Pan'Theon: Rise' of th'e Fal'Len - #1 Thread in MMO

Draegan_sl said:

As I've been working on my little MUD project using Unity I've learned a lot in a few weeks thanks to a few guys that are doing the engineering part. Unity is pretty limited in what it can do right out of the box.

Unity is just an engine. You still need a network and database solution and that isn't easy when it comes to Unity if you want your game to scale beyond like a handlefull of people and NPCs in your zone.

Unity doesn't allow you to create caves, "holes" in the terrain out of the box. You need to use 3d modelers (maya, blender) to create meshes, stick them into Unity and stitch terrain around them.

There is a lot you need to do, to make things work right for an MMORPG.

It's very easy to make a piece of terrain look cool, and drop a pre-made script and NPC into the game. It's hard to make it all work together without crushing your client with drawcalls and data management.

Click to expand...

And this relates to the development and testing of combat and class mechanics how? Yes, you need programmers to make a video game. No one disputes that. The question was, is the Pantheon team full of shit when they are saying they are working on combat now, when they don't have a programming team. My answer is "no, they arent full of shit. They could totally do it." Can't tell if you are agreeing or disagreeing?

Convo_sl said:

We did talk about it.. We want to see the game but we also see how dumb people are... I don't know if you have visited various forums but just in regards to the dungeon video, there were a ton of posts saying shit like "it's so dated looking!!" Basically ignoring any disclaimer that the game was in early development.. It's a damned if they do, damned if they don't type thing but I'd personally rather see it.

Click to expand...

It goes both ways, too. The ignorant retards say "It's so dated looking!!" and the ignorant fanboys say "I love the art direction and style!!!" The other 80% of potential players of Pantheon get scared and think the community is a bunch of ignorant retards and/or fanboys and haven't gained anything by a reveal that comes out too early, but have certainly lost some level of respect for the community and project in general.

Better to wait until at least the fanboys make some semblance of sense and don't look like a bunch of retards. I think that's actually a key in game design: make sure the released content is good enough the fanboys actually almost look like they're at all discerning instead of just blinding supporting something.

Well, I don't blame people for thinking it's easy. Especially when the Panth devs shit out "dungeon" with some of the premade Asset Packages from the Unity store. You can do that in 10 minutes if you know the basics.

If they wanted to fool someone, they should of entered their scene with a first person camera and walked around in a 30 second video and called it a day.

In order to make a really cool dungeon in Unity, it requires a lot of art assets and customization. It's easy if all your dungeon is is the same hallways of the same height like Diablo II. If you want to make large caverns with cool scenery, you really need to craft decent meshes to put textures on and link tunnels to them. Then on top of that, you need to make sure your meshes are low on vertices, resolution and polygon to keep performance up due to physics and network/db/client communications. It's really easy to make a fancy dungeon for a single player game because it's all done locally with one person, from one perspective. MMORPGs have a shit ton of considerations you have to make when building things.

Denaut_sl said:

I thoroughly enjoy using Unity, even with all its quirks, but Oloh you aregreatlyexaggerating how fast and easy it is to make an MMORPG even with such a good front-end.

Click to expand...

I am not exaggerating anything. I have been 100% consistent in my responses that there are two fundamental hurdles that using Unity creates. (1) seemless worlds are tough. (2) MMORPG networking is tough. I'll dig up the post weeks ago that I said the exact same thing and I have been saying it ever since. The question that people are asking is not whether or not you can make a MMO in Unity. Even if it was, I think you would agree that it is not only possible, but well within the skill set of a formal and professional game development house.

The question people are asking now, is "can they be doing work on the game without a team of coders." You clearly know about Unity, so I find it hard to believe that you would say anything other than "yes, they can do meaningful work without a team of programmers with Unity." Of course, at some point in the very near future, like I mentioned above, probably the first two positions hired, would be a database guy and a networking guy. I dont know how you took my response of "it is easy to get started doing work in Unity" and extrapolated it into "MMOGs are easy to make....no coding necessary"

Edited to add my post on my original thoughts on using Unity:LINK

"The networking options for MMO style games suck and you will probably have to build it from scratch."

