Is the travel system in this game and all the instancing temporary?

you are focused on the wrong use of proc gen.

yes, they do in fact use it for terrain already. go watch Chris make the landscape tutorial.

that height map you see? it ain't random, it is called procedural noise by a guy named Ken Perlin. Yes even all the otherfancy variants, he invented lattice noise. Google is your friend.

But that ain't the point.

proc gen can be used to boot strap areas with stuff, that artists and designers can then go back and tweak. they might already even do that. proc gen is simply an algorithmic way to generate content.

go look at dwarf fortress for a preview of what no mans sky is doing in 3d really. DF does not just do landscaping, but histories are procedural, characters, monsters, etc....

left for desd used proc gen in the AI unit. plenty of other examples in video games, it's not a forefinger concept by any stretch.

yes, they should plan and design for its use, but nothing is stopping them from doing that.

 

Time is stopping them doing that. Scheduling is stopping them from doing that. The fact they're not done with assets or still brainstorming assets that need to be plugged into the procedural generating system to work with one another is stopping them from doing that. The fact they have to pump out a game with those limitations at a reasonable time while showing constant progress towards the backers because they chose to be so open is stopping them from doing that.

The AI or the "Combat Director" in L4D just spawned mindless enemies that populated the area for your guns to mow down, not give great dialogue or give them proper senses that are limited by their terrain for you to sneak around in a randomly generated world instead of populating a static one or NPCs given work schedules. No Man's Sky being the only proper use of procedural generation in it's scope probably won't have anything involving the deep intricacies of the RPG SotA is trying to make. Yeah it'll have stats and enemies that'll scale and maybe varients of enemies, but you'll be able to see the pattern within a few planets (maybe, game isn't out obviously and I'd love to be proven wrong). There are plenty of reasons why they can't do procedural generation. It WOULD be awesome if they could later down the road and it is possible in the next episode once they have their core features locked down but they don't have it yet. If you could, point me to a game that has a completely 3D traversable world (not isometric 3D) where the AI interacts with the environment in a procedural generated landscape with resource gathering and enemies that aren't just meat sponges along with NPCs that have individual dialogue and work schedules with good pathing around the randomly generated world that is of a GOOD quality while not being so predictable that you know what you're easily getting into, then please show me. The latest entry in Elder Scrolls series didn't do procedural generating, Elder Scrolls Online and GW2 didn't do procedural generating, Final Fantasy XIV didn't do procedural generating (and they pumped the new one out in 2 years and made iconic locals), Black Desert or ArcheAge aren't using it either. All these BIG traversable ONLINE games didn't use procedural generation.

One key thing to add on here, SotA's world isn't just supposed to be static backdrop. Everything has to be interactable. Can you truly procedurally generate every single interactable item to work with the terrain that was procedurally generated with NPCs that were also procedurally generated? What game out there does this?

 

Yes you can. You have to design for it. Some close examples I already gave, like Dwarf Fortress, even Minecraft comes close to this without even trying to do it.

If you are a non-creationist, this is how the universe was/is and continues to be, created - procedurally.

Otherwise all that applied "math magic" would be meaningless in our practical world.


I mean, if you are seriously asking, not rhetorically asking =p

http://www.giantbomb.com/procedural-generation/3015-328/games/

http://store.steampowered.com/tag/en/Procedurally Generated/

While a large swath of these are definitely more "roguelike", you can see the ever increasing trend in applying proc gen to more than level creation/content and tree distribution.

Cheers.

 

Almost all of those games that I took a glance at were either isometric, 2D, or not traversable with limited interaction with the world or were shooters/turn based; games with systems you easily manage because their mechanics are simple. None of them are full 3D modeled world that you can traverse in the Z plane with more than 3 people on the screen that all have individual complex stats.

IE; those examples are not proving a point other than it is becoming a trend but have yet to reach a certain fidelity that it could be transferred over to a game like in the Elder Scrolls series, or Fable, or Risen, or the Witcher or any of those big traversable games where there is an emphasis on player interaction with the environment or with one another in a fully fleshed out world.

I still don't see a good example with everything that I mentioned working together.

