Should I make a new engine or just shut up?

ArChWind · February 2018

I'm going to give you some background here that is boring and maybe senseless but I want to assure you this is to bring the entire chips to the table kind of development effort that may complete this coming summer or maybe just go into the digital heavens. It entirely depends on what other input is and how it is received.

Back in September of last year I took a big step in my life. I retired from work. I wanted a project to occupy my time so I dug around and took out old Multiverse game engine. Multiverse game engine was released to the open source community in 2010. I thought maybe I would work on it as a hobby and keep myself busy with it. At first I had no wiki or any documentation so it was really a challenge to even attempt the project. Multiverse as the first engine I really had faith in as making a high scale MMO since it has the capability of clustered distributed computing.

I looked around the internet and found the project had literally vanished around 2012 (the answer why will come clear in a moment). I finally found a existing copy of the original board where there was a little information. I contacted one of the other developers through my account on the 'copy' board. I wanted to know what the status was and if there was wiki because the multiverse project had now gone into private development (I found this in Facebook) and the website was closed. The response was positive but here was no wiki in existence so I set out to make discovery of the existing system by reading code and working through the issues.

After a couple weeks I had created a working master server. The master server was the entire key to getting into a developers world but the master server seemed to only work if the game was made with a local server and it had to be run on same machine so it was by all intended purposes just capable of a single player game. It no longer worked as a MMO engine since at around the time in 2012 multiverse.net had closed down and the master server that the code wanted was no longer in existence. If you had a account at multiverse.net then your game world was no longer going to function. Closing down the master server literally killed the now open source projects.

I took this code and rewrote it to make the master work. I had to hardcode the client to not seek the real master but the new one I had on another machine. I literally rewrote the entire client side to no longer seek the real master or anything from multiverse.net. Now that this was working I had a master server, the game world server functional on a separate machine so I had indeed now a new MMO capable engine. I fired off email to this person and he was thrilled that it was now capable of MMO use. HOWEVER, that was the last communications I had with this person and was waiting for feedback on weather this new system was ever tested outside of my small setup.

I did mention that I was going to remove Axiom and make a new client with a different game engine which uses DX11 (now at DX12) I took the client and tools from Multiverse and started the process of rewriting the client. BUT, I hit a small snag! I have been programming for 39 years but I didn't know everything and I honestly was lost on how to make a scene manager. I did get contact from one of the guys that had worked on my MMO in the past. He was not really able to help on the issues at the moment since he had no clear understanding of how a scene manager was suppose to work. I was trying to keep the new client as close to Axioms design as possible but this lead to more or less a total confused state so I had to stop coding and start educating myself on scene managers.

Around late October I just kind of got busy with other things that were more profitable and kind of pushed the project off to the side.

I came back to the board here and see things are about the same as they were back in July and though maybe I would go ahead and work on the project again. Maybe look around and see if I have interest or maybe just take existing engines as they are and make a demo of my concepts with Axiom.

SO with that all said the questions I have are,

Would it be in anybody interest to see multiverse (a high saleable MMO engine) project come back from the dead?

Would it be of interest to just release this updated code with Axiom engine so developers can start making there own games with the intent to later release a new client?

Should I just keep this private and not make a new open source project?

Added poll.

iixviiiix · February 2018

Not in my field so i can't say much , no vote , but the "do it" always sound positive .

Amaranthar · February 2018

It can only benefit gaming.
Do it.

Then start working on a small scale project that you can sell cheaply and see where it goes.

I have an idea for you for a kiddy MMO. Simple. Might be a fun little thing for all ages.
I sent you a PM, in case you like it.

frostymug · February 2018

I vote do it.

Call me a hedonist, but I think any time you can spend doing something you enjoy is time well spent. If it has the potential to help others in some way, even better.

cameltosis · February 2018

I voted don't do it, but that's based on the presumption that you want to produce something useful for the genre rather than doing it purely for yourself.

