki] Z-ships
[EliteWiki] Baakili Far Trader
[EliteWiki] Tin of SPAM
User avatar
Micha
Commodore
Commodore
Posts: 815
Joined: Tue Sep 02, 2008 2:01 pm
Location: London, UK
Contact:

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Micha »

Still can't reproduce crash under Linux.

Start Oolite, load savegame A00G3Test3b, launch, turn away, press 0, press Shift-N, press n, watch jump animations, get bored.
The glass is twice as big as it needs to be.
User avatar
Svengali
Commander
Commander
Posts: 2370
Joined: Sat Oct 20, 2007 2:52 pm

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Svengali »

Micha wrote:
Still can't reproduce crash under Linux.

Start Oolite, load savegame A00G3Test3b, launch, turn away, press 0, press Shift-N, press n, watch jump animations, get bored.
I don't get CTDs too when build with debug flags, but without it sometimes happens on my box with the OXP Switeck has send.
User avatar
Kaks
Quite Grand Sub-Admiral
Quite Grand Sub-Admiral
Posts: 3009
Joined: Mon Jan 21, 2008 11:41 pm
Location: The Big Smoke

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Kaks »

While looking at the js reset problem I came across an internal timer that's only active when playing in window mode, and I'm sure there are other differences that I've overlooked.

<straw clutching mode activated> Playing in full screen or windowed might actually make a difference with this one...
Hey, free OXPs: farsun v1.05 & tty v0.5! :0)
User avatar
Zieman
---- E L I T E ----
---- E L I T E ----
Posts: 680
Joined: Tue Sep 01, 2009 11:55 pm
Location: in maZe

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Zieman »

I always play in full screen.
...and keep it under lightspeed!

Friendliest Meteor Police that side of Riedquat

[EliteWiki] Far Arm ships
[EliteWiki] Z-ships
[EliteWiki] Baakili Far Trader
[EliteWiki] Tin of SPAM
Switeck
---- E L I T E ----
---- E L I T E ----
Posts: 2411
Joined: Mon May 31, 2010 11:11 pm

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Switeck »

Kaks wrote:
Playing in full screen or windowed might actually make a difference with this one...
I get CTDs in both full screen and windowed modes.

Remotely possible this is a related crash?:

https://bb.oolite.space/viewtopic.php?f=3&t=10439
"I tried to change the code in my LongRangeScanner which allows the player to make jumps in a route by following the wormholes of another ship. The original code was using this.shipExitedWitchSpace and works perfectly, but when I changed the handler to shipWillExitWitchSpace it worked fine for a while and the play time was improved a lot. But after some jumps I got a CTD."

In the testing of my crashing OXP(s), I also discovered a program that can scan the video card's memory:
http://mikelab.kiev.ua/index_en.php?pag ... AMS/vmt_en
My video card's 1 GB ram showed no errors in those tests. It also looked possible for a small, simple program to reserve a section of the video card's memory if that memory section failed tests. It may need to block multiple memory regions for a rather damaged card. That might be useful to have for anyone who has graphics issues in Oolite or other games. Beyond this, the future of computing suggests powerful video cards may act as "math co-processors" and "number-crunchers" for tasks that may have little-to-nothing to do with graphics...though I hope Oolite won't need to do that. :lol:

Trend Micro Housecall also found no malware.

I managed to "see" 1 crash via Process Explorer -- Oolite.exe's DLL list goes almost completely red, then it all disappears from memory. Happened too fast for me to tell what crashes first, Oolite.exe or nvoglnt.dll. :(
If it wasn't for the compiled debug version of Oolite, I wouldn't even known to search out info for segmentation faults on nvoglnt.dll -- though I can't say the definition of that error helped me much:
http://en.wikipedia.org/wiki/Segmentation_fault
User avatar
Zieman
---- E L I T E ----
---- E L I T E ----
Posts: 680
Joined: Tue Sep 01, 2009 11:55 pm
Location: in maZe

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Zieman »

Did some more testing, with only 2 OXPs installed (my ridiculous test-ship & the troublesome OXP).
Looks like the JS errors are gone (using trunk rev 4660).

On Win XP 32bit the CTD ratio seems to be about 2/3 - 4/5 when using the equipment while cloaked. While not cloaked, the equipment success ratio is around 4/5 (TAF 1.0). Having TAF @ 16.0 drops the success ratio to 1/3, being stuck in wormhole-loop as the other option in both TAF cases, no CTDs.

On Win 7 64bit the ratios seem similar, unless you run Oolite as administator - then the CTD ratio while cloaked drops to somewhere between 2/5 and 1/2.
...and keep it under lightspeed!

Friendliest Meteor Police that side of Riedquat

[EliteWiki] Far Arm ships
[EliteWiki] Z-ships
[EliteWiki] Baakili Far Trader
[EliteWiki] Tin of SPAM
Switeck
---- E L I T E ----
---- E L I T E ----
Posts: 2411
Joined: Mon May 31, 2010 11:11 pm

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Switeck »

As for the actual jump-route used, it's a combination of intentionally bad jump-chains and managing to "miss" a wormhole when passing through it at TAF 16x that causes the wormhole-loop. If I put in only the best/easiest jump-chains and don't cloak, the success rate of reaching the destination is >90%.
User avatar
Micha
Commodore
Commodore
Posts: 815
Joined: Tue Sep 02, 2008 2:01 pm
Location: London, UK
Contact:

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Micha »

Can you please post the exact steps to reproduce, from starting Oolite until the crash, including which save-file, which systems you target, which order you perform the commands in, etc.
At this stage I can't help but conclude that it's either an oddity of specific Windows configurations and/or with the specific versions of the libraries which are shipped with the Windows installer. I certainly have been unable to reproduce any crashes under several different Linux installs.
The glass is twice as big as it needs to be.
Switeck
---- E L I T E ----
---- E L I T E ----
Posts: 2411
Joined: Mon May 31, 2010 11:11 pm

Re: [Solved] wormholes in interstellar space not quite fixed

Post by Switeck »

Load save game.
The starting location must be from a regular system, not interstellar space or an unreachable system.

Buy equipment from F3 menu if it's not already equipped on the ship.
F1 Launch from station.
Avoid running into stuff...
Cycle through equipment via Shift+N till equipment is armed.
Make sure the current system is the "destination system" on the F6 Short Range Chart.
F1 for forward view again (since most controls don't work from the F6 menu).
Activate Cloaking Device (0 key).
Press n key to activate equipment.
Increase speed to enter the resulting wormhole.

After ~20 wormholes, your ship will have crashed the game OR ended at the destination system OR is stuck in a wormhole loop. There is also a small chance your ship ends up at the wrong location and no more wormholes are made. If so, consider yourself lucky.

The game doesn't always crash on me even doing all the above. I probably only crash about 1 time out of 2-3. Considering there's been ~20 wormhole-jumps per reload, that actually means the odds may be closer to 1 crash out of 40-100 wormholes. The odds of crashes for ME seems worse at TAF 16x, but that could just be my perceptional bias.

Another minor discovery...

Removing this line seems to eliminate the crashes:

Code: Select all

player.ship.position =[-1000000-Math.random()*1000000, -1000000-Math.random()*1000000, -1000000-Math.random()*1000000];
Post Reply