71 replies to this topic

[Jasen]

Comassion, on 17 January 2013 - 12:04 PM, said:

My heart goes out to those programmers.

Guys, rewriting the game engine that your company bought and paid for is NOT FUN.

Considering I'd bet it was those very same programmers that recommended going with cryengine 3, without doing their research on it's netcode implementation, I don't feel bad for them at all.

Even in that other thread noone gave a single reason why CE3 was used over Source (which has amazing netcode), or Tech5 - which has amazing large world rendering and design abilities (would have been perfect for this game probably).

The choice actually reeks of gamers going... ooooh pretty... cryengine3, lets use that!!! - rather than developers going... what can this engine do for me, what am I going to have to do to it, and why should I use it over the other 5 commerically available awesome engines.

[SkyCake]

Vila deVere, on 17 January 2013 - 11:46 AM, said:

Maybe... but to keep introducing more light 'mechs into a game with a problem which gives lights a distinct advantage is perplexing. Why would you amplify a problem by emphasizing it?

because it comes at a time when significant work has been done to the movement portion of netcode... if you want to see if any improvements are being realized, releasing a fast mech is a good way to go about it... you arent going to notice the movement refinements as much in a slow mech... releasing a new mech gives players incentive to test these aspects of the game...

[Eddrick]

Vila deVere, on 17 January 2013 - 11:46 AM, said:

Maybe... but to keep introducing more light 'mechs into a game with a problem which gives lights a distinct advantage is perplexing. Why would you amplify a problem by emphasizing it?

A light was dew to come next. They have been going with a pattern from light to assault. Then light again.

[Protoculture]

Paul Inouye, on 17 January 2013 - 10:41 AM, said:

[redacted]I'll let you and the rest of the community in on something. Netcode is our #1 priority on the engineering side of things. The guys working on netcode have touched nothing but netcode. They have nothing to do with game balancing or adding new mechs/maps/items. Their main focus is netcode. i.e. Finding problems, re-writing a LOT of code that is part of the engine and making sure it doesn't break anything back up the chain. This is a VERY long process and not something that is addressed in a few days or weeks for that matter.

This is about to become my signature.

[Serapth]

Paul Inouye, on 17 January 2013 - 10:41 AM, said:

It is myself and David B. who do the game balancing. Both of us are not programmers and you DEFINITELY don't want us playing with netcode. I'll let you and the rest of the community in on something. Netcode is our #1 priority on the engineering side of things. The guys working on netcode have touched nothing but netcode. They have nothing to do with game balancing or adding new mechs/maps/items. Their main focus is netcode. i.e. Finding problems, re-writing a LOT of code that is part of the engine and making sure it doesn't break anything back up the chain. This is a VERY long process and not something that is addressed in a few days or weeks for that matter.

Hey Paul, if you are the guy in charge of balancing, does that extend to the matchmaker? If so, I would love an official word on balancing the number of premades on each team to make the game a bit more balanced and cut down on the pugstomping. Or does that fall into someone else's jurisdiction? Are such changes on the radar at PGI, because a great many of us are crying out for them, and god only knows how many newbies are simply rage quitting.

Edited by Serapth, 17 January 2013 - 12:49 PM.

[RickySpanish]

Jasen, on 17 January 2013 - 12:18 PM, said:

Considering I'd bet it was those very same programmers that recommended going with cryengine 3, without doing their research on it's netcode implementation, I don't feel bad for them at all.

Even in that other thread noone gave a single reason why CE3 was used over Source (which has amazing netcode), or Tech5 - which has amazing large world rendering and design abilities (would have been perfect for this game probably).

The choice actually reeks of gamers going... ooooh pretty... cryengine3, lets use that!!! - rather than developers going... what can this engine do for me, what am I going to have to do to it, and why should I use it over the other 5 commerically available awesome engines.

Here's a few reasons and educated guesses for the engines you listed:

* Id Tech 5 apparently isn't even being licensed outside of ZeniMax developers.
* Even if Id Tech 5 were available, the development pipeline has been described as quite different by Carmack (different in a good way obviously, but different nonetheless)
* Source is not particularly great at handling purely outdoor environments, as the engine still uses BSPs - a technology for partitioning 3D spaces for the purposes of managing the visibility of geometry - which were pioneered by the Quake engine.

Other reasons could include licensing costs, or the fact that no matter how you slice it, MWO is a very demanding multiplayer game - the developers really can't make many concessions with the physics engine without things looking obviously wrong (e.g. all limbs on a mech must be properly orientated).

[F lan Ker]

S!

