Jump to content

smiley

WFG Programming Team
  • Posts

    353
  • Joined

  • Days Won

    2

Posts posted by smiley

  1. 1 minute ago, Dizaka said:

    Though I believe you know that but just putting it out there.  VPN doesn't solve everything.  VPN is good for being a client.  But as a client your IP isn't compromised unless you host.

    Seems like you don't need to host to incur the wrath of the DoS gods.

    On 11/12/2020 at 1:02 AM, Grapjas said:

    Being Ddos'd into oblivion as we speak. Was playing in @nani 's game.

     

  2. On 10/12/2020 at 11:03 PM, mralex said:

    What method would a central server use to withstand the attacks?
    Why can't the same approach be applied to player hosts and clients?

    Basic enterprise grade hardware can withstand a DoS. A DDoS on the other hand, while expensive to launch is also expensive to mitigate, which is why you rent virtual servers on the cloud.

    The previous thread regarding this topic has somehow been locked down now, I am not sure if its global or just for me, but I can't reply to that in my own discretion now.

    I would want to be once again the bearer of bad news, but I no longer care and its getting old at this point.

  3. 4 hours ago, badosu said:

    I'd be up to making a mod to help with this, but I'm not sure I understand your tips.

    There are two functions exposed to JS in the lobby. You can use these functions. See, https://github.com/0ad/0ad/blob/d15248f72db6116fec09fe11b50f55a39aba5917/source/lobby/scripting/JSInterface_Lobby.h#L44
     

    void SendRegisterGame(ScriptInterface::CmptPrivate* pCmptPrivate, JS::HandleValue data);
    void SendUnregisterGame(ScriptInterface::CmptPrivate* pCmptPrivate);

     

  4. From your graphs, it looks like the router is being overloaded with packets, not necessarily bandwidth. A million tiny packets on a home router would still starve it.

    7 hours ago, badosu said:

    Is there a way to make some functionality to deregister from lobby after game starts? That could help.

    All this need in an interface to do. Alternatively, proxy the requests done through the lobby. That way, the DDoSer would need to connect to the host before he can get a public IP. Basically, don't advertise public IPs, just relay them to the actual host when they want to connect.

    • Like 1
  5. 5 hours ago, Angen said:

    some routers offer ddos protection, it is just not enabled by default. 

    also using VPN could help

    I doubt home routers will have that capability. I would assume who ever does this would know how to to actually knock out a router. DoS detection is a complex problem. Usually, DoS attacks starve out the end host, not the hardware in between. Home routers with limited memory aren't hard to starve unfortunately.

    Find where it comes from, find what it sends, null route all trafic that match the both criteria. You aren't running a server, you can safely block out entire regions.

    • Like 1
  6. 9 hours ago, badosu said:

    This could be doable on an experiment mod, I would expect performance to be incredibly poor. Another issue is defining the start quadrant, I'm not sure it's well defined (nerfs are the same regardless the initial unit).

    I made that mod and posted it on this forum sometime in 2018.

    Performance was alright, nothing intensive was needed to be done.

    I am not sure whether defining quadrants will be an issue. What I did was take the relative rotation of two entities in attack, translate for attack direction if needed and just take the circle quadrants based on angles.

    I recall having to revise it to only take the relative rotation of attacked and the attack, because missile attacks can have vastly different attacker and attack orientation depending on projectile speed.

    • Thanks 1
  7. Changing the terrain from a grid to a polygon map or something similar will allow for some remarkable art. If you can make it in blender, you can make it in Atlas kinda thing. Even allow for importing actual meshes to Atlas.

    Fun to think about, not fun to implement. Basically involves overhauling the entire terrain renderer.

  8. 3 hours ago, Nescio said:

    Wikipedia uses 13:57, 9 October 2020; github 9 Oct 2020, 13:57; both are based in the USA. The United Nations (headquarters in New York) uses 9 October 2020 as well.

    None of them are well known for their amazing UI/UX. I find it unnecessarily verbose and lengthy.

    And in my entire life, I have never found anyone who got confused by AM/PM. But by my own logic, I should be arguing for 24 hours.

    And I fail to see how moving from one arbitrary standard to another improves things. That's the point of locales. It's completely subjective. So the best thing to do is just have the default be the industry standard, which is also what all software and operating systems use out of the box. And maybe ask IPB devs to implement locales how everyone else does it.

    Regardless, this is not a hill I am gonna die on. I am pretty confident I can parse out dates in any locale.

  9. Given that en-us is the default in most cases and most of the world uses mostly American English these days, just keep en-us. And I don't believe there is anyone not knowing how to interpret the AM/PM notation. It's not hard, teaching them is easier than messing with locales.

    • Like 1
  10. 2 hours ago, m7600 said:

    and as if snarky comments was positive feedback.

     

    On 10/7/2020 at 9:14 PM, vinme said:

    XD what are you a communist? next thing u know ull be suprised at corruption being all like "it was supposed to be for the people! by the people! whaaa whaa whaaaaaaaaaaa"

     

    On 10/7/2020 at 9:26 PM, smiley said:

    You never fail to disappoint me. Sometimes, I see something and I am like, surely this is peak.

    Your snark-o-meter is malfunctioning. While I am not one to hold back in actual arguments, I went for the most neutral thing possible in that reply.

    Regardless, he takes pride in being the uncaring bada$$ who says whatever is on his mind unaplogetically as he mentioned in this thread. Don't dish out what you can't take.