If you're doing this project purely for yourself, because you find it interesting and enjoy doing it, then by all means crack on and have a blast.

If you're doing this project because you believe the genre needs a new engine then I would call it quits. Not because the genre is OK, far from it, we do need a decent MMO engine that can be used. I just doubt that you will be able to make one yourself. From what you've written, you're basing your engine on one that was developed 8+ years ago. That's not a good place to start. Second, you're just one person. You may be a great coder, but creating a game engine for MMOs requires some extremely specialised knowledge. Not only do you need to know programming, but you need an excellent knowledge of 4 dimensional physics, how to put that in a game loop, combined with extremely complicated networking code and server setup.

Now, maybe you have all that knowledge, I don't know. However, given that the majority of AAA MMOs don't have a good engine, certainly not one suited to MMOs, I have doubts. I just can't see how one person would be able to do a better job than massive teams with millions behind them. Not unless you have some revolutionary way of thinking that will transform the games engine market.

kitarad · February 2018

I don't know much about this area but are you confident enough to do this properly and not just sink a huge amount of time and come up with something barely passable. If you cannot do it then don't do it.

k61977 · February 2018

If you are doing it because it is a hobby and you enjoy it do it. If you are looking for extra income and not enjoying it then I would say don't. Overall no matter what you do if you work with an older engine you are bound to run into problems that you will have to think outside of the box to fix, like you did the master issue, but not having resources to call on to help you will make it very difficult down the road. That is why I say if it is a hobby project that you are doing for fun, enjoy!

ArChWind · February 2018

cameltosis said:

I voted don't do it, but that's based on the presumption that you want to produce something useful for the genre rather than doing it purely for yourself.

If you're doing this project purely for yourself, because you find it interesting and enjoy doing it, then by all means crack on and have a blast.

If you're doing this project because you believe the genre needs a new engine then I would call it quits. Not because the genre is OK, far from it, we do need a decent MMO engine that can be used. I just doubt that you will be able to make one yourself. From what you've written, you're basing your engine on one that was developed 8+ years ago. That's not a good place to start. Second, you're just one person. You may be a great coder, but creating a game engine for MMOs requires some extremely specialised knowledge. Not only do you need to know programming, but you need an excellent knowledge of 4 dimensional physics, how to put that in a game loop, combined with extremely complicated networking code and server setup.

Now, maybe you have all that knowledge, I don't know. However, given that the majority of AAA MMOs don't have a good engine, certainly not one suited to MMOs, I have doubts. I just can't see how one person would be able to do a better job than massive teams with millions behind them. Not unless you have some revolutionary way of thinking that will transform the games engine market.

Thank you for your honest input. You do bring up some interesting and quite frankly correct points.

First of all I have no intention of ever making another MMO game beyond a demonstration. I do not think that it would be successful enough make more than coffee money. I certainly would not go crowd funding and personal investment of money would never cover the cost. My interest has changed in that regard.

Yes, this engine is old and has some serious issues. I would not argue that ( I am more talking server replication which needs internal fixes). However MV server is the start of true streaming technology. By getting the basic core working It would be at open source level available to others to make improvement and expand its capabilities. Multiverse was a commercial project that ran out of funds and was released to open source with a MIT license.

There are several thing in the core that need completion which would require additional work. Probably beyond me. (Yes, you read that correctly and reason I would want open source to assist) The reason I say this is Multiverse engine is written in three different programming languages. The core engine is Java (do not confuse this with JavaScript) the interface to the server core is written in Python. The database is MySQL. The client is a mix of C# and Python.

Is Multiverse a good MMO engine?

The answer is the server is probably one of the best designed infrastructures available for free. It has high scalable cluster capabilities which needs further investment of code. The client is garbage it uses old technology and quite frankly probably the reason it failed as a open source project.

It failed twice so would it fail again?

