Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Remlap

Where Are The Coders?

Recommended Posts

This is something that has been eating at me since the first day of the kick-starter was announced, if I go by the staff list on the kickstarter you only have what looks like 5/6 coders on your team and 6/7 artists. I understand that you want the art etc. to look good but the perceived lack of coders really worries me about the project.

 

1e3c0a4b034e993504c17a766a979038_origina

 

Of these 15 people, the only Coders I see are Mike McShaffry, Howard Smith, Thomas Sitch, Mark Bourland , Bill Dalton, & possibly Thomas Blair. (I apologize if Thomas Blair is coding as well, but from what I can tell his industry experience seems to be more in design than coding)

 

This worries me because games that have 3-4 times the coders as you have released games that have looked fantastic, but ran horribly due to poor optimization and net code. This is only exasperated by the use of Voxel Farm, which as far as I can tell has only been used in Ever Quest Next/Landmark which is laden with performance issues. I'm also fairly certain that EQN has a much bigger development team than Arts & Crafts (considering earlier this month they laid off quite a few people to "make the studio profitable")

 

Now with the stretch goals it's to hire more artists instead of getting additional programmers to help make sure the game runs smoothly. I understand why, in todays market graphics will sell a game more often than game-play/performance. However, if the game performs poorly it's not going to sell at all.

 

If Jon Carmack at ID had released doom all those years ago barely able to run at 15 fps then the whole state of the modern games industry would be completely different. Which is quite honestly a fair enough comparison, because currently the market doesn't have anything like Crowfall. Whether you like it or not people and investors in the industry are going to be looking at your project to see how viable an mmo that breaks the formula is.

 

I really like the idea of Crowfall, there hasn't been a game like this in a long time. It's why I've donated to the Kickstarter, and really want the game to succeed. I just perceive the lack of coders as a general obstacle to the overall success of the project. Obviously none of us have played the game and my concerns could be unfounded, however if they aren't this could very well be the biggest obstacle to Crowfall being a success.

Share this post


Link to post
Share on other sites

I believe there are a few factors to this:

  1. They are using a game engine that has a large community (Unity) and is backed by a large catalog of 3rd party toolsets, such as animation, effects, etc which make the entire process much faster than developing an engine from the ground up.
  2. The programming language is pretty accessible and robust (C#)
  3. Typically, having 2-4x as many artists as programmers/engineers is necessary due to the amount of assets that need to be generated. Basically: it doesn't matter how flawless the code is, if the amount of assets is limited or their quality sucks, that is the first thing that people will see. That being said, a game with bad code won't play well no matter how good the art assets are.

I'm sure the 800k target for the kickstarter will be used in part to pay the salary of employees and bring on additional talent as needed (perhaps contracted). This is in addition to the 2.3m that has been raised by other means, which I'm fairly certain they haven't burned through in its entirety yet. 

Edited by Korren

Share this post


Link to post
Share on other sites

In Gordan Walton we trust.

Honestly there is a good chance most of us don't run a company that makes computer games but we got one here that's been involved in the making of UO, SWG and SWTOR. On his Wiki it says he has personally developed over thirty games and has overseen the development of hundreds more.


YannMLq.png

Share this post


Link to post
Share on other sites

It really only takes a hand full of coders to get and keep a game going once you have the core setup. Graphics and Art are what really takes up your resources. You can have amazing code but if your game looks like poorly made socks no one will play it. Any major problems can always be outsourced or licensed (hence the GFX stretch goal).

 

"Here's an example of the kind of effects that we'll be able to add to the game, by licensing the right technology and putting it in the hands of a pro."

7b2d75929412d8f299e6ee35132cf5e6_origina
Edited by thenebrosity

OQa1xvz.png?1

lol ok.. I wonder if I'll still be able to steal directly from people's inventories.. hrmmm

;)Twitch - Twitter

RIP DOC GONZO

Share this post


Link to post
Share on other sites

I believe there are a few factors to this:

  1. They are using a game engine that has a large community (Unity) and is backed by a large catalog of 3rd party toolsets, such as animation, effects, etc which make the entire process much faster than developing an engine from the ground up.
  2. The programming language is pretty accessible and robust (C#)
  3. Typically, having 2-4x as many artists as programmers/engineers is necessary due to the amount of assets that need to be generated. Basically: it doesn't matter how flawless the code is, if the amount of assets is limited or their quality sucks, that is the first thing that people will see. That being said, a game with bad code won't play well no matter how good the art assets are.

