If anyone still has a *working* ns-v1.04 binary, I'd like to get hold of it so that I can host a 1.04 server.

the sites I went to (all of them listed with a google search) cause a segmentation fault. 1.0 "runs" but is laggy as hell.

I'd appreciate any help on this that can be given.

Pebbs

The 1.04 binary from redphive.org works for me.

QUOTE (Lumpy @ Jan 6 2004, 10:08 AM)

The 1.04 binary from redphive.org works for me.

I'm getting a segmentation fault with that binary.

However the full NS 1.0 server works... but is laggy.

Any ideas?

Somewhat ironically I get DLL API not found errors with the original binary that came with ns_v1_linux_dedicated_server.tar.gz from redphive.org. However just extracting the 1.04 files/folders corrected all of that.

Do the 1.01 1.02 1.03 give seg faults also?

I'm guessing you extracted all of the files/folders from the 1.04 zip and not just the binary itself. It would probably give seg faults if you didn't extract the other files too which it probably relies on.

Also make sure you didn't have a brain fart like I almost did and extract the 1.04 into your 2.0 instead of the 1.0. I found myself in the middle of doing that. I guess I was thinknig I was trying one of the various 2.01's back before it became final.

I doubt you did anything like that but I figured I'd mention it to cover all bases.

Also I can't see how it would/should matter but....
FreeBSD 4.9-Stable
linux_base-debian-3.0.23 as my linux emulation
3.1.1.0 with the boffix patch.

I wish I could come up with something a little more concrete for you.

Additionally: I want to say I've seen 2.01 servers running ns_nancy. Possibly someone modified it to work with 2.01?

http://www.dexworld.org/forums/viewtopic.php?t=985

Actually there was a suggestion to play ns_nancy on a 1.04 server so that people understood what it was like in 1.04.

And to clear up some things, I did unzip all the contents to the server, I used to run a 1.04 server, and I remember making that mistake the first time...

I'll see if I can get this beast running.

Pebbs

You ever figure out the problem?

QUOTE (Lumpy @ Jan 11 2004, 09:00 PM)

You ever figure out the problem?

nope

I really should though.

The 1.0 server runs "fine" without the patch, just lagged to death. And the 1.04 server just seg faults all the time.

Pebbs

you got PM! (?)

QUOTE (TromPT @ Jan 12 2004, 11:56 AM)

you got PM! (?)

Thank you so much.

I am downloading at the moment.

Maybe you guys could get good pings to the server where it's hosted when it's up?

I'll definitely make a point of posting the IP if it can work.

Pebbs

I got a bit further with the NS I got from TromPT.

This is the problem I am getting now:

CODE

---------------------------------------------- CRASH: Tue Jan 13 00:37:12 CST 2004Start Line: ./hlds_i686 -debug -pidfile ./hl.pid -game ns +ip 69.56.235.110 +port 27016 +maxplayers 16 +map ns_nancyCore was generated by `./hlds_i686 -debug -pidfile ./hl.pid -game ns +ip 69.56.235.110 +port 27016 +ma'. Program terminated with signal 11, Segmentation fault. #0  0x4010d5ce in _int_free () from /lib/libc.so.6 #0  0x4010d5ce in _int_free () from /lib/libc.so.6 End of crash report ----------------------------------------------

Anyone understand what this means? Is my "glibc too up2date?"

Pebbs

It just occurred to me that you are, or may be, using steam. I only think this because of the hlds_i686 which only exists in steam for me.

1.0 gives the same DLL API errors in team. 1.04 works fine though. I run it via the ./hlds_run -game ns etc command and not the ./hlds_i686

I tried the ./hlds_i686 and got:

Error:libSteamValidateUserIDTickets_i386.so: cannot open shared object file: No such file or directory
Unable to load engine, image is corrupt.

Have you tried the ./hlds_run method instead? I was always under the impression that was the correct way to run the server and it then would load the correct cpu type binary. In my case it loads the hlds_i686 because I have dual 500mhz celerons. That seems correct to me.

I don't know how helpfull all of this is to you. I'm just posting it as it does work for me. Maybe the way I do things or whatever will ring a bell for you or someone else so you can get it working correctly.

Correct on that post being about steam.

