Canceled.
Haven’t played the game for months now, and I’m tired of waiting for TBS to come with a solution which have been talked to death several times in the past.
I’ll leave the GitHub repository behind as to allow others to use my code if desired.
Canceled.
I haven’t played the game for months now, and I’m tired of waiting for TBS to come with a solution which have been talked to death several times in the past.
I’ll leave the GitHub repository behind as to allow others to use my code if desired.
I asked the same question a while ago and never got a direct answer from the devs, so I assumed they didn’t know/try.
It may not be possible considering that Gameinfo mods are pratically serverside/standalone mods. Even if you do manage to cook the new flash content, the server still needs to be running your gamemode.
However, I did manage to make smaller changes to flash overlays and HUD elements by altering their attributes in Unreal Script. I couldn’t add new stuff, but I could alter what was in there to a certain degree.
but Chiv SDK (based on Unreal engine’s SDK) should provide you the same possibilities as standard Unreal Development Kit.
Not true, and it’s not even a matter of abstraction.
You are limited in several ways in Chivalry’s SDK. For one, you are unable to load your stuff outside of a Gameinfo subclass, which means you also are unable to alter the front end code (menus, etc).
TBS’ native code is also out of your scope, and most of them don’t even have documentation.
As for the language itself, Unreal Script is horrible (good thing UE4 will drop it). It doesn’t know whether it wants to be a high-level general purpose language or a high-level glue language.
Considering the fact that it lacks certain features that you take for granted even in lower level languages such as C++, it fails as a general-purpose language. Hell, sometimes it takes a single line of C to do a certain job, while it takes 10 to do the same in Unreal Script.
Since only fully licensed projects have access to native C++, it also fails as a glue language.
Data structures can be created using classes or structs (which unlike C++, are not classes in Unreal Script, they are more akin to C structs).
What you want to do is doable. However, it’s a big waste of time as there are better things you could do with the SDK.
Moreover, you don’t see people developing code modifications right now, that’s because TBS killed it with the lack of mutators, autodownload and a very long release time which provided us with nothing that we hadn’t in Beta already.
You SHOULD consider using another game with a proper SDK for a thesis though. Source SDK 2013 is pretty nice and you can work with a general purpose language (C++).
Stopped reading on “SDK is done”.
Do you even know what a software development kit is? http://en.wikipedia.org/wiki/Software_development_kit
Please do some research before making those claims. The ability to make maps doesn’t mean the game has a working SDK.
You also seem to have zero experience with TBS’ SDK, because if you did have some, you wouldn’t say that.
Lots of broken things (which they don’t seem to care, even if you bother to make a detailed report on it), lots of code that we don’t have access to (main menu, autodownload, etc).
I’ve worked with several IdTech 3 SDKs and Source’s SDK, and this is by far the worst one. It looks more like a SDK for a single player game rather than a SDK for a competitive multiplayer game like Chivalry.
If the SDK was truly done, the community itself could’ve made exactly what you are asking for in this thread.
Regardless, even if you erroneous think that making maps is the same as having a working SDK, once one of the problems with the current Workshop system becomes clear (I’ve detailed several of them already, but was duly ignored), the SDK will be done for, rather than “done”.
That’s a pretty low end system, I’m surprised you can even boot the game given how atrocious it’s resource utilization is. My system is like 8 times faster and I still have trouble. I bet you can run well optimized games like BF3 at medium settings though.
That was a good CPU when I bought it 6 years ago. :D
I just don’t have any new games that I care about, everything else that I currently play runs maxed out at 60+ FPS.
All the rest of the work I do goes just fine on a dual core at 3.2ghz, so not much need for me to bother buying new hardware right now. The fact that I’m no longer playing Chivalry only makes it even less relevant.
But yes, before the customization update I could run it fairly well, although I did some ini tweaking. GPU was still bottlenecked though.
To clear a few things out.
Chivalry’s code is not nicely written. Bad design choices, bad organization, overpacked conditional testing, expensive looping, etc.
Combined with the fact that UDK is single-threaded (the engine might create new threads, but the game itself will run with no parallelism nor multi-processing capabilities).
With that being said, overclocking your CPU WILL improve your overall game performance. It wastes so much CPU time in a single core/CPU that the faster your core/CPU runs, the faster you process all those useless instructions.
Back when the customization update was released, my performance significantly dropped to a point where I began benchmarking both CPU and GPU usage.
My GPU was always bottlenecked by the CPU, no matter which graphics setting I set it to.