Again, SotA COULD do it and they just might, but as of now, they're still working on the key features of the game, something you may want to have done before you create a system to procedurally generate anything.

 

Er, sorry, but then you did not look too hard. There are plenty of 3d games in there. There are even a bunch of First Person ones on there. Mostly ISO? Think this is now an arguement just to argue.

And I dont understand the hyperbolic jump to all or nothing. When did that start?

They can't leverage techniques we've been using in the CG industry, and obviously the games industry, for over 2 decades? because it has to be everything or nothing?

 

Those statements aren't there to argue, it's there to show that something that is supposed to have the complexity of a fully realized online game with multiple levels of detail cannot be hamstrung together with procedural generation as easily as some folks think. I'm still digging through those games and the only one that has a decent amount of detail is Daylight and even then that game is supposed to only be 30 minutes long in a single playthrough.

Like you said, they can leverage the techniques of the past to stitch the results together, but the issue with this whole thing is the perception that it's something that can be easily done. As I mentioned multiple times, they could do it but they can't do it now. They don't have their core mechanics in place for them to generate anything that would be worthwhile. Again, as you mentioned, they could go over everything after each instance was generated, but at that point you'd might as well do it manually and from scratch because of all the things they'd have to change or put in manually. What I'm getting at is that as of this moment, for this release of episode 1, it would be pointless to run a proc gen unless it's on those simple stagnant points like a forrest node or grassland node.

I also wasn't going for the whole "all or nothing" mentality either, but instead showing examples of how complex it'd be adding more than one proc gen variables at the same time to work with one another, especially on a game that is supposed to be as content heavy as this.

 

Different genre, but Elite: Dangerous will do this at the galaxy level. It will procedurally create, using astronomical models and real data, the whole Milky Way, and populate it with procedurally generated NPCs.

Everquest Next is also doing something of the kind, but less extensive; it will have a map that is mostly procedurally generated (and voxel based), though AFAIK structures will be placed by hand. The game's NPCs are supposed to use some innovative AI that gives them desires and motivations, making them far more realistic than typical MMO NPCs.

I'm not sure if going this way is the best route for SotA, but as far as being possible in a MMO, it is.

 

I wouldn't mind all the zoning really. Problem is most of the areas i can simply run strait through in a matter of minutes/seconds. The zones are not big enough but that can change. There really isn't much to do in each zone either, again that can change with time. If I am going to zone into an area. I would like to spend time in the area for what ever reasons. We could use things like...

Special materials this zone has (gathering)
Camps of monsters, or roaming monsters.
A reason for why this zone(scene) is so important for us to be in.
.......
.......

There are many more things we can talk about in a zone area. The main point is they just did another pvp/combat system testing. They have the templates, basics of an area put in to place. I am sure as time goes on they will focus on these things. We are in the building blocks of the game, I am very hopeful that they will follow there design ideas and make something great from this.

Basically I am saying it is to early to judge if zoning is worth it or not. At the current time, no it is not. In the future, maybe. We just have to wait and see what they have in store for the future design of this game world.

 

Everquest Next was the only game on my radar that could be a proof of concept of the whole procedural generated content with NPCs and I'll have to look into Elite: Dangerous, yet the only interaction I've been seeing with Everquest Next are destructible environments (which is a big deal) and the parkour (hero running?) but I admit I didn't know about their NPC templates. However, and this is a crummy excuse to use but a legitimate one, if SotA had the manpower and the funding from a big company as SOE to the point that they could create their own voxel engine as well as use an existing one that was also made in-house, then we could see SotA become a game that is of the same scope if not better than Everquest Next. SotA will have to do something to make it stand apart, and there are a few things that could be listed that they're already doing, but expecting them to have completely unique zones from one another is too impractical unless they were going to keep it bare. As mindmage44 alluded to, they could procedurally generate nodes or special monsters but I'm not expecting anything grand like each forest node being completely different from one another to the point that they're indistinguishable or having them try to merge all the hex's together so you could transition from one to the other.

 

A while back, they ran monthly contests in which the community would put together their own scenes. There were a lot of great submissions. That's a perfect way to utilize crowdsourcing and give each hex a more unique feel. When the game gets closer to release, and there is a lot more interest, I wouldn't be surprised to find that they could crowd source dozens of scenes a month that just need a little bit of polish before being implemented.