That was taken from debug.log

my hlds_run command looks like this:

CODE

./hlds_run -debug -pidfile ./hl.pid -game ns +ip 69.56.235.110 +port 27016 +maxplayers 16 +map ns_nancy

The output then spits out:

QUOTE

Auto detecting CPU Using Pentium II Optimised binary. Enabling debug mode Auto-restarting the server on crash Console initialized. ./hlds_run: line 412:  4893 Segmentation fault      (core dumped) $HL_CMD debug.cmds:1: Error in sourced command file: Cannot access memory at address 0xbfffd93c email debug.log to linux@valvesoftware.com Tue Jan 13 17:48:46 CST 2004: Server restart in 10 seconds Tue Jan 13 17:48:48 CST 2004: Server Quit

This is what comes from running the latest hlds_l_1111e server

command line:

CODE

./hlds_run -debug -pidfile hlds.ns.won.pid -game ns +ip 69.56.235.110 +port 27017 +maxplayers 16 +map ns_nancy

what is put in the console when I run it:

QUOTE

Auto detecting CPU Using Pentium II Optimised binary. Enabling debug mode Auto-restarting the server on crash Console initialized. ./hlds_run: line 339:  5572 Segmentation fault      (core dumped) $HL_CMD debug.cmds:1: Error in sourced command file: Cannot access memory at address 0xbfffdc4c email debug.log to linux@valvesoftware.com Tue Jan 13 17:56:11 CST 2004: Server restart in 10 seconds Tue Jan 13 17:56:13 CST 2004: Server Quit

from debug.log

CODE

---------------------------------------------- CRASH: Tue Jan 13 17:56:11 CST 2004 Start Line: ./hlds_i686 -debug -pidfile hlds.ns.won.pid -game ns +ip 69.56.235.110 +port 27017 +maxplayers 16 +map ns_nancyCore was generated by `./hlds_i686 -debug -pidfile hlds.ns.won.pid -game ns +ip 69.56.235.110 +port 27'. Program terminated with signal 11, Segmentation fault. #0  0x4010d102 in _int_malloc () from /lib/libc.so.6 #0  0x4010d102 in _int_malloc () from /lib/libc.so.6 End of crash report ----------------------------------------------

What seems to get me, is it's a problem all the time with the /lib/libc.so.6 file. Is there something I need to recompile to fix it?

Pebbs

I can't imagine there'd be a need to recompile the libc.so.6 file. That file was installed with my debian linux emulation and I never had to do anything to it myself. That may not pertain to your setup though.

After reading and then rereading your post though it almost sounds like it's an actual RAM issue. It states it can't access memory at a certain memory address. It's also exiting with a signal 11 and then seg faulting. I've always been told that signal 11 seg faults can be memory related.

I also found the following from this forum

"Segmentation fault means your app was trying to access memory outside its
address space. This means bug. What's the language? If it is C you could
compile it with debug and use gdb to analyze the core file that "may have
been generated". Has it?"

This sounds somewhat legit as in this thread verbose asked what gdb -q hlds hlds.core responded with. I think in this case he was looking to see exactly what signal killed the server. If that's all then it's somewhat pointless as we already know it was a signal 11.

When I run the command on my server it responds with

Deprecated bfd_read called at /usr/src/gnu/usr.bin/binutils/gdb/../../../../contrib/gdb/gdb/dbxread.c line 2627 in elfstab_build_psymtabs
Deprecated bfd_read called at /usr/src/gnu/usr.bin/binutils/gdb/../../../../contrib/gdb/gdb/dbxread.c line 933 in fill_symbuf
Core was generated by `hlds'.
Program terminated with signal 6, Abort trap.
/lib/libdl.so.2: No such file or directory.
#0 0x28092781 in ?? ()

In your case it most likely will state signal 11 which we already knew from the previous error messages. However that's just conjecture on my part. There may be some additional information in that output that I have no idea what it means. I'm never used a debugger before and have no idea what information to be looking for or anything like that. Maybe a PM to verbose will clarify exactly what the use of it may be in this particular case.

My guess, and my minimal testing supports this, is that 1.04 just won't run on the x.1.1.1 branch of HLDS.

3.1.1.0 + 1.04 fired right up.
3.1.1.1e + 2.01 no problems
3.1.1.1 + 1.04 segfaults
3.1.1.1e + 1.04 segfaults

Not ruling out operator failure here, but I'm fairly confident I didn't botch my testing.

