Rhinox

Some ISPs charge you for single static-IP more, than for 100/100mbit flat connection. Others do not give you static-IP even if you'd pay it with pure gold (especially those who use private-IP & NAT for all customers)...

Unfortunatelly, public servers are no more public, if you apply whitelisting. And we are again at the beginning...

May I ask what motherboard you have? I'm especially interested in eth-chip (easy to find if you tell me mobo-model). You'd be surprised, that many manufacturers include crappy network chips even on high-end motherboards. I had one of them some time ago, been struggeling with it for a few months and finall I had to swich onboard-lan off and put extra eth-adapter (pcie) inside...

http://forums.eagle.ru/showthread.php?t=84063

Believe it or not, but he already wrote a few replies here, to this topic, one even with link on youtube-video. I suppose it was removed since then and maybe his account here locked, but I'm pretty sure he is reading this...

Closing does not help. Hiding does...

Great. Just keep talking about "script kiddies" and "noobs", if you want to provoke him for even more actions, and he kicks the last few servers out of net... :-(

Nope. In case you did not read <LockOnInstallDir>/license.txt, I'll put it straight: you do not have any right to ask for refund, even if the software does not work at all. On the other side, this is quite common for majority of all softwares. Actually, software is the only product (or service, depends how we look at it) where customer does not get any warranty, and is actually buying "cat in a bag"...

A common praxis is that security advisory is not disclosed before solution exists...

With all do respect, let me add a few points: 1. this hack (at least the one with AI-respawning) is known for much longer than two weeks. How long exactly I do not know, but even if it were only two weeks, it is quite long time. 2. we are loyal long-time customers (I do believe at least most of us) who can not use product we payed for. We have full right to be impatient. 3. *If* this forum is not way of official support as you say, then I'm still waiting for ED's official statement concerning this problem. I checked DCS website, but did not find anything. Seems to me like playing dead fish...

Do not worry about miliseconds of response-time, they are just marketing crap. No standardized method of measuring exists, and manufacturers can write whatever they want (usually the lowest value, be it black-to-white or gray-to-gray). This value is never constant over the whole colour spectrum, but it is better to have ~8ms middle value and small deviation from it in the whole spectrum, than having 2ms in some special case and 20ms in others. I have Dell 2312, IPS with 8ms, and never seen any ghosting... What imho counts more is color quality, observing angles and contrast. All of those parameters are generally much better with IPS/PVA/MVA-based LCD. One more thing which might interest you is power consumption: Hanns.G HZ281HPB: 95W Asus VE276Q: 70W iiyama X2775HDS-B1: 42W (thanks to LED-backlight)

None of them. Time of TN+Film is over. Go for IPS/PVA/MVA, i.e. iiyama ProLite X2775HDS-B1/27" for about the same price...

1. any joystick is 3 levels better than the best mouse/keyboard combo 2. if you mean it with dcs:bs seriously, do not buy entry-level joy Try eBay. Sometimes starting price is as low as 1€. With a little luck you could buy it for less than 50$...

I suppose more servers are running in "stealth" mode these days: on non-standard port, and blocking connection to ED so that they do not appear on server-list (afaik in FC2 "master-server connection" is not needed). You can connect if you know ip/port... That offers at least some basic protection. Even if hacker knows IP of the server, he would still need to find ports. And compared to deep packet inspection, it is quite easy to block any client doing port-scanning (let's say more than 10 ports out of 65536 possible)...

iptables can definitelly do that (-m string --string "whatever" has been implemented some time ago), but there are a few caveats you must be aware of: First, filtering packets based on string matching requires quite much cpu-power if every packent is to be checked and low latency is required. I have done a few experiments with this on my web-server, and cpu-load increased signifficantly. My server was more busy doing all that filtering than by web/database-server, and I had to switch to application-level firewall... Second: packets might come fragmented. You could find one part of string in one packet, and the other part in different packet. And there are ways how to force fragmentation on sending side intentionally. And to make the problem even more complicated, those fragments might arrive "out-of-order". You could make conditional nested filtering rule(s), but number of combinations (even for such a simple string as "whatever") is very high, further contributing to complexity of filtering rules. So basically, you could protect FC-server by filtering out tcp/udp packets for known strings (provided you know unique attacking vector, which might in no case be used in valid communication), but imho you would need dedicated firewall-box with at least a few dozens of quite complex rules. And I'm affraid none of those small soho-routers is powerfull enough for this job... Oh, and not to forget: what if communication si encrypted?

