![[de]](/mwf/flags/de.png "Germany")
Hello together,already for Clonk Rage I wrote a masterserver in PHP. It is on MySQL and works with the standard Clonk "Masterserver-protocol" with new hosted rounds, Updates, and ends.
It even started a simple Leaguedevelopment for it. Would an open Masterserver be helpful for OpenClonk, and should I post the source for it here, or will the "official" masterserver script be used for online games?
i think the protocol should be opened, not the source. That would it make possible to develop multiple masterservers. a second idea is to extend the clientside of the game choose-menu for example with game-folders and a online-download-page.......
[edit:]... or extended liga functionality
[edit:]... or extended liga functionality
>i think the protocol should be opened, not the source.
I guess you can just look it up in the source code of OC.
>That would it make possible to develop multiple masterservers
Well, what would be the benefits of that? We already had multiple masterservers with CR (Quit coded one) and ~nobody used it
because it's hard to use multiple of them.
there should be more than one masterserver accessible
there should be more than one masterserver accessible
> there should be more than one masterserver accessible
Why? Why do you think it's necessary to split up the little amount of games?
for example you can arrange tournaments with KO system, league or an other challenge system.
Every masterserver could provide ingame-downloads for its rounds, it can also arrange different tournament systems.. aaaand when someone is hacking one of the masterservers, you can still play on an other server.
Every masterserver could provide ingame-downloads for its rounds, it can also arrange different tournament systems.. aaaand when someone is hacking one of the masterservers, you can still play on an other server.
If we were a community like one of the other great online games communities, it would be even useful... but this is madness.
>for example you can arrange tournaments with KO system, league or an other challenge system.
With the right interface you could do that on ONE masterserver as well.
>Every masterserver could provide ingame-downloads for its rounds
And why does the ONE masterserver not link to where to download the rounds? Would have the same result
>aaaand when someone is hacking one of the masterservers, you can still play on an other server.
I think when something like that happens we have other problems than where to play
![[ch]](/mwf/flags/ch.png "Switzerland")
>I guess you can just look it up in the source code of OC.
Wait, are you working for Microsoft? That's NOT how things should be documented ;)
> Would an open Masterserver be helpful for OpenClonk, and should I post the source for it here, or will the "official" masterserver script be used for online games?
Well, the clonk.de masterserver suffers from a severe lack of manpower. It's actually able to serve multiple games, but the highest priority are the paying customers. It's both unlikely that the source code will be made available or the clonk.de instance opened for OpenClonk players soon, so a free OpenClonk masterserver will probably get players. But the biggest challenge would probably be administration, to protect the masterserver from malicious individuals. The only reason the clonk.de masterserver is not flooded with fake game entries is that CR has a secret piece of code that authenticates official CR engines. That's not possible with an open source implementation...
How would you do this? The code is open, so everybody would be able to get the private key and just sign whatever the "real" engine would have sent.
This has already been discussed in the league thread, it is not possible without any closed piece of source.
we have to have some criteria to detect a fake game
some criterias ware for example:
- open ports
- unique IP adress
- unique player names
- registered player names
- some information about the used .c4d's
[edit:]opensource means hacker scene - when there would be a leck, the one who found it out would tell us about or fixes it.
some criterias ware for example:
- open ports
- unique IP adress
- unique player names
- registered player names
- some information about the used .c4d's
[edit:]opensource means hacker scene - when there would be a leck, the one who found it out would tell us about or fixes it.
Well, it is open for unregistered CR players, so including OC players shouldn't be /that/ problematic. For getting around the fake/flood protection, there are essentially two options:
* Use an official build (with obscurity code piece)
* Show a valid Clonk registration
So we only keep out people that both insist on compiling the source code themselves and don't want to support the project. Isn't that exactly the kind of "cracker" people we don't want on our servers anyway?
* Use an official build (with obscurity code piece)
* Show a valid Clonk registration
So we only keep out people that both insist on compiling the source code themselves and don't want to support the project. Isn't that exactly the kind of "cracker" people we don't want on our servers anyway?
I think we had this argument already in the internal redwolf design forum. :-) If somebody thinks he can fund a masterserver by selling registrations, I'm not going to try to hinder that, but I'm also not going to feel "supported" by those buying these registrations, and would resent the implication that buying a registration would "support the project".
> would resent the implication that buying a registration would "support the project".
I don't get your meaning. What else should it imply from your point of view?
It would be payment for the masterserver. Nothing more, nothing less.
No. Except maybe that the masterserver helps growing the player base, which leads to more developers. So somebody who told somebody else about Clonk, or just played regularly with newbies, would support OpenClonk more than somebody who paid money.
The objective of the project is to make Clonk as fun as possible, isn't it? And I consider a working infrastructure to be an integral part of that. In my view that's more honest than marketing a worse game to more people.