QUOTE (verbose @ Jan 14 2004, 01:42 AM)

My guess, and my minimal testing supports this, is that 1.04 just won't run on the x.1.1.1 branch of HLDS. 3.1.1.0 + 1.04 fired right up. 3.1.1.1e + 2.01 no problems 3.1.1.1 + 1.04 segfaults 3.1.1.1e + 1.04 segfaults Not ruling out operator failure here, but I'm fairly confident I didn't botch my testing.

Thank you Verbose, it never occurred to me to use the 1.1.1.0 branch.

Much appreciated.

Pebbs

Is 3.1.1.0c + the boffix patch the latest in that branch? Any known security problems with that combination besides the file-download problem?

1.04 works fine for me on 3.1.1.0c with boffix, 3.1.1.1e, and steam.

FreeBSD 4.9-STABLE
Debian linux emulation rather than the gentoo or the redhat emulation.

Have never used the gentoo emulation before as today was the first time I saw it. However in the past I was using the redhat emulations for a BF1942 server and ran into issues. It seems the debian just ran the server so much better. I tried installing the redhat emulation as a test but it kept running out of space on root. Wish I could come up with up my little theory below with some possible facts.

I can't imagine it'd work on freebsd with debian emulation but not on a real linux distro. I can't even imagine it working on one distro and not another. Yet it almost seems like that's what it's boiling down to somehow.

Be nice to know who else can and can't get it to work, and what OS/distro they're using.

Debian > everything else

Are you offering proof of that by also saying that 1.04 works in 3.1.1.0c with boffix, 3.1.1.1e, and steam for you on debian? Or are you just spamming?

QUOTE (Lumpy @ Jan 14 2004, 11:40 PM)

1.04 works fine for me on 3.1.1.0c with boffix

this is my setup and no problems
adding mm 1.17 amx 0.9.7+plugs

QUOTE (TromPT @ Jan 16 2004, 12:15 PM)

QUOTE (Lumpy @ Jan 14 2004, 11:40 PM) 1.04 works fine for me on 3.1.1.0c with boffix this is my setup and no problems adding mm 1.17 amx 0.9.7+plugs

TromPT rocks!

But besides that, working on:

Red Hat 9.0
hlds_l_3110c+boffix

by tonight I'll announce the IP of the server so that people can play.

Pebbs

It's up and running.

See this topic for more info: http://www.unknownworlds.com/forums/in...=8&t=59395&st=0

Pebbs

QUOTE (Lumpy @ Jan 14 2004, 08:40 PM)

1.04 works fine for me on 3.1.1.0c with boffix, 3.1.1.1e, and steam. FreeBSD 4.9-STABLE Debian linux emulation rather than the gentoo or the redhat emulation.

Nice find Lumpy: using the debian compat setup.

I confirm that NS 1.04 works with 3.1.1.1e under FreeBSD using the linux_base-debian port instead of linux_base. For those wanting to try this, I did nothing special when installing that port (and accepted the various defaults when prompted).

You'll still want a Linux /proc FS mounted, as 3.1.1.1 verbosely complains about while using the debian libs: "Unable to open /proc/85646/stat" scrolling by. mount -t linprocfs linprocfs /compat/linux/proc

Every combination of ns1 or ns2 and any HLDS I tried worked using linux_base-debian. w00t debian.

I did notice an issue when installing Debian on a system that did not already have linux compatability installed. It seems when installing any of the redhat bases from the ports at one point you can see it actually enable linux compatibily. It looks very similiar to the linux_enable="yes" command in rc.conf. However the debian port didn't do this, at least not the last time I installed it. It would die with an elf binary unknown error. Or something very much like that. The fix I found was to enter linux_enable="yes" into the rc.conf and reboot, even though I didn't actually have any compatability layer installed. I then installed the debian compatability without any issues.

I found this to be easier and less time consuming than installing one of the redhat bases first just to then install the debian base and then uninstall the redhat base.

P.S. Thanks for posting that verbose. I felt somewhat foolish posting my possible theory without any real data to back it up.

Help i am trying to set up a 1.04 server and i have tried using metamod and not useing it on server versions 3.1.0.9 - 3.1.1.1 including 3.1.1.0c and i keep getting
Host_Error: Couldn't get DLL API from /usr/steam/oldserver/hlds_l/ns/dlls/ns_i386.so!
The only place i could get the 1.04 patch was redphive and i have tried 1.0-1.04. plz help me.