How much money would it take?

[Drakolis]

To fix this game? New Client? more staff? IF I won the lottery and say wanted to help this game out. how much would it take?

[Chaldor Itieles]

10 to 20 devs at 6000-8000 a month (all inc) for 12month = 720000 - 1920000 $

(hmm worst case)

[Garfonso]

One should aks what the Quake III Engine or Unreal (whatever) Engine costs..
Spending them a License would be a very good step...
But then they'd still need those 20 devs for over a year, I think.

[Steelclaw]

at 6000-8000 a month

Whoa! What kind of development do you need done? Are you hiring?

[MadDwarf]

WoW cost just under $20 million to produce and release. It was a very polished game at release but still needed a bit more work.

So I would imagine $20 to $25 million and 4 years would fix horizons. But if you were going to do that you might as well make a whole new game and call it Horizons 2.

How to fix this game ? There isn't a developer that will touch it. The code as I am sure EII has now found up is worse than a large bowl of spaghetti. Working with it to do any fixes or additions is a herculian task. Its just not something a developer who is looking to pad their resume will really want to touch unless the money is really really good.

[Deth]

One should aks what the Quake III Engine or Unreal (whatever) Engine costs..
Spending them a License would be a very good step...
But then they'd still need those 20 devs for over a year, I think.

They thought about engines like those, but they would not make the world seamless so dropped those ideas.

[LaughingOtter]

Since Tulga had 17 people on staff, plus contractors, interns, and players who were volunteering their time, I'd say a staff of 30, and ballpark 3 million per year for headcount, just to make sure I could offer wages that are competative with the big boys.

Hosting and bandwidth - 10-20k per month, depending on usage and provider costs. Figure another quarter of a million dollars.

Software licensing: Depends on who you talk to. Going with open-source solutions brings this cost down, but might result in a longer development cycle and higher costs over the long run. This is why I would budget a lot for headcount; that way I can hire people who, unlike myself, know something about the MMO industry. ;-)

Add another two million for unforseen expenses, marketing, office costs, travel, and lawyers.

So, DID you win, Drakolis? :-D

[Spartikus]

Wow a lot of misinfo here

@Maddwarf: Actually wow cost just over 50 mill to produce.

@OP: Seeing as TulgAE invested just under 30 million dollars over the last 5 years and this is where the game is at.. I don't think the game can be fixed. I think it needs a ground up rewrite.

The reality is this. You can buy prefab MMO engines today that allow you to prouce block buster MMOs in under 24 months with less than 20 people. Start to finish dev costs including the engine less than 5 mill. NC soft has produced many asian based MMOs for less than 5 mill.

[LaughingOtter]

Wow a lot of misinfo here

@OP: Seeing as TulgAE invested just under 30 million dollars over the last 5 years

More like 15 million. Roughly six million under David Allen, another six million or so under James Jones, and three million under David Bowman. I do recall an occasion in IRC where David stated that he was giving away an 18 million-dollar game, in reference to the free download & trial. Given that these are rough numbers and that the person I got them from implied (but did not state conclusivly) that they were pre-launch development costs, I'd say 2-3 million in post-launch costs would be believable and get to the 18 million David mentioned.

There are indeed some very nice MMO solutions out there, but I think the real cost saver would be having a single, coherent, reasonable set of design goals for the project and by making sure that a given solution really IS a viable solution before committing to it.

[Spartikus]

LO: dont let's you and me get into another numbers "discussion" ok

Ima not going to beat up on you for your numbers but suffice to say I'm very confidant in the Jones/Bowman era spending just under 20 mill. I will, however, agree to disagree on the exact figure

Either way though, the real question is what would it cost to "Fix" or bring it up to current triple A standards. My opinion is that nothing short of a ground up rewrite of the game using a proven technology backend would do it. I know the network layer is a writeoff and I know the asset system (object management) is a writeoff as well as the database layer. The patcher/login system is pretty bung'd up as well although probably salvagable. The only thing I am not sure about is the rendering pipeline. If it's salvageable thats a good chunk of work there. Yeah the current game has real issues with Framrates etc but a ton of those current problems are related directly to the datastream and asset library. (some of the 3D objects are 10s of thousands of polys which is unheard of in an MMO and should never have been put ingame.) I think in one interview David said he was surprised to find out that one of the bone piles they have ingame was over 30 thousand polys.. To put that in perspective, alot of MMOs have poly budgets of less than 1000 for most 3D ojects (weapons, armor, swords, buildings etc) and their highest poly count game assets are the player models and those are budgetted at under 10k generally.

I was a big critic of the Horizons rendering pipeline at first untill I saw the underlying problems with the asset library and datastream.

If I was at the wheel I'd start with a great engine that works and go from there. but.. sadly.. it's not up to me

PS: This post is not meant to reflect negatively on EII or Horizons. It's just some basic facts about the Horizons performance and stability issues that are well known and documented in a hundred different places. Please don't take it as anything else.

