Wow the hero engine, you got to be kidding listing that engine.  what a hunk of junk. 

Originally posted by joshuahalls

This is back from 2011 mind you, but was during the 30% royalty bit.  I don't recall anything changing since then beyond the seat counts as that would be a rather drastic change to EULA and just cannot be snuck in.

  * We use Visa PlaySpan to provide 84 billing methods in 122 countries. They handle all the money. They handle the initial purchase of the subscription, and then using the same systems will handle charging your players. After financial transaction fees (credit card, mobile, facebook coins, paypal) you get 70 percent, we get 30 percent, and we have to pay all the costs for licenses and servers and bandwidth, so it's more like we get 12 percent, which all of a sudden feels like we are working way too hard for only 12 percent. 

Obviously he is making numbers up there, but they pick up the infrastructure bandwidth cost and that is a rather variable number based on design choices and downloading of client data.

Yes Josh that was stated but did you read the full contract? I remember this part and futher down is stil stated that you pay the bandwith cost. If you plan to make a free game and absorbe the bandwith cost then go right ahead.

Several paragraphs later it again states a bandwith cost but no where does it state 12% BUT now tack on 12% +30% nd that is 42% of your revnue of which 52% has to pay the other developers and yourself.

Josh, the only one making money in this is HE.

Like I said you are one of the founders and only have to pay 10% and 12% then that is 22% which leaves you with a lot bigger profit margin.

Originally posted by VassagoMael

Originally posted by joshuahalls

This is back from 2011 mind you, but was during the 30% royalty bit.  I don't recall anything changing since then beyond the seat counts as that would be a rather drastic change to EULA and just cannot be snuck in.

  * We use Visa PlaySpan to provide 84 billing methods in 122 countries. They handle all the money. They handle the initial purchase of the subscription, and then using the same systems will handle charging your players. After financial transaction fees (credit card, mobile, facebook coins, paypal) you get 70 percent, we get 30 percent, and we have to pay all the costs for licenses and servers and bandwidth, so it's more like we get 12 percent, which all of a sudden feels like we are working way too hard for only 12 percent. 

Obviously he is making numbers up there, but they pick up the infrastructure bandwidth cost and that is a rather variable number based on design choices and downloading of client data.

So is is that why so many games have such disasterous launches? They are giving power to initial bandwidth to a 3rd party company that gives them as little as possible so then there are horrible queues?

There are a lot of reasons there are disasterous launches from releasing to early to not allocating enough servers to handle the load to not enough bandwidth to poor design to a full moon.  Mostly from companies wanting trying to keep up with the hype they built and 1-2 million people showing up to destroy their servers.  

Originally posted by Yamota
Very interesting post. I am a Java developer and always wanted to build some simple persistant multiplayer game but the problem is that my skills are limited to Java and I have no 3D skills at all. And only Java engine I found which seems feasible is jMonkey but it has no multiplayer support :/

Most of Multiverse server is primarly Java based. Some of it is Perl and python but the core engine is java/python. What I like about this design is it is entirely portable so a Sun System is a valid option. In other words the server can run on a main frame.

 Maybe you may find it useable. Much of the client is Python but if you have good skills you may be able to port it to Java.

---

Originally posted by ArChWind

---

Originally posted by Nasa To the OP, I would like to know your view of:

• Unreal3 (used in TERA) Cryengine3 (used in Archeage)

  And thanks for the nice read  

---

The reason these engines never made it to my list.

Cry engine indie at the time I tested it was not very stable. I had problems even getting a basic block into it without a CDT. If one bought the full commercial package I am sure this would work for a MMORPG but the requirement for this list is low cost and good performance engine. Cry fails on low cost for a commercial game.

For a zone based indie game Unreal engine is fantastic. Only testing I had was UDK so I did not have the source to evaluate beyond that. With UDK it would be possible to make a (somewhat MMO) reason I say this is because each server would be very limited on the number of CCU. Some of the issues with both engines is the fact they are not built to stream areas out of the box. Similar to HE they tend to stutter on the boundary crossing when using the seamless design.

