Sierra shutting down master servers

1356723

Comments

  • Hamachi is definitely only to be used among small groups of friends. Trying to conduct a whole multiplayer experience with Hamachi is a horrible idea.
  • Hamachi is only a temporary fix until we can get our actual replacement master server online after testing. If you read the PNet front page we don't state the use of the Hamachi software because we're using it as a temporary fix.

    I would not make a public announcement like I did if we were planning to use just the Hamachi software.
  • TThyth AKA Electricutioner
    Well, then I'd like to hear how you plan to implement an authentication system with perfect forward secrecy. There is just about no information on your website on how you plan to implement a system. I discuss the necessity for perfect forward secrecy in a secure authentication system on page 3 of this thread.

    On my end, I'm doing final debugging of IPC with the Ruby interpreter (about 95% done). Krash has sent me some modified login GUIs, but I haven't had a chance to look at them yet. The core of the Tribes 2 client/game-server side code is almost complete. The only component that hasn't been started yet is the auto-update system.

    At the moment, I think I'll be ready for a release on Wednesday.
  • The only reason we don't have anything on how we're planning on doing this because we don't want to get it online and then watch it fail. Now, if you want us to host your script, with complete credit to you of course, then we will. Just let us know what you need and we will give it to you. That's basically what I'm trying to say, just ask and we'll give. I personally don't want to see Tribes 2 or its community, cease to exist. I play the game almost every day, I'm a member of TKB (Tappa Kegga Brew).

    All I really want to do is help. Please, just let me know what you need and I'll make it happen.
  • Construction Mod Comunity, we're badass, especially the guys who do badass things, and not so specially the people like I who don't do anything.
  • Hi all,

    Just to let you know, I'm SA on the GameArena T2 server which is an Australia based server.

    We're very interested in seeing this working as we currently run two PUGs per week, and sporadic pubs during the week too.

    Please keep us posted on the TribesNext progress. If it isn't ready in time for November 1 we might run with the LAN option until it is.

    Cheers,

    G.
  • This is in support of what PNet-Michael stated... we (Rumble - War Mod) has chosen to fall into support of EWO and affiliates. We discussed Hamatchie and now plan to build a new system from the ground up. That said, it's not as bad as it sounds. I've been developing C# for many years, have M$ certs, etc, and descried this all to the EWO core players, and I have so far received positive consideration.

    My first, temp suggestion is that reputable players fill out a form with EWO, to be manually added to a web html posted list. Others (Rumble, probably V2Rebellion, etc) will then copy and post the same list, probably every 1, 2 or 3 days. Sound familiar to you old guys? The original ARPANET (pre-internet) shared host files before BIND (Dns) came out. As you probably know, Bind, a version Dns, is a Yellow Pages like system that allows computers to register their presence and find other computers. We could share a list at first, as we build the new automated T2 game server registration system (development which is already under way). Once the developed system is in place, there will be small scripts added (downloadable) on clients games and game servers that will allow these player game pcs and game hosting servers to use the new system for gamers to find servers.

    Basically a server will start, register with the new game server advertisement system, and allow players to grab that list and sign in.

    There is one more very important component to this thought and (probably today) I am going to post some of this information, including a diagram with the (considered) login system overview. But right now, a person logs in to a server in LAN mode, they can use nay name. I am warning you all right now, if you join a basic LAN / -nologin server, someone could login as anyone, and tick you off, or otherwise try to ruin a reputation.

    But our system will provide secure, end to end security. You will log into a web page, select the player you have registered, and get a ticket to login to other game servers as that person. If you, and the game server, both use this system, you will be yourself. And if someone tries to sign up as you, we have ways to determine who is who, which includes both online and player vouching... using our system, you will be you.

    I remind you all, my idea is a proposal, but easy to build. The EWO guys will decide and the more teams who jump in will have a voice at the table. They may be multiple systems. Some may work and some may not. Not everyone has to use this system, or any system. You have the freedom to do what you want. And I want you all to be sure, I know the head EWO guys, and I am allowed to speak for Rumble. We want the best for the community, and we will listen to you, whether it's a critique, or a suggestion. We may likely need T2 scripter’s and C# (and SharePoint 2007 also) developers. But we love this game and we will do all we can to support it. If you can find a way to help, please do.

    Sincerely
    ViciousStarFish
    http://www.tribalcombat.com
    http://www.tribalcombat.com/forums
    http://www.elitewarriorsonline.com
  • TThyth AKA Electricutioner
    How will you stop people from forging tickets?
  • Good question, and there is a consideration, but also please check your PM's.