[Kulamata]

I'd start by seducing two or three good headhunters to find a superb project/program manager and a good advanced development group. I'd lure Tango and Smeglor and Amon and ? back with untold riches for part of the team. I'd allow ? months to develop a proposal and get a top level design map written, and then the more detailed full design document. Then start hiring implementors. I'd guess (and I am guessing) that the Evolution Engine would be worth leasing and that the client would need to be completely rewritten. Some of the tools I think were good, but they need to be cleaned up, made easier to use, and of course, documented. I'd shamelessly bribe the Beagle Boy and have a quest group. Another group would be cleaning up the models that have silly poly counts. There would be a QA manager, who reported only to God, with dedicated tester(s). I'd try to find some of the originals, who thought so expansively.

And I'd make a small fortune;




out of a large one.

[sethvir]

the game needed to be recoded during beta... and no I am not bitter or being sarcastic. I have been in an out of this game since beta and the same problems that existed in beta still exist.

If your going to sink $50m into this game... take the craft, dragons, and just rebuild the engine... the mechanics are ok.. but the graphics engine always was horrid.

[Drakolis]

Ok, when I win 100 million (after taxes) I will sink 50 million into this game.

[Garfonso]

How uptodate is that info about the high polygon models?

I heard DB talking in IRC about people going through all the assests and reducing polygon count whereever possible.
I think I even read an update note on that.

But after all I'm not too sure about that anymore, so I figured, I just might ask

[MadDwarf]

Ok, when I win 100 million (after taxes) I will sink 50 million into this game.

this game wouldnt be worth that kind of money

With $50 million you could start your own studio and hire a team of crack developers. In 2 to 3 years you could easily create an MMO of your own and still have $10 million to spend on the wife and kids .

[Spartikus]

Yeah Garfonso, it was tabled a couple times that they would have someone redo the crazy resources. It's another one of those things that never got done. I think a couple of the 30k bone piles were replaced with a different object but I believe they still exist in a few places in the game. As to the rest of the resources, the huge poly count items are still in game and being used. Least ways they were as of 6 months ago.

[Solitaire]

Cleaning up the assets was an ongoing project that the Tulga devs were working on. Just about every time we patched there would be new versions of existing world assets downloaded. Don't know how much time was being devoted to them, but it was part of the optimizations they were working on to try to coax as much performance out of the existing client as they could.

I wish that the wireframe display mode had not been disabled in the client. At the time, I was intent on learning 3D modeling and I used to turn it on to see how things were made. Even knowing as little as I did then, I learned the difference between a well optomized model and a poorly done one. And there were a lot of the poor ones. It was also funny to see some of the forgotten things here and there that were hidden from view, like a dead body entombed in some rocks NE of Last Stand where the arbotus used to be.

[LaughingOtter]

Fixing up the models was indeed an ongoing project, along with fixing the occlusion data for them and a whole bunch of other stuff. Just because a play-by-play wasn't posted doesn't mean it was being ignored.

[Tzael]

Personally I'd be leaving the database and simulation layers alone if I were in a position to authorise further development for Horizons.

My focus would be turned to the client: rendering engine and audio.

We all know of inherent problems with the audio in Horizons, number of channels available for playback, audio distortion etc. The Miles Audio System is actually pretty good and is still used in a lot of high profile games today so I'm fairly confident that the problems we know of are inherent with the implementation and not the main audio component itself.

Redoing the rendering engine is a prett hefty piece of work. All original assets such as models and textures would need to be exported to compatible formats once a new rendering engine has been chosen. Using an existing rendering engine would save on development time but would require the licensing of additional technology. Replacing the rendering engine would be time-consuming and I would be very interested if EI announce that they will continue development of the new client that Tulga were focusing a lot of their development time on. It makes perfect sense to look at how developers who know the product in great detail have approached a ground-up rewrite of the rendering engine implementation, as a project of this nature would have involved cleaning up a lot of the existing code.

Ultimately players can live with a bit of server lag, take WoW as an example. What players can't deal with is poor client performance and it is this in my opinion that has prevented Horizons attaining the growth that it deserves.

[LaughingOtter]

The database layer apparantly has a significant issue in the way it stores object information. I am not a database expert, but as I understand it, it was initially set up so that every single object - right down to your bonus salt rock - is stored as a discrete entry with all the common information about that type of object duplicated for each entry. Normally, there would be one entry for "sandstone brick" that had all the information defining a sandstone brick and all other sandstone bricks would reference that entry. Not in HZ, for some reason.

The result is, as I understand it, every node that pops takes up about 1.5 mb of space in the DB. Same with monsters, NPCs, whatever. There may be a reason it was set up that way, but nobody I've talked to knows what it might be, and the person who set it up is long gone.

That excess data may be part of the data stream bloat Spartikus keeps mentioning, it may not; I don't know. But if I were doing a rebuild, that's way high on my list of fixes. As I understand it, though, fixing it is a massive undertaking and involves a lot of recoding - client, server, and simulation alike.