Draegan_sl said:

MMORPGs have a shit ton of considerations you have to make when building things.

Click to expand...

You act like the people on the team are doing their first project. It's funny when people say "what have they been doing for all these months" on one hand and then on the other hand, people like you say "figuring all this shit out is harder than it seems." To be clear, if it wasn't already:

1. Combat prototyping and mechanics are easy to do in Unity. The Pantheon team has more than enough expertise on it right now to start designing the combat and the classes. It is easy in Unity and with a little practice, the technical expertise of many of the people on this board would be sufficient to do it, let alone a team of pros. This is what the team is saying they are doing and there is nothing in Unity that makes that statement sound anything but truthful.

2. They will need to hire a shitload more people to actually get past the combat prototyping stage and into full production of a MMO. Chief among these are (1) a pro networking guy b/c Unity's networking is not made for MMOs, nor is it easilly adaptable to MMOs and (2) a DB guy because prototyping typically involves hard-coding a lot of stuff that will later be handled by a DB (chief among them, is the player itself). While not easy, the networking piece is possible and can be done by someone with experience in managing and running such apsects of a MMO.

3. They will need to hire experienced artists that understand limitations of PCs to make the game look good. I can't believe I need to type this, but apparently Draegan was mistaken in thinking that I didn't realize it.

Does anyone, even the naysayers, dispute any of the above?

Oloh_sl said:

And this relates to the development and testing of combat and class mechanics how? Yes, you need programmers to make a video game. No one disputes that. The question was, is the Pantheon team full of shit when they are saying they are working on combat now, when they don't have a programming team. My answer is "no, they arent full of shit. They could totally do it." Can't tell if you are agreeing or disagreeing?

Click to expand...

Testing combat? Nothing I suppose. Developing combat? Everything. If you're creating animations, combat physics, and everything else that comes with it like evasion, collision or if your client can handle 10 people fighting one creature or fifty. Everything needs to be measured properly so your server isn't getting chunked.

They were discussing some fairly complex combat system that isn't essentially "Press 1 to cast a fireball, target does a simple resistance check." You need a programming team (or person) for that, especially balancing all of that between client and server.

So I don't seem how they are doing that unless they already have solutions for their games infrastructure in place. If they mean by "working on it" they are creating a design doc, then no they aren't full of shit. I have no idea who is on their team right now and what their abilities are (I don't care), but you can't just "work on combat" without getting all of your ducks in a row first.

Oloh_sl said:

I am not exaggerating anything. I have been 100% consistent in my responses that there are two fundamental hurdles that using Unity creates. (1) seemless worlds are tough. (2) MMORPG networking is tough. I'll dig up the post weeks ago that I said the exact same thing and I have been saying it ever since. The question that people are asking is not whether or not you can make a MMO in Unity. Even if it was, I think you would agree that it is not only possible, but well within the skill set of a formal and professional game development house.

The question people are asking now, is "can they be doing work on the game without a team of coders." You clearly know about Unity, so I find it hard to believe that you would say anything other than "yes, they can do meaningful work without a team of programmers with Unity." Of course, at some point in the very near future, like I mentioned above, probably the first two positions hired, would be a database guy and a networking guy.I dont know how you took my response of "it is easy to get started doing work in Unity" and extrapolated it into "MMOGs are easy to make....no coding necessary"

Click to expand...

If that isn't what you meant, thank you for clarifying. That was how I took it, and I think so did several other people. But, now that you've clarified your posts it is much more clear what you meant, which is reasonable.

I actually, even with our extremely good tech, would not attempt to make a seamless/zoneless world in Unity (at least with the budgets we work with). It is almost certainly possible, but I dread to think of how long and difficult such a task would be, and what you would give up for it.

As to making an MMORPG with Unity, there really isn't even a question ofifas it has already been done several times. However, MMO "engines" are kind of a funny thing because they aren't really a single engine, they are more like 2 engines, the client engine and the equally if not more important server engine. So yea, there are many examples of MMORPGs with Unity, but they only use that as their client engine, all of them had to build their own back-end (as far as I know).