I'm sure the 800k target for the kickstarter will be used in part to pay the salary of employees and bring on additional talent as needed (perhaps contracted). This is in addition to the 2.3m that has been raised by other means, which I'm fairly certain they haven't burned through in its entirety yet. 

 

Unity just raises more concerns honestly. In my experiences with the engine for clients I've worked for, it's robust enough but takes quite a bit of work to get it to get it to make various types of games. The net code is also primarily all going to have to be built from scratch, which is going to be a daunting task by itself. Obviously there are plugins available where you won't have to code all of the net-code but those then have to be worked in, along with creating the server. Just seems like a lot of work with only one dedicated server programmer, and it's arguably the most important.

 

I'd disagree on the artist part however, majority of the work has to go into the pve, environment crafting, etc. for traditional mmo's. Obviously you want the armors and things to look different, but that can be easily obtained by a bump map and a texture without having to remodel it. The terrain is confirmed to be randomly generated so the environment artists primarily just need to make a handful of things for each area instead of having to hand sculpt out every realm. This won't be the case for all of it, but there modeler seems talented enough to be able to do techniques to speed up the process.

 

I'm sure they will as well, I just wanted to bring attention to it. Because it's honestly a concern.

 

It really only takes a hand full of coders to get and keep a game going once you have the core setup. Graphics and Art are what really takes up your resources. You can have amazing code but if your game looks like poorly made socks no one will play it. Any major problems can always be outsourced or licensed (hence the GFX stretch goal).

 

Majority of the time spent coding is usually spent in developing the engines and tools needed to make other parts of the game, obviously with them using an engine and outside technology parts of that are going to be alleviated. However, that in no way changes that there are a lot of parts to this game. It's not just a handful of core systems, there are quite a few that all have to optimized to work together and quickly all while sending the information across the network. It's quite easy for such things to quickly bog down a project and require quite a bit of re-coding to undo or find whats causing too many packets to send etc.

Share this post


Link to post
Share on other sites

Unity just raises more concerns honestly. In my experiences with the engine for clients I've worked for, it's robust enough but takes quite a bit of work to get it to get it to make various types of games. The net code is also primarily all going to have to be built from scratch, which is going to be a daunting task by itself. Obviously there are plugins available where you won't have to code all of the net-code but those then have to be worked in, along with creating the server. Just seems like a lot of work with only one dedicated server programmer, and it's arguably the most important.

 

I'd disagree on the artist part however, majority of the work has to go into the pve, environment crafting, etc. for traditional mmo's. Obviously you want the armors and things to look different, but that can be easily obtained by a bump map and a texture without having to remodel it. The terrain is confirmed to be randomly generated so the environment artists primarily just need to make a handful of things for each area instead of having to hand sculpt out every realm. This won't be the case for all of it, but there modeler seems talented enough to be able to do techniques to speed up the process.

 

I'm sure they will as well, I just wanted to bring attention to it. Because it's honestly a concern.

 

 

Majority of the time spent coding is usually spent in developing the engines and tools needed to make other parts of the game, obviously with them using an engine and outside technology parts of that are going to be alleviated. However, that in no way changes that there are a lot of parts to this game. It's not just a handful of core systems, there are quite a few that all have to optimized to work together and quickly all while sending the information across the network. It's quite easy for such things to quickly bog down a project and require quite a bit of re-coding to undo or find whats causing too many packets to send etc.

 

More coders means the game will be completed and hotfixed faster.. that's always a good thing. I get where you're coming from... heck we already reached the kickstarter goal, how about the next stretch goal be to higher another coder or two? I'd be down for that.


OQa1xvz.png?1

lol ok.. I wonder if I'll still be able to steal directly from people's inventories.. hrmmm

;)Twitch - Twitter

RIP DOC GONZO

Share this post


Link to post
Share on other sites

A little off topic perhaps, but this concerns me:

 

I'm a programmer myself. I thought about their bold ideas to make voxel based MMO game, and came to conclusion, that it might be doable, but it's times more complex than with traditional static world MMO.

 

Warning: technical stuff ahead.