Probably. This is the biggest concern and the reason for this thread. I don't see enough interest at this point to even consider it valid reason to invest in yet another project.

The bottom line here is, this is no toy. It is a very complex system to setup and get functional. The server requires a network administrator level of knowledge to set up. It is by no means a pushbutton instant make a MMO. Those using this as a basis for their projects are still required to write their own functions so it has a few strikes against it already.

ArChWind · February 2018

Amaranthar said:

It can only benefit gaming.
Do it.

Then start working on a small scale project that you can sell cheaply and see where it goes.

I have an idea for you for a kiddy MMO. Simple. Might be a fun little thing for all ages.
I sent you a PM, in case you like it.

Thank you for the offer but as I already stated I have no interest in making a MMO. I do what to clarify that this is open source and therefore non-profit.

cjmarsh · February 2018

I said "dont do it" but not for the reason I originally expected to after reading the op and comments. You're an experienced programmer so you understand the perils of writing a core client in anything other than C++ (with whatever other language for scripting). Frankly, if you're looking for an advanced hobby project I'd suggest rolling your own with Vulkan and C++ or better yet, using an existing engine like Unity, Unreal, or Xenko to create the client and focus on the backend server solution for the cloud.

Ridelynn · February 2018

cjmarsh said:

I said "dont do it" but not for the reason I originally expected to after reading the op and comments. You're an experienced programmer so you understand the perils of writing a core client in anything other than C++ (with whatever other language for scripting). Frankly, if you're looking for an advanced hobby project I'd suggest rolling your own with Vulkan and C++ or better yet, using an existing engine like Unity, Unreal, or Xenko to create the client and focus on the backend server solution for the cloud.

I mostly agree with cjmarsh. I think you gotta make the decision - are you trying to develop a game engine, or a game? Your post ~sounds~ like game engine there. So I'll continue forward under that assumption.

From what I remember of Multiverse - they wrapped up Firefly rights, which I was excited about. But their marketing angle was to be a game engine that was going to be easier for developers - not really to push games out. Honestly, it sounded like a shovelware platform to me at the time.

That was ... almost 15 years ago.

Today, I think Unreal, Unity, Cryengine/Lumberyard, Source (you could argue that's dead now too) and probably a handful of other less famous engines, have probably already achieved, and surpassed that original goal.

If you want to make a game engine - good on you. Absolutely nothing wrong with that, and I wish I were smart enough to do the same thing. But Multiverse doesn't hold any special attraction anymore, there's no brand recognition or anything it promised to do that hasn't already been done, and improved upon since. If you want to pick up Multiverse and bring that project back from the dead, that's cool too, I want to be clear I'm not dogging your ambition at all. But resurrecting MV just for the sake of making a game made from MV... nah, no interest in that.

Hashbrick · February 2018

I wouldn't bother, the engine itself is going to be your Achilles heel, it's old and techniques have been refined to make things run smoother and faster.

You might find a little community that will work on it with you but do you really have the time to direct, document and appraise it? I once you knew a smart engineer who said something of great intelligence. "As an open source developer, you are 10% of a programmer, 20% community adviser, and 70% an author. Write a book instead." If you love to program do your own thing, because an open source project will kill that hobby for you.

ArChWind · February 2018

I want to address the last several posts. Lets start at exactly the reasons for not using such engines as Lumberyard, Unity, Unreal and so on.

The misconception is these engines are for all intended purposes client engines. Actually on rewrite I was using Urho3D since it has a lot less overhead. Could I have used Unreal? Yes. Could I have used Unity? Yes. Could I have use Lumberyard? Yes. They do not and are not server end it is interfacing the client to the server which is the core of making a MMO.

Well Yeah, I see most here will not understand the difference.

Hashbrick · February 2018

Even as a server technology it is old, there is much better and more efficient ways of doing things to make it respond quicker and more reliable.

But it sounds like you are trying convince yourself not us, I think you are doing more soul searching than looking for an honest answer.