For reference, I’m running a C2D E6750 overclocked to 3.2ghz and an ATI HD 5770 overclocked to 950/1350.
Maps like Citadel were unplayable for me with the server running at 32/32.
What TBS should’ve done instead of adding even more code into an already bad structure (Deadliest Warriors), was to hire someone with the single purpose of restructuring and cleaning their code.
The more you add into an already bad code, the more expensive and long the process to restructure it becomes.
Restructuring their code would lead to several benefits such as improved performance, improved debugging, improved code expansion and removal of a significant quantity of bugs as the process goes (mainly the tricky ones).
Projects that start as amateur eventually tend to do that as they become more professional.
But meh, that’s TBS being TBS.
Besides, most of these maps are bad for multiple people. The fps will likely go way down on maps like diningroom and courtyard.
To avoid the spawn issue however, you could spawn people in different realities and then bring them back to reality 0 when they are out of the spawn point.
Changing the spawn position around a spawn point could cause you to spawn a player inside an object or in a non-preferred volume (water, out of level, etc), and performing checks to avoid that could be expensive to the game (remember that the game wastes a lot of cpu time on its own).
If you are thinking about team duels, this is the best that I could think of viewtopic.php?f=51&t=16263
However, thanks to TBS, it’s been hanging for months.
I must say I’m not happy with the progress of the SDK. Pushing for workshop support is nice and will definitely help with adoption of custom content, but I really don’t think it should have been a priority.
Certainly. This makes even more sense when you’ve easier, cheaper and faster alternatives.
I’ve implemented more complex auto-downloading mechanisms (not in a game though) with C only in less than a day. It’s simply not acceptable that they can’t implement an alternative in a fully developed engine and with access to its C++ structure.
Not everyone can/want to upload/download their content to/from the Steam Workshop. Forcing me to distribute my content in the way you see fit isn’t the proper way to do it.
At the moment mods will not run on dedicated servers and although several people have expressed interest in hosting servers and playing custom content, I must tell them to wait for a fix.
You can run your mod as a native package on serverside. Clients can connect to it as long as they have said content in CookedSDK (they don’t need to change their config).
If you can make enough people download/install your stuff, you should probably do that. Bear in mind that this is a Steam game though, 90% won’t even know about your mod, much less manually install them.
I know you have to quality control live releases but seriously, it’s mods and maps, people EXPECT some bugs and weird stuff (plus, it’s not even working at the moment anyway). Having runnable (although somewhat difficult to install) content in live should have been a priority, and could have been achieved a while ago. For those people desperate for new stuff, following some small install guide isn’t a big deal. Workshop could have come later and tied everything together for everyone else.
It’s a big deal when every server must compile and cook their combined modset.
Which, as I’ve already said in the past, is not fitted for Workshop.
Moreover, smaller mods like mine are intended to run over vanilla aswell as with any other mod around. Making each player download said smaller mods just to play isn’t viable either.
As of now I’m also boycotting Steam Workshop for Chivalry.
Even if at some point they add code-dependant mod support to the Workshop, I’ll not be uploading my stuff to it.
I would recommend all the other modders and mappers to also do the same until we are given alternatives to the distribution of our content.
The SDK is not fully released yet. There are some workshop visual issues that we’re trying to get resolved with Steam. So until things are up, running, released and smooth as butter, then we can proceed with other fun things.
Funny. Two weeks after that pseudo release and tadadada… you are still worrying and putting more effort into the Workshop, like we didn’t get enough of this in Beta.
Ironic how not even a Valve engine uses the Workshop as a primary mean to distribute their community content.
However, if you do think that’s the highlight of the evening, wait just a second, the game will download every single content that is installed within a server when you first attempt to join it, not just what it’s currently running.
Old engines used such a “feature” and every server ended up disabling autodownload.
Wanna play in my server hosting 100+ custom maps? Come back in a year.
Keep in mind, not everyone is working on DW.
It’s clear that CrustaceanSoup is deeply involved in both projects. Probably leading coding development in both, if you read his posts and code comments.
It was rare of me to openly criticize TBS before this lastest patch, but at this point it appears they are no longer worried about people bashing them, they are bashing themselves, on purpose.
This my children, is how you should not build your company.
A completely lack of vision for the usage of your project in the future.
Killed the development of most code-dependant mods around here (is there an active one at the moment?).
What are you going to do once you’ve overhaul mods and more complex maps (full of imported assets) when the Workshop limits it to 100 MB?