I did hear that Unreal had a atlas program but I haven't seen it in action.

Unity has a atlas and background streaming in the professional package so if either of these engines used the paged method I would expect the game to run a bit more smoothly.

The nice thing about Esenthel and BigWorlds and for that matter Multiverse is they are streaming engines. They are designed form the ground up for massive sized worlds. (although the bug in Multiverse needs fixed becasue it loads the terrain data but does not unload it which means you run out of memory). They break the data up such a way that background loading and unloading are in their own threads so you never get a big jump in FPS unless you screw with making models too many vertices or just plain to big for AOI.

 

---

Both UE3 and CE3 support streaming...

It's been a staple of UE3 ever since the engine was announced. A lot of games emply it, from Mass Effect to Tera.

CE3 does streaming differently but offers it nonetheless. It is however a bit limited in size as it ultimately doesn't really /stream/, limited by the floating point precision (10^63 for 64 bit) for geometry.

Originally posted by ArChWind

Originally posted by joshuahalls

This is back from 2011 mind you, but was during the 30% royalty bit.  I don't recall anything changing since then beyond the seat counts as that would be a rather drastic change to EULA and just cannot be snuck in.

  * We use Visa PlaySpan to provide 84 billing methods in 122 countries. They handle all the money. They handle the initial purchase of the subscription, and then using the same systems will handle charging your players. After financial transaction fees (credit card, mobile, facebook coins, paypal) you get 70 percent, we get 30 percent, and we have to pay all the costs for licenses and servers and bandwidth, so it's more like we get 12 percent, which all of a sudden feels like we are working way too hard for only 12 percent. 

Obviously he is making numbers up there, but they pick up the infrastructure bandwidth cost and that is a rather variable number based on design choices and downloading of client data.

Yes Josh that was stated but did you read the full contract? I remember this part and futher down is stil stated that you pay the bandwith cost. If you plan to make a free game and absorbe the bandwith cost then go right ahead.

Several paragraphs later it again states a bandwith cost but no where does it state 12% BUT now tack on 12% +30% nd that is 42% of your revnue of which 52% has to pay the other developers and yourself.

Josh, the only one making money in this is HE.

Like I said you are one of the founders and only have to pay 10% and 12% then that is 22% which leaves you with a lot bigger profit margin.

 

Yes and had talked to them about it as well when it was released as we were under another contract.  I haven't seen anything added about paying bandwidth/infrastructure costs and reread the license agreement as well (the latest one). I could be mistaken, but not seeing anything that would say otherwise.  Either way, a lot of great options out there now adays and make sure to get solid feedback and make sure to get your hands dirty with multiple products to get a solid evaluation of what is out there.

Originally posted by Arglebargle
Threads like this should be required reading for anyone who starts to say, "I don't understand why developers  don't do this simple thing (X) that I want."

It's like building a space shuttle without blueprints, and finding out halfway through completion that your boss wants the interior rearranged and people don't really like the seat covers.

At least that's how it seems to me.

Originally posted by ArChWind

Originally posted by joshuahalls

There technically isn't an "MMO Engine" there is a graphics renderer and then a collection of many tools and systems to turn a multiplayer game into an expansive persistant online game.  People sometimes tend to think this game or that game failed so it must be Hero Engine, it must be Unity, it must be Cryengine, etc.  All 3 of those engines used properly and more importantly used toward their strengths can make very successful games.  If you want 400 people in one spot, you have design your game content from the ground up to support that style of play otherwise you are not going to get that.  A few distinct choices in design and that number gets cut down drastically and is design choice more so than an engine limitation.  Not to say that some engines might be better at it from a technologial standpoint, but they aren't as far off as most people want to think.