  11. 1 hour ago, GunChleoc said:

    Sometimes, one just needs a break. It's not the end of the world or the project.

    Its not quite as pretty as "everyone's taking a break". But i have no stake in any of that discussion, so lets just leave it be.

  12. 8 minutes ago, vinme said:

    @smiley yeah..umm..idk how the whole 0ad power hierarchy system works and all that but tf r u on about here?

    are you seriously suprised that ppl who dont get paid to do something dont do that thing properly?

    XD what are you a communist? next thing u know ull be suprised at corruption being all like "it was supposed to be for the people! by the people! whaaa whaa whaaaaaaaaaaa"

    You never fail to disappoint me. Sometimes, I see something and I am like, surely this is peak.

     

    2 minutes ago, Dizaka said:

    In conclusion, I will actually defend the devs and all the volunteers.   They are doing an amazing job with this game and I wouldn't be posting here if they weren't.

    Alright, you are the victim here. You words are somewhat more meaningful lol.

  13. 4 hours ago, Stan` said:

    So tell me, what would you want to hear?

    I look for two things in a project. Engagement from users and engagement from devs. If said project is meant to amount to anything, devs need to take the thing seriously.

    The cruel reality of the this endeavour is that this is neither easy nor fun. If everyone is looking for something fun to do, I guess that makes a lot of sense. Noone is tackling the big issues. Noone wants to go through hell for N months for fun.

    "We are doing this for fun" makes for a good tagline. But when things amount to something, it's no longer fun, it's work. Most projects aren't abandoned because there isn't usage. Most are abandoned because something someone did for fun became useful, others found it useful, and it became an actual thing that demanded a lot more than fun, which the maintainer couldn't offer.

    I don't have authority to dictate how people should act or feel. But that attitude wouldn't really deliver what the users of the software expects. Maybe that's alright. Maybe this codebase is a tech demo for people to code on and having an end product 0AD, the RTS game built by the community for the community, is just a byproduct.

    It's not about what people want to see. It's about what people expect to get and what you can offer.

    • Like 2
  14. Ok this is just getting unprofessional from devs now...

    1. Imagine if someone with a rather pricy internet connection is on the recieving end.

    2. This is actively ruining the multiplayer experience from what I can tell.

    3. Thread was made in September 13 and no one with a blue nick has even bothered replying here.

    At least implement a central relaying proxy so people don't have to expose their public IPs.

    I will reply to my own post as well because I literally know the response.

    "This is an unpaid volunteer project"

    • Like 1
  15. 16 hours ago, Dizaka said:

    The bad IP addresses:

    104.31.64.171
    172.67.180.106
    216.105.38.13
    45.129.33.81

    All of them are from the United States from my lookup. Some behind Cloudflare. The last one from an ISP that usually host servers.

    Seychelles has some blocks close to that last IP, but it doesn't own that specific range.

    45.66.35.0/24

    45.67.14.0/23

    45.134.12.0/24

    45.141.59.0/24

    45.148.164.0/24

  16. 1 hour ago, Stan` said:

    Sure. Open Source has often exceeded the expectations. But it doesn't mean you should require it to.

    Well, of course, nobody owes anyone anything. That is a two way road though.

    When the answer to everything is being an unpaid volunteer project, it says a lot more than what I presume you even intended to.

    That statement pretty much says, "we are out of our depth here, don't expect much from us, this is an inherent aspect of the project we can't ever change, deal with it or move on". The message isn't something current and potential users and contributors like to hear. And most of the time, they will choose the latter and just move on because well, seems like it will all be a heathen effort.

    And truthfully, people don't owe usage of the software being developed as much as they presumably aren't entitled to criticism because we are all altruistic here. And at that point, then what? development for development's sake? the means becomes the ends?

    That last statement may seem hyperbolic, but that's exactly what I am seeing here now. "It's an unpaid volunteer project. Take your entitlement elsewhere". Sure, but then what?

    • Like 2
  17. 13 hours ago, Stan` said:

    that's the most you can expect from volunteer work.

    This part is not entirely true. Its the same argument people and companies were using against open source for years. Its just a bunch of dudes working during their nights. It wont go faster than a snail. But time and time again, it has been proven to be wrong, because those same people deliver. With all the complaining, the burnout, the turnover, things still move forward.

    And its kinda disappointing to see this from a FOSS point of view because the principle behind the movement is being contradicted and we are to believe community development is a dead end. I guess you yourself believe we are marching towards oblivion.

    I have said my interpretation of why that's not the case here, but alas, they were neither constructive nor on topic and were removed.

    • Like 1
  18. Turns are a concept of the lockstep model. Events are received and fired with the only limitation being latency. There is central authority which allows such flexibility without having to juggle knives.

    Another important consideration is that once there is central authority, the game doesn't need to have the same exact state in each client. CFixed is no longer necessary and I recall a RedFox post where he experimented with CFixed and the actual overhead it has compared to fast floating point math. His figure at the time was 25% performance boost. I can't verify that, so take that with a grain of salt. Any slight deviation would just be overridden by the server overload.

    I cant think of a recent game where devs opted to go for a lockstep model. (Even AoE2:DE revised its predecessor's networking code too).

    The more efficient things are made, the more load the server has to take, so that's a trade off. In everybody else's case, Valve is ready with open arms with game servers.

    To answer your question, lockstep is impossible to make secure. Which was one of the main reason devs switched to a full client server model in the first place, that and not having to worry about slight cpu optimizations and things across different architectures completely derailing the game.

×
×
  • Create New...