Ungood · February 2018

Do what you enjoy doing, screw what anyone else thinks.

cjmarsh · February 2018

ArChWind said:

I want to address the last several posts. Lets start at exactly the reasons for not using such engines as Lumberyard, Unity, Unreal and so on.

The misconception is these engines are for all intended purposes client engines. Actually on rewrite I was using Urho3D since it has a lot less overhead. Could I have used Unreal? Yes. Could I have used Unity? Yes. Could I have use Lumberyard? Yes. They do not and are not server end it is interfacing the client to the server which is the core of making a MMO.

Well Yeah, I see most here will not understand the difference.

Every MMO has a client and a server. I clearly stated that you should use one of the existing clients and focus on the custom backend. Maybe you should read more carefully before leaping to conclusions and asserting other's incompetence. You certainly wouldn't make a good open source manager if that's your modus operandi.

cameltosis · February 2018

ArChWind said:

cameltosis said:

I voted don't do it, but that's based on the presumption that you want to produce something useful for the genre rather than doing it purely for yourself.

If you're doing this project purely for yourself, because you find it interesting and enjoy doing it, then by all means crack on and have a blast.

If you're doing this project because you believe the genre needs a new engine then I would call it quits. Not because the genre is OK, far from it, we do need a decent MMO engine that can be used. I just doubt that you will be able to make one yourself. From what you've written, you're basing your engine on one that was developed 8+ years ago. That's not a good place to start. Second, you're just one person. You may be a great coder, but creating a game engine for MMOs requires some extremely specialised knowledge. Not only do you need to know programming, but you need an excellent knowledge of 4 dimensional physics, how to put that in a game loop, combined with extremely complicated networking code and server setup.

Now, maybe you have all that knowledge, I don't know. However, given that the majority of AAA MMOs don't have a good engine, certainly not one suited to MMOs, I have doubts. I just can't see how one person would be able to do a better job than massive teams with millions behind them. Not unless you have some revolutionary way of thinking that will transform the games engine market.

Thank you for your honest input. You do bring up some interesting and quite frankly correct points.

First of all I have no intention of ever making another MMO game beyond a demonstration. I do not think that it would be successful enough make more than coffee money. I certainly would not go crowd funding and personal investment of money would never cover the cost. My interest has changed in that regard.

Yes, this engine is old and has some serious issues. I would not argue that ( I am more talking server replication which needs internal fixes). However MV server is the start of true streaming technology. By getting the basic core working It would be at open source level available to others to make improvement and expand its capabilities. Multiverse was a commercial project that ran out of funds and was released to open source with a MIT license.

There are several thing in the core that need completion which would require additional work. Probably beyond me. (Yes, you read that correctly and reason I would want open source to assist) The reason I say this is Multiverse engine is written in three different programming languages. The core engine is Java (do not confuse this with JavaScript) the interface to the server core is written in Python. The database is MySQL. The client is a mix of C# and Python.

Is Multiverse a good MMO engine?

The answer is the server is probably one of the best designed infrastructures available for free. It has high scalable cluster capabilities which needs further investment of code. The client is garbage it uses old technology and quite frankly probably the reason it failed as a open source project.

It failed twice so would it fail again?

Probably. This is the biggest concern and the reason for this thread. I don't see enough interest at this point to even consider it valid reason to invest in yet another project.

The bottom line here is, this is no toy. It is a very complex system to setup and get functional. The server requires a network administrator level of knowledge to set up. It is by no means a pushbutton instant make a MMO. Those using this as a basis for their projects are still required to write their own functions so it has a few strikes against it already.

I'm still unclear on your motivations and the purpose.

From what you've written, it sounds like there is some core tech within the engine that you think is extremely good, better than what else is available. This is particularly related to scalable server technology and what you've called "true streaming technology".