 

You can't just stream voxels over net to all clients. That's a giant socks load of data, no way modern network can handle this. What can be done?

 

First time character enters new word, the server sends random seed used for world generation. Client-side then generates that world itself (might take some time). Then, the server has to keep all influences, that caused world modification, in order of application. I. e., vectors and strength of force applied. The world healing technology is there to alleviate them, so most of old influences, which was fully healed, should be eliminated. But keep in mind, that intersecting influences must be processed in group, and taking the time passed into account. For example, the plain, that had hemisphere digged out, then partially healed, then another hemishpere digged out of it with small shift -- the result may differ depending on the stage of healing of first digging.

 

Client side has to receive the list of influences from the point of last logging off, and process them, applying to the world. Every change to terrain that happens during online, should also be passed as influences (vectors/force, etc.).

 

As you can see, single-player tech demo might not work at all over net, and will need considerable rework.

 

 

These are just theoretical speculations. It might work, but there might be some pitfalls. I'm sure, ACE programmers are competent people, and could handle all this mess. Just I've never heard anyone talking about the difference in complexity between static world and destructible voxel-based world, and that keeps me at unease.

Edited by Gremour

Share this post


Link to post
Share on other sites

I'm a programmer myself. I thought about their bold ideas to make voxel based MMO game, and came to conclusion, that it might be doable, but it's times more complex than with traditional static world MMO.

Landmark/EQ Next uses the same voxel library (VoxelFarm), and I haven't heard about data problems with that game. I don't have access to the source or documentation of VoxelFarm, but I suspect the library already solves such a compression problem. If not, I don't think it's that big of a problem.

 

The rest of my post is even more technical speculation, so take the following with a pinch of salt: You assume that they naively send each voxel "by itself", so the full amount of uncompressed data, which would probably be enormous. But when dealing with voxels, more often than not, a lot of voxels of the same type are clustered together (think about all the amount of air/nothing, the earth underneath a field, or even a castle wall entirely made of stone). So instead of stuffing a 643 volume of voxels into an array (like Minecraft does, or did when I last worked with it), you can use data structures like Octrees to compress the voxels. For example, a volume of 323 voxels, where every voxel has the same type, would then ideally only take x bytes rather than 32768 bytes. Such a degree of compression probably needs to be used on the client and server to store the voxels, as well, not only to send them around the network, because it really becomes impractical without compression when the voxels are really small (you don't have infinite RAM, and neither has the server). 

Share this post


Link to post
Share on other sites

Unity just raises more concerns honestly. In my experiences with the engine for clients I've worked for, it's robust enough but takes quite a bit of work to get it to get it to make various types of games. The net code is also primarily all going to have to be built from scratch, which is going to be a daunting task by itself. Obviously there are plugins available where you won't have to code all of the net-code but those then have to be worked in, along with creating the server. Just seems like a lot of work with only one dedicated server programmer, and it's arguably the most important.

 

I'd disagree on the artist part however, majority of the work has to go into the pve, environment crafting, etc. for traditional mmo's. Obviously you want the armors and things to look different, but that can be easily obtained by a bump map and a texture without having to remodel it. The terrain is confirmed to be randomly generated so the environment artists primarily just need to make a handful of things for each area instead of having to hand sculpt out every realm. This won't be the case for all of it, but there modeler seems talented enough to be able to do techniques to speed up the process.

 

I'm sure they will as well, I just wanted to bring attention to it. Because it's honestly a concern.

 

I would not argue about adding additional programmers. And yes, the Unity engine is good about hashing out the base idea quickly, but takes a fair amount of time to refine and optimize (like any engine). The topic of programmers has brought a question to mind though which is: how many players on screen is Crowfall aiming for? Since PvP is such a core focus of the game, I'd honestly be worried about incurring crippling lag between the destructible terrain, effects, and raw number of models running around.

Share this post


Link to post
Share on other sites

I would not argue about adding additional programmers. And yes, the Unity engine is good about hashing out the base idea quickly, but takes a fair amount of time to refine and optimize (like any engine). The topic of programmers has brought a question to mind though which is: how many players on screen is Crowfall aiming for? Since PvP is such a core focus of the game, I'd honestly be worried about incurring crippling lag between the destructible terrain, effects, and raw number of models running around.