I don't think the pre-release, at this point, is supposed to feel unique when you jump into random hexes on the map so I'm not too worried.

 

Again my 2 biggest issues, with any game really, is IMMERSION and INTERACTION.

These are the 2 biggest losses we have suffered in the MMORPG genre since Ultima Online, yet these to me have the be the core principles in any worthwhile MMORPG.

Richard spoke a little at E3 about how we have created tools in the game for the player to essentially "dumb down" the whole game in order to bring in new people who might of found tradition RPG games too tedius.

He used WoW as a example, where nowdays in most MMORPGS you can write the best material and lore, but most people only see the the exclamation point over the npc's head, the spot on the map to go, the quest log, and then the exclamation point on where to return it.

While he wont be doing this in SotA, he to me is looking past what he will be doing in his own game possibly that will be just as bad, and that is allowing players to basically bypass interaction.

I found myself falling victim to this already. Even as sparsely populated as the world currently is, I happened to run across a person who was doing exactly the same thing I was doing and was farming the encampment. As soon as I showed up the guy left, kinda huffy that I had come along, and instead chose to go into single player online, where he could continue on with what he was doing, without having to interact.

Trouble between points a and b? Either go into single player online or just use the overhead map and bypass it completely.

Again I realize the mindset that the team is taking, this game is going to be more about the expansions they will be selling to the storyline and less about a MMORPG world, but to me they should of stayed that course completely instead of injecting MMO elements into the game, because as I currently see it, this will really only be there so people can showcase their home building skills, and create pvp hotspots(towns) for PvP, which might as well be persistent instanced pvp.

Look before you go biting my head off, I know the game is still in its early stages, and the game hasn't been put to the test yet, and possibly my issues with the game will be alleviated, I may be making mountains outta molehills at this point, I just want to the developers to understand my issues with the path they have taken.

It feels like Diablo'esque to me to be quite honest, which would be fine, if you didn't have the wonder that was Ultima Online on your resume. I think in this case having that title under your belt in the end will be what this current game will be judged by, rightly or wrongly.

I am split on this because had you not added any MMO interactions to the game, and sold it purely as a ULTIMA 11(or 10 depending how you look at it), it wouldn't get the hype this game is currently getting, and I doubt it would of gotten the backing it currently has.

 

Heh, I never thought in a million years I would be saying this, or would of thought my MMORPG friends would be saying this, but we are at least a little interested in EQ:N hehe.

 

You already know it's PRE-alpha... meaning its BARELY got its key features implemented.

I'm not expecting more than it was already mentioned during the kick-starter presentation. If they can achieve what they've been talking about and do it with attention to detail and polish, then I'll be a happy player.

Heck, R9 was probably the best online experienced I've had in a long, long time. It was more about people and less about the system(s). That shows a lot of promise considering there's hardly a game yet.

Sign me up!

 

what you see is all your going to get sorry. sadly i agree with you .

 

It is incredibly annoying when people use caps to belittle something someone posted.

I don't think you even get the issue that I was discussing. At the moment there are way more tiles then there are scenes. What we are going to get with the way they are doing it so far is a lot of similar scenes linked together under those tiles.

I did not ever say do the entire game seed like...but for non important hexes it would sure be nice to have a bit more veriety. If you could procedurally generate a scene to go with a tile on the map you would be able to populate the world with a lot more unique(well relatively) scenes. It's not the perfect solution, but if you think they will have enough time to do them all be hand then you are dreaming.

Now, my other thought is why don't they have a program that makes scenes for them and then they could just add them in game after some touch ups. I have suggested in the past they open it up to the community to make a bunch of scenes but I haven't heard anything on that.

This limitation(if they need to stop to see it then well...I'm not sure what to say) has to be mitigated somehow...unless you want to run through the same forrest over and over.

 

It is also incredibly annoying when someone doesn't read. Especially when I agreed with one of your points.

Your point on non-important hexes being set to an algorithm is something I could see happening, as I mentioned, but they aren't going to go that route until they set up some prep models for that specific system as well as finish what the character limitations are on how they interact with the world.