    I'm adding an edit as I wanted to answer quick then come back... Thyth I've been suggested by Mike P-Net / EWO to speak with you about combining forces. I've read your thoughts and think quite highly of them. I beleive my plan is good on the server side, and I of course have not divuldged all information, just suggestions / teasers. But the game and servers are best reinforced with solid T2 scripting and Ruby, as you have suggested. In fact, my considerations are best NOT client side (which mean game ser as well). It's best considered server side. But wether my components speak XML web services, tokenized strings via http gets or posts, or net-tcp or sockets (or any combo in between), it can be built with an interface that is best communicated with Ruby on the client side to interact with the game (or game servers). Thus your code becomes a requirement for game and game server download / modification. I have been thinking of the new master service as "TNS" as a cross between the game name and DNS. So I may slip into saying that. But further, the TNS service is also an excellent into into SharePoint (WSS 3.0) as the User / Team pages, but that is another systems and another conversation. You of course would have full access to the C# code, and every effort will be made to provide a fast, stable and documented interface.

    ALong the lines of your question, encryption is easily done, but, although I am not totaly reliant on security thrue obscurity, there are a few points that are best described in private at first.
  • (btw, anyone who may have tried to email me at [email protected], forgive me, I was having a problem with that address, which has been fixed)
  • TThyth AKA Electricutioner
    Picture is worth a thousand words. See the attached image.

    Process isolated Ruby interpreter is operational. It has a couple of minor bugs that will need to be fixed before other developers poke at the Ruby interpreter, but otherwise, it's alive. I'll make a few enhancements to the interpreter in later updates of the re-re-release.

    I'll explain the contents of the console briefly (for those interested). You can see how to invoke Ruby from Torque Script, and how to invoke Torque script from Ruby in the first two commands. After that, I'm generating a new RSA-512 key to test the authentication principles (in reality, I'll be using a 4096-bit key for this purpose). I then print out the decimal contents of the RSA key -- the e, d, and n values are all what you would expect if you are familiar with RSA. I then create a test account for myself by putting the account name, a random GUID, a previously generated public key (e and n parts) to create an account template. I then use the SHA1 digester to produce a hash for that account template. I then use the Auth key to "decrypt" (sign) the account. I then verify the signature by performing an "encrypt" (verify) on the previously generated signature. I then concat the account template with the auth server signature, which makes a client certificate.

    In reality, obviously people will generate their own keys, and will send them to my auth server for signatures.

    Now, I said I thought I would be ready for a release sometime today. Unfortunately, getting to this point with the Ruby interpreter for the crypto took longer than I thought. There are still about half a dozen scripts to write, and a few minor modifications to the DLL (such as hard coding a few RSA public keys).

    The good news is, this is about 85% of the new system, right there. Editing closed source software is an absolute pain, and an absolutely slow process. Everything is down hill from here, and we will definitely be online by the deadline. I'll be working with Krash over the next few days to get a basic website online. At the very least, some download instructions for a patch installer will be there.

    Visit us in a couple of days at: http://www.tribesnext.com/
    • authSys.jpg
  • Wow, great, so we're already starting the post-master server era with the community split in half.

    Good work guys. I have more confidence in Electricutioner than in EWO.

    If you want any help with the website, let me know.
  • Hello,

    I'm here on behalf of the Euro T2 community, and I have been reading all these ideas but unfortunately, my coding skills being limited, most of it is all Chinese.

    However the fact is that the American T2 community is by far the only one with players still actually interested in the game. I can speak for the euro side and say, most of our "coders" abandoned the game long ago, leaving the few players left, alone with nothing.

    Meaning, that I'm sure you guys will find a working solution to Nov 1st, since you are all dedicated in keeping t2 on its feet.

    I have some questions nonetheless: how will this new master server work for the Europeans? and how would getting private and public servers work? Like for example if a team would want, like old times, to pay for a private server for their scrims etc...

    Anyways, keep up the good work,
  • Krypt0n!te wrote:
    Hello,

    I'm here on behalf of the Euro T2 community, and I have been reading all these ideas but unfortunately, my coding skills being limited, most of it is all Chinese.

    However the fact is that the American T2 community is by far the only one with players still actually interested in the game. I can speak for the euro side and say, most of our "coders" abandoned the game long ago, leaving the few players left, alone with nothing.

    Meaning, that I'm sure you guys will find a working solution to Nov 1st, since you are all dedicated in keeping t2 on its feet.

    I have some questions nonetheless: how will this new master server work for the Europeans? and how would getting private and public servers work? Like for example if a team would want, like old times, to pay for a private server for their scrims etc...