I'm curious about the last one. Are you saying that the game would be 100% processed on the server and the resulting images just streamed to users? Something like OnLive but just for a single game? Or have I misunderstood?

I don't mean to be disparaging about your project, ultimately you're now retired and can do what the hell you want, so if you start this project you can always give up in 6 months time with no repercussions. So, you have nothing to lose except your own time, but if you enjoy it then that's time well spent.

That said, I'm a software engineer myself so I have some understanding of what you're attempting. I built a physics engine for my final year dissertation (it sucked) and have worked on a variety of projects, ranging from large scale military software down to real basic websites. So, whilst I don't know the specifics of building an engine like this, I have an appreciation for the scale and complexity of such a project. Only the most dedicated people tend to be able to finish a project like this, let alone finish it well, so my advice is usually not to bother, take on something more manageable but equally rewarding.

ArChWind · February 2018

cameltosis said:

I'm still unclear on your motivations and the purpose.

From what you've written, it sounds like there is some core tech within the engine that you think is extremely good, better than what else is available. This is particularly related to scalable server technology and what you've called "true streaming technology".

I'm curious about the last one. Are you saying that the game would be 100% processed on the server and the resulting images just streamed to users? Something like OnLive but just for a single game? Or have I misunderstood?

This is what makes it about the only redeeming thing to want to push it out. I think the core design as true streaming beings the world itself is never on the client. The client holds all the assets but the server holds the worlds description and passes this description based on location. Think of it as a database of the world. This allows the clients to have literally unlimited world space. (I did test the client on a 16 gig pixel world and it did work but the client had a memory leak. At the time this was done the client was still in commercial production)

The other thing that makes it valuable is the fact it can cluster and use RPC as its intercommunication.

Beyond that it will require considerable amount of work so it may not be worth the time and effort. As someone her posted I may be trying to convince myself this is probably true but before I completely wipe it off the box I thought I would put it up for discussion.

Mendel · February 2018

Another point, which many fail to consider.

Building an engine is one aspect, but maintenance and support are equally important. If you are looking to simply rework the engine as a personal learning experience, you're probably only concerned with the building aspect. That's fine. If you're involving third parties, such as licensing this as a product, then there's a commitment to the maintenance and support aspects you probably need to consider. Even if you intend to simply revamp the entire project as an updated open source project, there will be some degree of life-support and bug-fixing necessary.

My advice to you would be to make sure you are certain what you commit to. Be certain on a very firm answer to the question "How much time do you plan to devote to this project?". A means to divest yourself of future commitments may be the most important thing you do.

ArChWind · February 2018

Mendel said:

Another point, which many fail to consider.

Building an engine is one aspect, but maintenance and support are equally important. If you are looking to simply rework the engine as a personal learning experience, you're probably only concerned with the building aspect. That's fine. If you're involving third parties, such as licensing this as a product, then there's a commitment to the maintenance and support aspects you probably need to consider. Even if you intend to simply revamp the entire project as an updated open source project, there will be some degree of life-support and bug-fixing necessary.

My advice to you would be to make sure you are certain what you commit to. Be certain on a very firm answer to the question "How much time do you plan to devote to this project?". A means to divest yourself of future commitments may be the most important thing you do.

Actually this is what I am considering. Revamp it, put it out as open source, keep on expanding.

BUT this is also important that if it has no interest from anyone then doing this just adds more work to the project. If it goes south then I am left maintaining a project that probably should have not been released to begin with.

Sovrath · February 2018

It really depends on why you are doing this. If it's just for your own edification then sure continue. If you want more out of this then you are going to really have to do more research and see if what you are doing is going to yield the results you want.

Amathe · February 2018

Keep in mind that if you posted that you had just made your own aircraft in your garage, and asked if you should personally test it off a cliff, 80% of the people here would say yes.

Howdy, Stranger!

Should I make a new engine or just shut up?

Comments

Howdy, Stranger!

Quick Links

Should I make a new engine or just shut up?

Comments