Well, I don't. That might be because I never played much online, granted. A masterserver definitively makes Clonk more fun for a fraction (which even might be a majority) of the overall player base, but a better game makes Clonk more fun for every player. So giving the impression that everybody who wants to support Clonk has to pay for the masterserver is imho wrong. But also imho everybody who uses the masterserver and who doesn't support Clonk in other ways, say by testing beta versions or creating content, should consider supporting the masterserver with money.
Is a masterserver really essential? Wouldn't it be possible to distribute the info on games in a different way?
What do you think, what does the masterserver do?
Where should the clients get the informations about the IPs of the hosts from?
![[se]](/mwf/flags/se.png "Sweden")
Somehow those concept makes me wanna take out the "open" from open clonk.... Semi-open source projects who calls themselves open are lame IMO. :/
Maybe tackle the problem from a different angle? Like replacing a master server with something else or something.
Maybe tackle the problem from a different angle? Like replacing a master server with something else or something.
Well, OpenClonk will still be open. But you just will not be able to play on the league server with your own builds (aka "ha, I disable my FoW, lol") but with official ones. Sounds reasonable to me
![[us]](/mwf/flags/us.png "United States")
We had long discussions about that.
But as many relevant things (like FoW, Health-Bars, Inventory) rely on client-side code to enforce (non)-visibility, it's hard to guarantee fair games without any such restrictions. And as other games and experience showed us, if such a possibility to cheat exists, they will be used... (wasn't there such a thing in an early version of Soldat where you could adjust the fire-rate of your weapons? or am I mixing things up?)
But as many relevant things (like FoW, Health-Bars, Inventory) rely on client-side code to enforce (non)-visibility, it's hard to guarantee fair games without any such restrictions. And as other games and experience showed us, if such a possibility to cheat exists, they will be used... (wasn't there such a thing in an early version of Soldat where you could adjust the fire-rate of your weapons? or am I mixing things up?)
see at linux! it is a good example how "open" works. there are quite more hackers for linux than for windows but the linux-hackers are on the side of linux (the kernel foundation makes hacker congresses etc.), the hackers of the proprietary windows work against it - because it's commercial.
the more hackers an opensource project has, the more secure it will became.
the more hackers an opensource project has, the more secure it will became.
>the more hackers an opensource project has, the more secure it will became.
And how are you going to prevent an individual from using a cheat-engine (for example no Fog-Of-War) in public games? Because something like _that_ was the question. Not how secure linux is.
you need no cheat engine to turn down FOW - you just have to disable alpha adding
>you need no cheat engine to turn down FOW - you just have to disable alpha adding
I think even you understood that Fog-Of-War was just one example out of many.
and i thaught you understood that turning down FOW was just one of the possibilities to cheat in Clonk without changing the engine itself.
Oh. Please list those possibilities or enter them into the bugtracker. I think we should change that then.My question from here is still unanswered. Even if you take another aspect and not the Fog-Of-War.
oh, sorry for not posting those cheats, but clonk is still a proprietary software ;)
![[at]](/mwf/flags/at.png "Austria")
Stop trolling, kekeke?
you cannot prevent it - there is no total security. you can introduce some techniques to detect cheaters but there are also problems that you cannot solve and for those it is better to trust the players. just look at soldat.pl - it says "we are a hobby project and if you are cheating you will destroy the game fun - we are not commercial so we don't want to spend time and money to fix cheating bugs"
Well. Soldat is not Open Source I guess and therefore a very, very bad example.Of course, you can use a hexeditor/disassembler and analyse the source code and do weird stuff afterwards. But that mostly is not very easy. In an Open Source project you could on the other hand just take the source, comment stuff (like FoW) out and compile it. Veeery much easier.
Of course we cannot detect cheaters, but we can at least remove the very easy and obvious ways to crack the game for online play.
there is a way of giving the cheaters no chance:
video streaming.
with todays 32MBit-internet connections it's not the problem - and when it's too much anyway you can just stream the object positions. that would avoid any cheating.
video streaming.
with todays 32MBit-internet connections it's not the problem - and when it's too much anyway you can just stream the object positions. that would avoid any cheating.
> with todays 32MBit-internet connections it's not the problem
You are aware that 32 MBit/s connectons are not available everywhere, and that there are, in fact, households that aren't provided with broadband connections at all.
but you can request a cable ethernet connection - i think cable TV should be available in each village >500 people.
if not - just await sattellite connections or this project: http://tu-dresden.de/aktuelles/news/easy-c/newsarticle_view?cl=en ("32 MBit everywhere")
believe me - when the OC project will be in a playable version, those broadband connections will be available. (streaming object positions is not that much data)
if not - just await sattellite connections or this project: http://tu-dresden.de/aktuelles/news/easy-c/newsarticle_view?cl=en ("32 MBit everywhere")
believe me - when the OC project will be in a playable version, those broadband connections will be available. (streaming object positions is not that much data)
Powered by mwForum 2.29.7 © 1999-2015 Markus Wichitill