    Anyways, keep up the good work,

    I think everything will work almost exact the same as it does now. You'll just need to install a patch, and the server will have to be running the patch as well. Where you live shouldn't make any difference at all.
  • Kryand wrote:
    I think everything will work almost exact the same as it does now. You'll just need to install a patch, and the server will have to be running the patch as well. Where you live shouldn't make any difference at all.

    I believe he's right. As I understand it, the new server will be web based and the web is... well... worldwide! ;)
  • Well thank you for the quick response. This is good to hear because I am not ready to quit on T2, and neither are some other Euros. We have big plans for the euro community and I guess we'll have to wait after Nov 1st, to see how smooth things are going with this new server.
  • TThyth AKA Electricutioner
    Krypt0n!te wrote:
    I have some questions nonetheless: how will this new master server work for the Europeans? and how would getting private and public servers work? Like for example if a team would want, like old times, to pay for a private server for their scrims etc...

    Anyways, keep up the good work,
    The new server will work pretty much the same way. Login GUIs will be a little bit different to accommodate the new system, but once you're in the game, finding servers and joining them will work pretty much the same way as it does currently.

    Some of the design goals is to keep compatibility with existing mods. After the patch is installed on the client and server side, there shouldn't be any additional changes needed. In really rare cases, mods that change the logic of the GameConnection::onConnect function (in server.cs) in certain ways (i.e. packaging it) might break -- since this function will be hooked to provide authentication verification.

    I expect passworded servers will still work, so you can set up private servers the same way you do now.

    Finally, if you know anyone in the European community who is multi-lingual, we could use their help to provide translation into non-English languages after the system is online.
  • Well give me a list of needed languages and I'll ask around.
  • Translate it to Finnish. :D
  • -[Hastis]--[Hastis]- -]JagenSteel[-
    But what do you want to translate exactly? The game have never been translated in any language... no? :confused:
  • Wow, great job Thyth! At least someone got the work done. As I've stated in my PM to you, if you still want that server I will give it to you, the Tribes 2 community I mean, at no cost. It's free.

    I have several Premium Hypernia Dedicated Servers. One of them will be for the Tribes 2 community to make use of, it should also allow the IRC chat to come back online as well as anything else you may desire.
  • Thats good to hear.

    Long live Tribes 2
  • Phantom139Phantom139 PGD Founder
    Thyth, a friend of mine wants a linux port for this.
  • TThyth AKA Electricutioner
    Tell your friend to use Wine.
  • Looks like some talented individuals have been working overtime to keep this game up and running. Maybe even improve it. Not sure what I can bring to this project except maybe some players with bulging bank accounts.

    Let me know if I can help.
  • Phantom139Phantom139 PGD Founder
    Thyth wrote:
    Tell your friend to use Wine.

    will do.
  • -[Hastis]- wrote:
    But what do you want to translate exactly? The game have never been translated in any language... no? :confused:

    That was my next question. If its for the game itself, I think we all can manage English, however if its for something else.....

    Which ever it is, I can get translators for German, Finnish, Italian, Spanish, French, Danish, Dutch, pretty much any language you need. Just let me know so I can start contacting the ones concerned, which sometimes isn't as easy as it seems.
  • TThyth AKA Electricutioner
    As we approach the eleventh hour, I'll provide another nightly update.

    First, I've done some benchmarking of RSA key generation at various bit lengths. The longest step is the large prime number search, consuming more than 99% of the computing time. RSA-512 keys take under a 30 seconds to generate. RSA-1024 keys take up to 4 minutes to generate. Exact times are pretty variable, since there is a random element to the prime number search. In any case, RSA-512 bit keys are of acceptable cryptographic strength -- cracking them is out of the means of most people to do in timescales of under a couple of years, and generation time is pretty short. However, I think I'll also allow generating RSA keys of 768 bits and 1024 bits in the account creation system for those willing to wait a few minutes for a key generation. Cracking RSA-768 keys and RSA-1024 keys are beyond the capabilities of even the 3 letter agencies.

    At the moment, RSA key generation is a blocking operation, just like any other call to the Ruby execution environment. This basically means that the game appears frozen for the duration of the key generaton. I'll have large red text in the account creation UIs indicating this, but I'll try incorporating some sort of threading and a progress bar in later versions.

    Secondly, I have finished writing the game server side script for the login process. It creates an additional authentication phase before running the traditional loading sequence, which should last under a second. The server needs to perform only two RSA encryption operations, both with small mods, so performance impact shouldn't even be noticeable. Plenty of work went into writing this script defensively, so a client can't hack his or her way around properly identifying themselves. This script was pretty trivial to write after all of the Ruby execution stuff was integrated into the game.