Obviously, a little bias toward Hero Engine as that was the path we chose and would chose it again even with all the alternatives out there now (would probably still evaluate some of the new ones and didn't get to dig into esenthel much as it was fledging at the time), but at the end of the day building an MMO is an insanely complex project and building the toolset from scratch even more so.  Having a something like BigWorld or Hero to take care of some of the more complex parts is helpful and you can spend your time on development versus building the toolset.

As far as Hero goes the 30% is a tad steep, but as mentioned it does include server hosting/support fees as well so those costs are inclusive as part of that so it brings that number down a bit as you would normally have to pony up that money for the instrastructure.  We design our world in a grid design as their seamless support works fairly well for that, but we do it in a slightly different way than most people are doing it as we have scripts that automate it making the process fairly quick and can theoretically keep expanding the size and scope of the world out as far as we needed.  BigWorld when we evaluated it was restricted to 8-10km in a single heightmap even though it was chunked.  That might be old news now of course as everything is constantly moving forward and updating.  From a collaborative standpoint I haven't seen anything like it out there right now and was one of the major aspects that allows for rapid development as there is no nightly builds or fighting with repos to deal with.  Both were pretty solid products though, but not something you just jump into and expect to have a shovel made MMO ready to go.

That being said, the options available now compared to 4 or 5 years ago is night and day and it will be interesting to see what is available in another 4 or 5 years.

You're one of the lucky ones Josh, you bought the engine at a higher price so you are not going to get hit very hard 10%  but for the people comming into HE the 30% pluse the bandwidth make it lower on my listing. It is a fantastic engine but I think end costs are a bit to steep and for my project it would not coorporate. :).

30% isn't that bad.  They handle payment, distribution, and server hosting.  those three combined aren't exactly free under normal circumstances.  For a game that's hoping to attract maybe 50k or 60k people, for example, it makes very little sense to go out and spend a bunch of money on servers and a datacenter location, as well as find a payment portal who can handle payment for you (at a cost of course).

Hero Cloud just isn't targeted at games larger than that.  That's what they have the actual Hero source code for, where you can handle all of your own stuff and modify the server however you like, and not get hit with 30% fees.  Of course the upfront cost is much higher, like any other licensed engine.

Originally posted by joshuahalls

This is back from 2011 mind you, but was during the 30% royalty bit.  I don't recall anything changing since then beyond the seat counts as that would be a rather drastic change to EULA and just cannot be snuck in.

  * We use Visa PlaySpan to provide 84 billing methods in 122 countries. They handle all the money. They handle the initial purchase of the subscription, and then using the same systems will handle charging your players. After financial transaction fees (credit card, mobile, facebook coins, paypal) you get 70 percent, we get 30 percent, and we have to pay all the costs for licenses and servers and bandwidth, so it's more like we get 12 percent, which all of a sudden feels like we are working way too hard for only 12 percent. 

Obviously he is making numbers up there, but they pick up the infrastructure bandwidth cost and that is a rather variable number based on design choices and downloading of client data.

Plus bandwidth is nowhere even close to $100 for a GB.  It isn't even that for a TB.  Maybe if you are talking 1 Gbit unmetered connection, but you would be consuming a lot of bandwidth at that point.

Yeah the price was way high back in 2011 I would of never considered it back then. They have changed lots since then including the price plans. If you look at the prices back in 2011 wasn't a seat $300 - $1000 ? 

Originally posted by erictlewis

Wow the hero engine, you got to be kidding listing that engine.  what a hunk of junk. 

 

Just because SWTOR was a hunk of junk does not mean the engine is :) just saying form your post history.

Also the fact The Repopulation is being made on HE with all them features the engine can not be that limited to make it a hunk of junk.

The Cryengine is the best imo. Entropia has been useing it for a while and it is impressive what youcan do with it.

Originally posted by Yamota
Very interesting post. I am a Java developer and always wanted to build some simple persistant multiplayer game but the problem is that my skills are limited to Java and I have no 3D skills at all. And only Java engine I found which seems feasible is jMonkey but it has no multiplayer support :/

http://www.lwjgl.org/

Minecraft was made with it, you would need to learn some opengl, networking, and DB (unless you roll your own flat file).