If I understand correctly, your 2 options are: a) Keep current system and instead get 3 new 3D Vision 2 capable monitors b) Upgrade motherboard /processor/memory with an i7-3930K Well, I'd suggest something different: c) Upgrade your GPU. 2xGTX480 is nice for common user, but I think it would bottleneck in case of 3x full-hd screen setup. Especially if you play BS/A10 on higher settings (LO is not so gpu-hungry)...

OMG, not again! Why people write questions before they use search-engine remains mystery to me. (removed, search yourself!)

Of course they can.

I think you are a little "off the track". Multiplayer mode of LO/DCS (and actually most of online/network-games) is a little different from common client-server software model. In this case, "intermediary-client" model is probably more suitable name. Server does not take input from clients (key/joy/mouse) to run simulation for them. Clients run application locally, and send to server only "results" of their user-inputs (i.e. motion-vector of player controlled aircraft). Server then feed those "results" to other clients. I suppose server does not check in detail how a client calculated that motion-vector. And it is quite obvious why: you'd need 20x more powerfull server if it had to check/recalculate results of 20 clients being sent to it. The same for respawning: it is determined by locally running application (on client-side), not by server. That is why server can not be "ultimate authority". Server can only check some critical client-files (i.e. request checksum) and compare them with its own. But this can (of course) be deceived...

"...suggested retail price of US $249..." I always considered trackir expensive as hell, but it seems in the light of kinect I will have to change my mind... On the other side, SDK is free and devs do not have to pay for license. This might speed up developing of games having kinect-support...

Actually, I'm surprised there are not many more. Usually such an exploit spreads itself very quickly. Could it be maybe mod-hack (weapon, flyable plane, etc)? Or does it happen on clean LO with standard integrity-check too?

Do not bullsh*t me EtherealN please. I think I know more about this particular subject, than you. BTW, check my previous thread concerning this: http://forums.eagle.ru/showthread.php?t=56356 I could tell you more about it were I not affraid you'd remove my message (or lock the topic, as in the above mentioned case) as soon as I put some details. So the last 2 points: 1. LockOn 1.0.x runs in virtual environment 2. LOFC/DCS does not run in vmware/vbox-guest, because StarForce detects game is running in virtual environment, and prevents starting it. As soon as you circumvent StarFroce, you can run DCS/LOFC in vmware/vbox-guest without any problem. You say no deliberate measures have been made? Hm, why does then SF check for strings like "virtual" "vbox", "vmware", etc in hw-id? It's hard to debug that SF-beast, but not impossible! ;-) AFAIK, none. But even if you loose 50% of your native hardware performance using virtual-guest, it might still be enough for running game-server. That is all I needed...

Simply put, you are wrong. LO/DCS *can* (an in the case of original LO 1.0.x even *did*) run in virtualised environment, but ED intentionally took measures to prevent it. They were quite successful in their effort, yet that protection coud be circumvented...

Nope. GTX550Ti is not "very close" to GTX560: GTX550: chip GF116, 192bit memory-interface, 192 stream units, 32 texture units GTX560: chip GF114, 256bit memory-interface, 384 stream units, 64 texture units GTX560 is about 1.6-2x faster than GTX550 and price-tag corresponds with it. If you can, go for GTX560. BS2/A10C is really gpu-hungry...

Is it the same for de-activations? IIRC there is a limited number of them too...