    Thirdly, I have about 80% of the script backend for the client side account creation and login process complete. There is a lot of GUI interaction that has to happen, and that has barely been started, but the client component of the authentication process with game servers is complete. Account creation/retrieval/storage still needs to be finished.

    Fourthly, here is a breakdown of the completed/in-progress/planned components of the system and a status of each of them.
    1) Process isolated Ruby interpreter: 99% done (auth server RSA public key needs to be hard coded into the executable)
    2) Cryptographic provisioning written in Ruby: 100% done
    3) Server side identity verification script: 95% done (debugging is left)
    4) Client side login/account generation script: 80% done
    5) Client side GUI additions: 15% done
    6) Server side list server interface: 30% done
    7) Client side list server interface: 30% done
    8) Auto update: 0% done
    9) SierraUp drop-in replacement: 10% done
    10) Automatic installer: 0% done
    The idea is to have all 10 of these completed in some form by Saturday. It's still a lot of work.

    Finally, there were a couple of comments from a couple of weeks ago that I wanted to react to.
    To me it seems like a bunch of people wishing they could make a Tribes 2 authentication server, merely wishful thinking on everyones part. The way we used to set up our servers guarantees far less of a community which lets be honest when the master servers were up the community was still very small, and fragmented, in this case there will be less of a variety of servers, less players because most are lazy and just want to click and icon and login to a server.

    Lonewolf and EWO are going in over their heads if they think they can do what Thyth was talking about. It would take far too many resources, and someone who is actually experienced with client-server architecture and various coding languages such as C++
    DUN_DUNNA wrote:
    I did some reading since I posted this and unless Thyth is blowing smoke, you are absolutely correct.

    There are free downloads for Ruby, SHA(1-?), and RSA... that stuff can be found and implemented for free but the time, knowledge and expertise to get it done... net code, scripts to redirect DNS hosts... very few could pull this off.
    I hope that I've been able to demonstrate a cohesive plan for an easy-to-use community system, and an ability to pull it off in the last few days. It's pretty clear that without a knowledge of C++ the types of modification made to the T2 executable would have been impossible.

    Also, it's very easy to make uneducated claims from comfy armchairs. Yes, Ruby is free, but an interpreter integration with a closed source system is not something you can go onto SourceForge and download, it's a ton of work, and it happens to be an essential component to building this. Properly utilizing cryptography is also a lot harder than throwing around the terms RSA and SHA1. DUN is right that very few could pull this off -- I just happen to be one of them.
  • sloiksloik Krypton Developer
    DUN is right that very few could pull this off -- I just happen to be one of them.

    We're glad to have a coding god on our side, Thyth. Thanks for all you've done.

    Viva Tribes 2!
  • Yes, this is all very interesting. Yet I am curious about the white noise being generated about the overly described importance of cryptography and the lack of other topics that are required in a massive, multi-point communications system that, of course, takes security into account. Many languages and systems can do this, and no serious developer would ignore the required amount of security required. Furthermore, offers were extended to work in a cohesive manner, but it's become clear that some want to be first to market, cohesiveness be damned, thus accumulating allot of valuable and possibly misguided appreciation. There are multiple ways that all of this can be done and I respectfully point out that the community has always been fractured. This truly is a shame, as it is primarily driven by self-importance and immaturity. Many will also miss the disassociation between open and closed development that seems to be fueling this, although is it not fully obvious. The truly good news about all of this is that “first” may equate to vaporware, but not to market finances. It may quite likely be that multiple sites agree to shared lists at first, and they are always a backup to systems that have functional problems. And in that realm, files with preset favorites can easily be downloaded from pages that display solid servers and show instructions to get in manually. Then come along active systems that are intended to replace interactive functionality. This is where consolidated efforts would truly go a long way toward consolidating the community. But, alas and as often is the case, personalities can be found trumping principals. And in such cases, although cohesive opportunities are missed, people (players) will get choice. And votes are cast by way of usage. In the gaming community, you can always vote twice (or more) as I am sure many will. Although it would normally be considered that time will tell between systems, it is often the case that for many, friendships, associations and affiliations will happen, and color some peoples vision. Most players will probably decide to use multiple systems, since they really don’t care about who did what when, but today, tomorrow and next year are always new days and new choices. I am especially happy that by all accounts there will still be fractures to bitch about next year ;)
Sign In or Register to comment.