 

Which is why it concerns me they only have about 5/6 programmers. It's not about how many players they are aiming for, but rather how many players are going to do. If you look at the most notable examples we have with Darkfall & Shadowbane there were hundreds of players in sieges, and generally performance issues to go with it. It's unavoidable to some degree when you have that many people in one area, but there are things you can do to avoid it.

 

But couple hundreds of players with state of the art particle systems run by Physx for example, and rather quickly lag becomes and issue even on rather robust systems. Even going by strategy game standards, at any time you have hundreds of units in the game. Which is why there is quite a bit of LOD's (level of detail, basically objects far away are low poly) for when you aren't rendering right next to units. Unity has LOD built in, but how will voxel farm integrate with that? This is why it worries me when we can only see a handful of programmers on the project.

Share this post


Link to post
Share on other sites

Which is why it concerns me they only have about 5/6 programmers. It's not about how many players they are aiming for, but rather how many players are going to do. If you look at the most notable examples we have with Darkfall & Shadowbane there were hundreds of players in sieges, and generally performance issues to go with it. It's unavoidable to some degree when you have that many people in one area, but there are things you can do to avoid it.

 

But couple hundreds of players with state of the art particle systems run by Physx for example, and rather quickly lag becomes and issue even on rather robust systems. Even going by strategy game standards, at any time you have hundreds of units in the game. Which is why there is quite a bit of LOD's (level of detail, basically objects far away are low poly) for when you aren't rendering right next to units. Unity has LOD built in, but how will voxel farm integrate with that? This is why it worries me when we can only see a handful of programmers on the project.

 

LOD will help, as it can apply to models, Physx, and FX, but I am still curious to know what should be attainable, even at a pre-alpha stage.

 

So yes, I can agree with you that it is somewhat concerning if they only have a few programmers. Hopefully there are a few more lurking in the background somewhere coding furiously. :)

Share this post


Link to post
Share on other sites

For what its worth, I worked on LEGO Universe as a 'engine engineer' (mostly server stuff) then switched over to running the live team (doing mostly operations / server stuff). We were running server farms that supported 120k CCU. We did tons of load tests with bots from EC2 pointing to our server farms.

 

Once the base network code was built and utilized (in our case raknet); we didn't have the need to touch the actual base networking layer on a frequent basis. I did a lot of the CSR / operations / devops type stuff that dealt with the network code; but i was ONE out of 20~ programmers and I was (for the most part) only working on the networking backend systems.

 

The teams were broken up kinda like this:

 

6 people for tools (this isnt needed AS MUCH since unity has its own toolset); they'll still have a programmer write custom tools here and there, but they have a nice toolset out of the gate.

3 Engine Programmers - we had a SLEW of 3rd party tech that was integrated, a lot to achieve what unity does already.

5 'Build System' Programmers - This was a team that oversaw the building of bricks and their systems.

2-3 UI Programmers - We used Actionscript on top, Unity solves a little bit of this. Can this be worrysome... sure, but most programmers can go in and do UI programming.

2-3 Gameplay programmers

1 Database Guy - All us programmers at the time weren't very versed in DB tech at the time, a lot of us that did work on that project end up becoming able to handle mostly anything for this position. 

1 TD

 

People rolled on and off at times, but that was the programming team supporting a very very large game. That team had a mix of junior - lead programmers on there. Yes CF only has 5-6 at the moment, I can see them going up to 10~; however, just note that those 5-6 people are VERY Senior level, that have done this before... several times. We had 3-4 people on our team of 20 that had worked on Auto Assault MMO previously. Everyone else was new and learning as we went (including myself)

 

After LU i went to work at Turbine, and the engineering team was even smaller. We had 6-7~ engineers for Infinite Crisis (+ UI engineers). (we had a internal tools team / engine team that I'm not counting); but that can be somewhat the same as using unity. 

 

5-6 Is what they can use to get done with what they've promised. Once more stretch goals are met, this perhaps will lead to hiring more engineers.

Share this post


Link to post
Share on other sites

 

It really only takes a hand full of coders to get and keep a game going once you have the core setup. Graphics and Art are what really takes up your resources. You can have amazing code but if your game looks like poorly made socks no one will play it. Any major problems can always be outsourced or licensed (hence the GFX stretch goal).

 

 

 

That is not true. You need both.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...