You are also completely right in that you can do meaningful work in Unity without programmers, but meaningful work means bearing concrete fruits. Personally, my expectation from a team of professional designers after 2-3 months of work would be a fully playable client-side prototype using placeholder/asset store art that does an extremely good job of letting you fully play a small vertical slice of content (such as a dungeon with simple NPCs and a group of player characters that can use abilities). Much like what you and your friend built actually. What is nice about Unity is that there is little excuse fornothaving something like that.

Denaut_sl said:

To echo Draegan, I highly recommend Unity to anyone out there that is interested in making games.

Click to expand...

Ditto.

Draegan_sl said:

but you can't just "work on combat" without getting all of your ducks in a row first.

Click to expand...

Which ducks, again?

Oloh_sl said:

With all due respect, I don't think you understand what a "designer" is as used in the industry. A designer isn't a pure "thinker." A designer essentially implements content. Its isn't like, for example, a quest designer writes up a quest in a word document and then sends it to a programmer for implementation. The designer actually does the implementation via scripting. He might not be a full on programmer, but the tools for the game are typically good enough that you don't have to be. The programmer's job is to make the infrastructure so that designers can do their work efficiently. Another example is with player abilities. Its not like a programmer is needed to implement a fireball...especially in unity.All the tools you need are already in there, whether you are making a TERA like action game or a WoW like tab targeted game. In Unity, much of that work is already done, so a "designer" could start designing from day one without any additional coding needed.

Obviously, if you want to add more niche systems like player housing,a mail system and other stuff, then a coder would get it all started. But you could 100% be designing the combat and class system with not a single programmer on the team. Again, this isn't me just spouting off. I've actually done it.

Click to expand...

This was in response to Flex saying 9 designers and one coder is a bad combo for a startup MMO company. This is the part that really puzzled me as I was under the belief a team should be more programmers and engineers and Oloh's take in refuting Big Flex was to say he was wrong.

Denaut_sl said:

However, MMO "engines" are kind of a funny thing because they aren't really a single engine, they are more like 2 engines, the client engine and the equally if not more important server engine. So yea, there are many examples of MMORPGs with Unity, but they only use that as their client engine, all of them had to build their own back-end (as far as I know).

Click to expand...

There are several third parties that say they have a working MMO style network infrastructure set up and ready to go. They each cost like 1000 bucks per license (which is cheap for development studio, but not so cheap for me and my buddy fucking around). Our work around was to do this:

While we are prototyping the game, we essentially bifurcate the script on all objects into what we call "smart code" and "dumb code" using Unity's default networking "MyNetwork" testing. Said differently, we have an if statement on all relevent scripts that says, in essence, If I'm the player, use the smart code, if I'm not the player use the dumb code." This approach is not efficient at all, and is totally a "hack", but it allows us to effectively prototype for a MMO without developing the full server infrastructure on a budget.

The idea is that, once we are willing to spend some cash on the project, we will be able to, with some effort mind you, move the "smart code" to the server and the "dumb code" to the client. Also, we set up separate script to serve as the "database" where all it does is set variables. Also inefficient, but saves us the hassle of managing a DB with a static address. We can just each mess with the full code base, all within Unity at all times. We use git as version control, and we haven't had any issues at all with it.

And, we are just a couple of guys that thought "all mmos suck right now, lets make one." I am sure that a team that knows wtf they are doing could prototype combat for a few months in a way that doesn't screw them/force a complete re-write when they go full MMO.

Oloh_sl said:

1. Combat prototyping and mechanics are easy to do in Unity. The Pantheon team has more than enough expertise on it right now to start designing the combat and the classes. It is easy in Unity and with a little practice, the technical expertise of many of the people on this board would be sufficient to do it, let alone a team of pros. This is what the team is saying they are doing and there is nothing in Unity that makes that statement sound anything but truthful.

Click to expand...

I agree... but where is it? I would expect it to have been avaliable from the get go.

Oloh_sl said:

2. They will need to hire a shitload more people to actually get past the combat prototyping stage and into full production of a MMO. Chief among these are (1) a pro networking guy b/c Unity's networking is not made for MMOs, nor is it easilly adaptable to MMOs and (2) a DB guy because prototyping typically involves hard-coding a lot of stuff that will later be handled by a DB (chief among them, is the player itself). While not easy, the networking piece is possible and can be done by someone with experience in managing and running such apsects of a MMO.