Things like Paul posted could be put up in a blog or something and the thread locked out of comments. People could see where the things are going etc. But nevertheless good to hear things are being worked at

[DragonsFire]

Serapth, on 17 January 2013 - 12:48 PM, said:

Hey Paul, if you are the guy in charge of balancing, does that extend to the matchmaker? If so, I would love an official word on balancing the number of premades on each team to make the game a bit more balanced and cut down on the pugstomping. Or does that fall into someone else's jurisdiction? Are such changes on the radar at PGI, because a great many of us are crying out for them, and god only knows how many newbies are simply rage quitting.

I'm not sure if this falls directly into Paul's lap as it does technically fall under the balance category, but I believe he was referring to weapon and mech balance specifically.

That being said, I know some of the upcoming changes to Phase 3 of the MM are intended to address the balance issue there with the ELO rankings. It won't be perfect, but it's a step in the right direction, and likely will provide answers as to how it can be tweaked after it's been implemented. I don't know that they have a hard and fast date for it yet, but the last Dev post had mentioned implementation of it in January.

[MustrumRidcully]

3rdworld, on 17 January 2013 - 10:48 AM, said:

I wouldn't have accepted responsibility for the balancing in public.

He's taking one for the team.

"And they piled upon the Inoyues's white hump, the sum of all the rage and hate felt by their whole forum community. If their center torsos had been a auto-cannon, they would have shot their hearts upon it."

Edited by MustrumRidcully, 17 January 2013 - 01:06 PM.

[Serapth]

DragonsFire, on 17 January 2013 - 12:54 PM, said:

I'm not sure if this falls directly into Paul's lap as it does technically fall under the balance category, but I believe he was referring to weapon and mech balance specifically.

That being said, I know some of the upcoming changes to Phase 3 of the MM are intended to address the balance issue there with the ELO rankings. It won't be perfect, but it's a step in the right direction, and likely will provide answers as to how it can be tweaked after it's been implemented. I don't know that they have a hard and fast date for it yet, but the last Dev post had mentioned implementation of it in January.

Frankly that is the clarification/feedback we are looking for.

ELO is this big pending question mark, but we dont actually know what MM changes are coming. Many of us thought ELO was going to drop in the last patch, so once it didn't all the questions start to pop up again.

[Heeden]

Serapth, on 17 January 2013 - 12:48 PM, said:

Hey Paul, if you are the guy in charge of balancing, does that extend to the matchmaker? If so, I would love an official word on balancing the number of premades on each team to make the game a bit more balanced and cut down on the pugstomping. Or does that fall into someone else's jurisdiction? Are such changes on the radar at PGI, because a great many of us are crying out for them, and god only knows how many newbies are simply rage quitting.

I imagine designing such a system would fall under Paul's jurisdiction but implementing it would probably be down to the network coders. Hopefully things like this will be a lot simpler than their current tasks so once the lag issue is resolved we should see a more regular turnout of features.

[Serapth]

RickySpanish, on 17 January 2013 - 12:50 PM, said:

Here's a few reasons and educated guesses for the engines you listed:

* Id Tech 5 apparently isn't even being licensed outside of ZeniMax developers.
* Even if Id Tech 5 were available, the development pipeline has been described as quite different by Carmack (different in a good way obviously, but different nonetheless)
* Source is not particularly great at handling purely outdoor environments, as the engine still uses BSPs - a technology for partitioning 3D spaces for the purposes of managing the visibility of geometry - which were pioneered by the Quake engine.

Other reasons could include licensing costs, or the fact that no matter how you slice it, MWO is a very demanding multiplayer game - the developers really can't make many concessions with the physics engine without things looking obviously wrong (e.g. all limbs on a mech must be properly orientated).

That still leaves a number of engines ( http://www.gamefroms...e-Round-up.aspx ), the most obvious of which being Unreal, which powers a number of MMOs and hundreds of FPS type games. Or Gamebryo, which powered for Oblivion and about 100 more games, and a dozen others. CryEngine does seem to have it's ..... warts. Can't say from what I've seen ( I played around evaluating the download of CryEngine, Unity and UDK a while back and CryEngine certainly had the least polish and developer support. Certainly is pretty though.

Heeden, on 17 January 2013 - 01:07 PM, said:

I imagine designing such a system would fall under Paul's jurisdiction but implementing it would probably be down to the network coders. Hopefully things like this will be a lot simpler than their current tasks so once the lag issue is resolved we should see a more regular turnout of features.

Many of these things can be worked on in parallel to the lag issues. Some things obviously need to wait for net lag fix... its probably going to require a lot of the weapons get re-balanced once we can start shooting more accurately and the effect on the usage of streaks will certainly be interesting. Other changes should be able to happen in parallel since it is a dedicated team.

[DragonsFire]

Serapth, on 17 January 2013 - 01:07 PM, said:

Frankly that is the clarification/feedback we are looking for.

ELO is this big pending question mark, but we dont actually know what MM changes are coming. Many of us thought ELO was going to drop in the last patch, so once it didn't all the questions start to pop up again.