Nice list, really i dont think any engine is "best for mmorpgs", what should be the case is you find the engine that best fits your design, a key point is to understand a projects deveopment requirements and ensure the choosen engine will be able to achive the design goals.

Me personally i prefer engines that are just wrappers really, its hard as a programmer trust "blackboxes" where you cant get into the source code or cant see what its doing, and could even end up writting alot of unreusable code,  yeah ok that means i have to roll alot of code, but if you follow good coding principles, you can make reusable code, and switch between graphic wrappers like no tomorrow, c\c++ will never die.

Originally posted by erictlewis

Wow the hero engine, you got to be kidding listing that engine.  what a hunk of junk. 

 

I take it from that statement that you are an experienced game developer who has extensive knowlege about the HE.

Could you please list your credentials and the games you have developed??

Originally posted by PyrateLV

Originally posted by erictlewis

Wow the hero engine, you got to be kidding listing that engine.  what a hunk of junk. 

 

I take it from that statement that you are an experienced game developer who has extensive knowlege about the HE.

Could you please list your credentials and the games you have developed??

Yeah I'm waiting with baited breath for this reply,I would like to know from this intellectual insight his opinion on the transition to Hero Engine 2.0 or does he actually realise this I really doubt it.

Btw brilliant post OP. 

Originally posted by ArChWind

Originally posted by Yamota
Very interesting post. I am a Java developer and always wanted to build some simple persistant multiplayer game but the problem is that my skills are limited to Java and I have no 3D skills at all. And only Java engine I found which seems feasible is jMonkey but it has no multiplayer support :/

Most of Multiverse server is primarly Java based. Some of it is Perl and python but the core engine is java/python. What I like about this design is it is entirely portable so a Sun System is a valid option. In other words the server can run on a main frame.

 Maybe you may find it useable. Much of the client is Python but if you have good skills you may be able to port it to Java.

Well I looked at the Esenthel engine and it looks really interesting and since I just want to create a simple online game, and not a full fledged MMORPG, the limitations on network code is fine.

However I am not sure how hard it is to learn C++ if you have 5 years of professional Java experience. Do you know?

Also, how hard is it to import 3D models to use? Since I dont have any 3D modeling skills I would look at the many free models out there and possibly buy some simple one's as well.

Holy crap!!!  Finally a MMORPG.com user who actually knows his stuff about graphics and engines!  You sir are a gem, and I wish more users would brush up on their graphic engine knowledge.  Then people like me who are anal about their quality/performance ratio wouldn't have to spend 12+ hours tweaking/testing graphic options and .ini files on games like TERA and Skyrim :(

I think tweaking .ini files is fun though, so meh.

Originally posted by mmoski

Originally posted by Yamota
Very interesting post. I am a Java developer and always wanted to build some simple persistant multiplayer game but the problem is that my skills are limited to Java and I have no 3D skills at all. And only Java engine I found which seems feasible is jMonkey but it has no multiplayer support :/

http://www.lwjgl.org/

Minecraft was made with it, you would need to learn some opengl, networking, and DB (unless you roll your own flat file).

Nice list, really i dont think any engine is "best for mmorpgs", what should be the case is you find the engine that best fits your design, a key point is to understand a projects deveopment requirements and ensure the choosen engine will be able to achive the design goals.

Me personally i prefer engines that are just wrappers really, its hard as a programmer trust "blackboxes" where you cant get into the source code or cant see what its doing, and could even end up writting alot of unreusable code,  yeah ok that means i have to roll alot of code, but if you follow good coding principles, you can make reusable code, and switch between graphic wrappers like no tomorrow, c\c++ will never die.

Read some of the tutorial and it looks pretty straight forward. However it seems that this is just an interface to opengl and as such does not contain a physics engine and so on?

In any case I am going to try it out and see if I can get a simple space 3D object moving around using the mouse.

Thanks!