3. They will need to hire experienced artists that understand limitations of PCs to make the game look good. I can't believe I need to type this, but apparently Draegan was mistaken in thinking that I didn't realize it.

Does anyone, even the naysayers, dispute any of the above?

Click to expand...

Sort of. In theory you could get away with 3 people for pre-production. Brad (as Designer), 1 client programmer, and 1 server programmer. Assets you'd buy from the Unity store or farm out a a Chinese studio. Realistically though, for a really solid pre-production, they could get away with a team like this:
* 1 Project Manager
* 1 Designer
* 2 Client programmers
* 2 Server Programmers
* 1 GUI programmer (at least 50%)
* 1 General Purpose Artist (Modeling, Concept, GUI)
* 1 Animator/Technical Artist

That is a really well rounded team skill-set solid team. Maybe throw in an additional general purpose artist or 2. You actually don't need a DB Admin at all anymore. Modern document databases like Couch and Mongo make them pretty obsolete, your server programmers will generally handle "administering" the database. I use the term loosely because there is really no such thing for lots of NoSQL databases.

The question is... what are all those extra designers you had during the prototype going to do? Nothing really, for a while at least until the tech is usable. Better off just starting with the small 3 person team I described at first and adding what you need when you need it. Lots more designers really only come into play once you go to full production after a vertical slice. According to Brads own timeline that is at least a year away at the absolute earliest.

Oloh_sl said:

There are several third parties that say they have a working MMO style network infrastructure set up and ready to go. They each cost like 1000 bucks per license (which is cheap for development studio, but not so cheap for me and my buddy fucking around). Our work around was to do this:

While we are prototyping the game, we essentially bifurcate the script on all objects into what we call "smart code" and "dumb code" using Unity's default networking "MyNetwork" testing. Said differently, we have an if statement on all relevent scripts that says, in essence, If I'm the player, use the smart code, if I'm not the player use the dumb code." This approach is not efficient at all, and is totally a "hack", but it allows us to effectively prototype for a MMO without developing the full server infrastructure on a budget.

The idea is that, once we are willing to spend some cash on the project, we will be able to, with some effort mind you, move the "smart code" to the server and the "dumb code" to the client. Also, we set up separate script to serve as the "database" where all it does is set variables. Also inefficient, but saves us the hassle of managing a DB with a static address. We can just each mess with the full code base, all within Unity at all times. We use git as version control, and we haven't had any issues at all with it.

And, we are just a couple of guys that thought "all mmos suck right now, lets make one." I am sure that a team that knows wtf they are doing could prototype combat for a few months in a way that doesn't screw them/force a complete re-write when they go full MMO.

Click to expand...

Which then brings us back to Pantheon. If they truly have been at this since September, how come the KS doesn't bear that out? Maybe a simple tech demo showcasing some of this?

Oloh_sl said:

There are several third parties that say they have a working MMO style network infrastructure set up and ready to go. They each cost like 1000 bucks per license (which is cheap for development studio, but not so cheap for me and my buddy fucking around). Our work around was to do this:

While we are prototyping the game, we essentially bifurcate the script on all objects into what we call "smart code" and "dumb code" using Unity's default networking "MyNetwork" testing. Said differently, we have an if statement on all relevent scripts that says, in essence, If I'm the player, use the smart code, if I'm not the player use the dumb code." This approach is not efficient at all, and is totally a "hack", but it allows us to effectively prototype for a MMO without developing the full server infrastructure on a budget.

The idea is that, once we are willing to spend some cash on the project, we will be able to, with some effort mind you, move the "smart code" to the server and the "dumb code" to the client. Also, we set up separate script to serve as the "database" where all it does is set variables. Also inefficient, but saves us the hassle of managing a DB with a static address. We can just each mess with the full code base, all within Unity at all times. We use git as version control, and we haven't had any issues at all with it.