It's possible that the silence is due to a lack of solid answers or is motivated by marketing. I believe it was Garth that mentioned this the other day, but they don't give dates or answers on stuff that isn't solid, or if it's going to be part of a marketing effort, which is standard. I know there are features that we work on for addition to our code that customers really want, but because it's a marketing decision, we can't breathe a word of it until release.

[Jasen]

RickySpanish, on 17 January 2013 - 12:50 PM, said:

Here's a few reasons and educated guesses for the engines you listed:

* Id Tech 5 apparently isn't even being licensed outside of ZeniMax developers.
* Even if Id Tech 5 were available, the development pipeline has been described as quite different by Carmack (different in a good way obviously, but different nonetheless)
* Source is not particularly great at handling purely outdoor environments, as the engine still uses BSPs - a technology for partitioning 3D spaces for the purposes of managing the visibility of geometry - which were pioneered by the Quake engine.

Other reasons could include licensing costs, or the fact that no matter how you slice it, MWO is a very demanding multiplayer game - the developers really can't make many concessions with the physics engine without things looking obviously wrong (e.g. all limbs on a mech must be properly orientated).

Admitedly I never looked into actually licensing Tech5 so I had no idea that you couldn't... Doesn't seem to make sense that they would invest all that dev time into an engine they wouldn't even license...

I am familiar with source being pretty limited in large outdoor worlds, however... look at the maps... they aren't exactly large. Although I do see that limitation. (Funny enough, I used to make maps for Quake 1... very familiar with the BSP format and its limitations)


That said, I guess cost could have been the issue but I doubt CE3 was cheap...

I just hear this talk about adding networking features to CE3 and completely surprised that they would actually buy an engine they would have to modify so heavily.

[DragonsFire]

Jasen, on 17 January 2013 - 01:47 PM, said:

Admitedly I never looked into actually licensing Tech5 so I had no idea that you couldn't... Doesn't seem to make sense that they would invest all that dev time into an engine they wouldn't even license...

I am familiar with source being pretty limited in large outdoor worlds, however... look at the maps... they aren't exactly large. Although I do see that limitation. (Funny enough, I used to make maps for Quake 1... very familiar with the BSP format and its limitations)


That said, I guess cost could have been the issue but I doubt CE3 was cheap...

I just hear this talk about adding networking features to CE3 and completely surprised that they would actually buy an engine they would have to modify so heavily.

The maps that are currently out aren't massive, but they still have some good size to them. But there are much bigger maps coming down the pipe.

Testing internally is quite a bit different than real world deployment. I would wager their internal testing yielded the best results for what they were looking for while utilizing CE3 or perhaps it was the most manageable. When it hit deployment, issues not previously seen were brought to the fore, and they are now being addressed as a result. Again, not being in their shoes, I can't say that this is exactly how it happened, but being a developer as well, I would wager that it's not far off.

The adage 'The network was perfect, and then we let the users on it.' comes to mind.

[TheSteelRhino]

Paul,

Thank you for the commentary. I'm actually a Systems/Infrastructure Engineer so I get it when you say "you don't want me touching the netcode".

I actually appreciate the candor in telling us "netcode" is top priority. Some of us get it. There are some problems that need to be fixed first. Considering it's extremely difficult to target light mechs period I would imagine it's kind of hard to find that "balance" And of course if you get netcode fixed, then knockdowns can possibly be viable again, and people wouldn't be quite so dependent on streaks......etc.

Just keep us in the know best you can please, because a lot of us really do want this game to succeed. And I honestly think you'll have a great game when you get the bugs ironed out and can move on to things like Communicty Warfare.

Keep hammering away!

[PoLaR]

Thanks Paul,

I appreciate the great work you and you're team have done for us. Some people (including myself) just don't understand how long these things take. I look forward to seeing you're dream grow in the future.

Long live MWO!

:]

[Mxxpower]

Paul, balance is pretty good (except for ECM) ;P

[Solis Obscuri]

DragonsFire, on 17 January 2013 - 02:00 PM, said:

The maps that are currently out aren't massive, but they still have some good size to them. But there are much bigger maps coming down the pipe.

It's also fair to say that most of the maps are bigger than the currently avaialbe area for 8v8 play; we've seen some expansion in the Conquest mode, for example, and it's possible we may see further expansion with 12v12.

[DragonsFire]

Solis Obscuri, on 17 January 2013 - 03:05 PM, said:

It's also fair to say that most of the maps are bigger than the currently avaialbe area for 8v8 play; we've seen some expansion in the Conquest mode, for example, and it's possible we may see further expansion with 12v12.

This is a great point as well. Any time you bring you up the map, there is a cross hatched OB area that expands out beyond the regular map, as well as a larger area available for which the map could expand out to.