 

In theory, EQN's NPCs are capable of things like autonomously fighting unscripted wars, moving camp when conditions around become bad (such as, for example, too many players farming a bandit camp ) or when they find a better place to live, asking for help (thus creating unscripted quests), etc. I'm skeptical, but if SOE can deliver just a fraction of what they are promising with their NPCs, it will already be a fantastic game.

 

I understand the gist of the comment. The issue in my view is that I have not personally played a game with instances so small and as many loading screens in a typical gaming session. It is to an extreme that feels uncomfortable for me. It will have an effect on my playstyle and how often I play as I would imagine it will on others.

I will likely not explore the world and will avoid mechanics that attempt to force me if they exist. I will rarely enter an instance unless gathering resources in the instance or for RP opportunities that exist in that instance. I doubt I will do much PvE other than what is absolutely necessary. It is a clunky mechanic and unenjoyable for me.

Edit: I should have said "clunky design". I understand the need for the mechanic and have played games with it but it has never been as disruptive as I find SotA to be.

 

It's almost a sub-genre of online multiplayer RPGs, but there are a few out there that are incredibly popular. Games like Monster Hunter or Phantasy Star Online/Universe have these finite zones that you could traverse in 30-60 seconds and those games sell like crazy in Japan and there is a hardcore fanbase in the states. However, those games have an emphasis on live-combat and are heavy in the action and actively dodging and taking down bosses. Because of that, you're almost begging for a loading screen to run out of the zone that had the monster so you can catch a breather. But again, different type of game, but it is a solid example of a game with small instances and a lot of loading that is doing very well.

 

I wouldn't state Elite: Dangerous as expansive as such, the scale will be, but David Braben already did the same thing with Elite : Frontier in order to save disc space way back in the 90s... I still have my original Frontier discs for the Amiga in fact, an entire Milky Way in 800kb of floppy disc. And the reality is that it generated a few data points (number of stars, number of planets) then an attribute for each (type of star, planet) and simple variables like orbital inclination etc. That's much, much less an order of magnitude than mapping every square inch of a planets surface, which an MMO will have to do. There have been more modern attempts (Spore for instance a few years back for very simple planet surfaces) but we're a long way away from having unique, finely detailed planets, especially with all the potential path finding etc done on it to ensure it's playable.

The Everquest Next is an interesting comparison because it proves the point; procedurally generating a landscape is easy, it's just a random height map. But to make it function, have any kind of character everything on top of that has to be done by hand.

Character for NPCs? Again, you can look at older games like Black And White by Peter Molyneux, it's easy to give creatures a huge range of stats, randomize them (or jusr give them inputs that through being blind to their meaning, you don't understand what lessons are being learned) and have emergent character come out of them... because the characters don't need to be that deep, just noticeably different. Again, in Frontier you'd get little quests from random NPCs, because there the cargo running, pirate blasting was more important than the backstory of the random name/face in the Bulletin Board. But for an MMO, where roleplaying is a vital part of the immersion, the technology just isn't there yet to create actual characters in the sense of a believable personal narrative; Dwarf Fortress comes close, but that's because it's a lot of simple text with an awful lot of viewer suspension of disbelief behind it.

Which is why "Design an NPC" is quite a high end backer reward here; Portalarium is going to crowdsource a lot of that. I wish it were a lower end reward, or even just thrown open to any constructive entries like the player written lore, but there we go.

That's not to say we won't get something procedural; and did you notice the requests to hire a world builder in recent Shroud emails? But you should be realistic; to make procedure it a gameplay component on the Spore/No Man's Sky model would require the entire game being dedicated to it, and a whopping great budget beyond what we have here... or accepting that you might spawn into a zone where the trees are entirely impassable, and just dull, dull, dull to explore.

Or, they could crowdsource basic maps too; there won't be many with the expertise to make them and it would require at least a few hours checking over them for hidden swastikas in the trees but the advantage of this model is that these can be added in later when authorised, presumably. Thus I'm not too worried for now. I just wish I had the time and Unity tools to have a crack at it myself... but having made a few maps for Dawn of War, believe me... it's really an enormous amount of work, even when you start with just a random height map to make a believable scene.