And, we are just a couple of guys that thought "all mmos suck right now, lets make one." I am sure that a team that knows wtf they are doing could prototype combat for a few months in a way that doesn't screw them/force a complete re-write when they go full MMO.

Click to expand...

What you are doing is awesome, and a really clever way to do things so that they fit your needs. I am just trying to explain that this way is great for a quick prototype/hobby project, and doesn't scale even a little into anything remotely approaching a commercial product. Now, it really doesn't have to for you guys, so honestly it doesn't matter. Whatever allows you to work fast and efficient is the best choice in that situation so you did the right thing.

However, if you ever actually wanted to release it as a commercial project that scales to MMORPG size, even a small one, you would probably have to rewrite everything from scratch. I don't know for certain of course, because I haven't seen your project, I just know that what our server/erlang programmers do is arcane magic to me and even with my own hobby projects I am under no illusion that they would ever be commercially viable.

Seriously, I spent about 5 days trying to learn to script in Erlang. After attempting to push through the"there are no variables" and "there are no loops" I just said screw it, I'll leave this server programming/scripting crap to the experts.

Denaut_sl said:

That is a really well rounded team skill-set solid team. Maybe throw in an additional general purpose artist or 2. You actually don't need a DB Admin at all anymore. Modern document databases like Couch and Mongo make them pretty obsolete, your server programmers will generally handle "administering" the database. I use the term loosely because there is really no such thing for lots of NoSQL databases..

Click to expand...

I have to disagree here. If you are discussing a small server with few transactions, then yes you can have a programmer handling this. If you are discussing a large MMO with huge amounts of I/O and a need for cache tuning and with calls to packages then your database and most likely your MMO will be a trainwreck due to being CPU bound or hung on I/O wait, regardless of if you are using a hierarchical flat file NoSQL DB or a more traditional measure.

This is the problem with designers running technical projects. You are making the same mistake for your DB that Oloh seemed to make for his overall project.

Denaut_sl said:

Stuff.

Click to expand...

I agree with everything you are saying 100%. My goal is really just to tell folks that don't know about game development generally, or Unity specifically, that by any standard, nothing that the Pantheon team has said officially has been out in left field. It all makes sense.

If they came out and said "we don't need a hardcore networking guy with Unity" I would be the first to post "BS!" but they haven't said that, or even hinted at it. If it gets funded, my guess is that they will have a network guy on the team within a month and then backtrack and reconcile everything to its synced up the way it should be.

Gecko_sl said:

Oloh seemed to make for his overall project.

Click to expand...

In fairness, my long term projected scaleablility is probably around 6 users, so not that huge of a mistake .

@quaid With respect to the earlier discussion. Sony now forecasting a 1 billion loss and is exiting pcs and tvs. So no Smed to the rescue for pantheon and if eqnext is not huge soe is probably on the chopping block. Since I don't think pantheon funds I should probably consider p99 as emus may be the only old school option going forward.

@mkopec the trick is to give the AI something other than "kill healer" while still.making combat interesting. EQ and WOW have done that with taunts,summoning, boss aoe abilities, rampage, etc. I just think they need to pair those abilities with a more dynamic decision tree: one time in band camp the raid boss will summon, another time aoe damage, as opposed to completely scripted encounters.

Gecko_sl said:

I have to disagree here. If you are discussing a small server with few transactions, then yes you can have a programmer handling this. If you are discussing a large MMO with huge amounts of I/O and a need for cache tuning and with calls to packages then your database and most likely your MMO will be a trainwreck due to being CPU bound or hung on I/O wait, regardless of if you are using a hierarchical flat file NoSQL DB or a more traditional measure.

This is the problem with designers running technical projects. You are making the same mistake for your DB that Oloh seemed to make for his overall project.

Click to expand...

Functional Programming languages and NoSQL databases work much differently than traditional I/O, so we don't need (or have) an in-house DB guy. The data-centers that host our servers probably have them, but then you are sort of paying him through your hosting costs. It isn't quite the same.

Just saw this posted on the KS comments, lol...

"Momo
11 minutes ago
Is Brad around anymore? Not a good sign when the captain of the ship disappears when the ship shows signs of sinking."