Trials & Tribulations of Running a Server

[Mezmorki]

And another question ...

Qrack/ProQuake support downloading of mods? Where does it download the mods from? Is it possible to add other mods to that repository?

[Mindf!3ldzX]

And another question ...

Qrack/ProQuake support downloading of mods? Where does it download the mods from? Is it possible to add other mods to that repository?

I can't speak for ProQuake, but I just tested Qrack's mod autoDL feature and this is what I came up with

Qrack auto downloads the mod stuff from downloads.quake-1.com , IIRC that is run by Baker.

It autodownloaded the star.mdl (for 3wave,this is the mod I test it with, @ ctf.clanhdz.com) to quake\qrack\progs\here and other mdl files needed for 3wave.

and quake\qrack\maps\here for the map files

Quake 1 Files Depot just has a note dating back to 2009. I'm pretty sure you could get another mod/map setup for usage on that service, the problem is. Baker has not been back since 6/1/2015 when he popped into a post of mine in the Venting Forum. long story short, yes mad. BAKER Y U NO COME BACK/Y U TROLL MY POST/Sorry, I hope I have helped.

btw, the source url can be changed in both proquake and qrack, in qrack its under the cvar

cl_web_download_url "insert http address here"

R00k would know how to implement this onto a separate/different URL , as he implemented it into his engine Qrack, that means he could help you replicate it somewhere else, at least until Baker shows back up, or doesn't. I don't know if anyone else has access to the Quake-1.com repository to be honest, I sure don't haha. R00k might, Sole might.

[Polarite]

You can also use these sources for map downloads.

cl_web_download_url "http://bigfoot.servequake.com/"
cl_web_download_url "http://damage.servequake.com/"
cl_web_download_url "http://newyork.quakeone.com/"

[Mezmorki]

Alright ... getting closer ...

I figured out my lag spikes ARE from using wifi for my client connections. Switching to ethernet / hard wire lets me connect to other servers just fine with no lag spikes.

However ... when I'm connected via ethernet I CANNOT connect to my local server at 192.168.1.100. If I go back to wireless, I can connect just fine. Any ideas? I've tried purging all the network profiles, using static ip vs dhcp. Nothing seems to help. What could I be missing?

[bluntz]

Alright ... getting closer ...

I figured out my lag spikes ARE from using wifi for my client connections. Switching to ethernet / hard wire lets me connect to other servers just fine with no lag spikes.

However ... when I'm connected via ethernet I CANNOT connect to my local server at 192.168.1.100. If I go back to wireless, I can connect just fine. Any ideas? I've tried purging all the network profiles, using static ip vs dhcp. Nothing seems to help. What could I be missing?

Look in your admin settings for the router.It's set this way so the admin is on Ether (Lan)and so cant be hacked from outside,ie (wan).They likely use diff subnets on Lan wire/wireless but look the same on the surface.
IF you use the DDrt firmware as I used it for many years as Quakeone.org with the free dyndns hostname to back up much of the content from this place and the shubhub,I think what I did was to name the router the dns name and forward the portrange 26000-26020.I left it open and being on Nix nobody ever broke anything.
You can further trick the router by cloning the mac address,loop it with a hosts file and also make it fly up it's own ass but be careful not to arp poison yourself.

[Mezmorki]

Not sure what you are suggesting that I try here. My networking knowledge is fuzzy, unfortunetly.

I suspect this all has something to do with wireless vs. wired gateways or subnets, or something. How to sort it out?

I disabled DDNS for now (I'll deal with that later once I get the internal network working correctly).

[Mezmorki]

Still plugging away...

this thread ....
http://quakeone.com/forums/quake-hel...ces-qrp-2.html

Is maybe related to my problem. My client (laptop) has two network adapters (ethernet and wifi). The wifi can connect to my local server on the same hub but my ethernet cannot. I tried adding "-ip xxx.xxx.xxx.xxx" to my command line for my ethernet ip address (on the client) but that didn't help either. Something screwy with routing tables (WTF are those anyway?)?.

Other links:

http://quakeone.com/forums/quake-hel...orwarding.html
http://quakeone.com/forums/quake-hel...-exe-host.html

[Mindf!3ldzX]

I hate that your having to crawl through glass to achieve your goals bro

[Spike]

single port servers (either of the two that I mentioned earlier) - use them or pull your own hair out in frustration.
these servers can reply without getting blackholed by a nat. these servers don't have to force single-homed sockets to get the correct reply route. these servers don't require every single udp port being open to the world.

but hey, it is... was... your hair.

[Mezmorki]

I'll give FTE a shot with the NQ code turned on. That will work as a server with clients connecting using Qrack/ProQuake?

[Spike]

both fte and dp should work as nq servers with various vanillaish clients like qrack aned proquake.
put the following in your fte/autoexec.cfg file or something, and it'll be more nq-like.
sv_listen_nq 1; sv_port 26000; sv_nqplayerphysics 1
and if fte doesn't work, you can always harass me until it does work (revenge is sweet, so they say). or yeah, there's dp, but I'm not that knowledgable about that (other than from a technical perspective), but other people here should know a bit more about it.

now, I'm not saying they won't have other issues... just that basic connectivity is trivial when the server uses a single port.
and yeah, I get frustrated reading about people strugging with stuff that my own project is meant to have solved a decade ago (or 12ish years, I just checked). I wonder how many bugs it still has with it... :s

[Mezmorki]

@Spike,

Unfortunetly FTEQW exhibits the same pattern of of behavior. I can connect FTE-to-FTE with the server on ethernet and client on wireless, but if the client computer uses it's ethernet (turning off wifi) then I can no longer connect.

FYI, with both Qrack + ProQuake, I could not connect to an FTE server even when on wireless and using the config settings you mentioned. I'll play around with it a little more.

The bigger issue is that my computers simply cannot ping each other when the client is on ethernet. They both can ping the router at all times (192.168.1.1), and they can ping each other when the client is on wireless, but not when using ethernet. This is supremely frustrating and I don't know whether it is a router problem or an issue with my client computer. I need to borrow another hub/router and see if it still can't connect.

I suspect maybe it's a Win7 problem? I have another laptop with win7 that trying to use it as a "client" has the same issue

[Spike]

sv_port 192.168.1.$WIFI:26000
sv_port_ipv6 192.168.1.$ETHERNET:26000
should ensure that the operating system can't pseudo-randomise the physical interface that is used as a source address in the reply. that said, I'd like to think that an icmp would directly reply on the same inerface though, so maybe its a completely different issue...
you can add more addresses/ports with sv_addport, but you might need to reissue the command if the server is temporarily disabled, but that shouldn't normally happen when its a dedicated server at least.
an fte server's status command will display all ip address+ports, along with the category of the address etc. the ipv6 link-local addresses may also be interesting but note that the %foo bit specifies the local interface rather than the other machine's.
but yeah, make sure you're not using your server's public ip instead of its private/lan address, because its public ip will require a hairpin connection via your nat, and these are often not supported by your nat - the wifi and ethernet connections will normally have different ip addresses, and if you try to connect to the servers' wifi address over your ethernet connection then it shouldn't normally work (it shouldn't respond to the ARP requests and should result in 'destination host unreachable' errors once the ARP stuff times out).
I guess its also possible that windows is detecting that both interfaces are on the same lan and then decides to disable one, I can't say I've tried this myself though.

[bluntz]

You may have to put the router in bridge mode to get it to work.

[Mindf!3ldzX]

I would have done put it in DMZ mode and called it done at this point