The dynamic universe and background simulation leaves something to be desired

Is not a problem of capacity/technology, but of developing time. Your average desktop computer today can run a 'simulation' multiple times more complex that what we got in-game right now, tracking billions of 'agents' and flows and with way more logic that what we got now, and all that in real-time.

So no, it's not a problem with processing power.
 
I'm not sure what Frontier would do without your visionary input, OP. Rome wasn't built in a day, you know.
I don't feel like I'm stuck with anything. See? Opinions.

I almost forgot what forum i was on! i got through 11 pages before I saw a blowhard fanboi who can't take criticism of his game. Seriously when i read the OP i thought he would be flamed to hell before the first page. Im glad ALOT of people are having a coming to jesus moment about what ED IS and ISNT.

A few months ago I wrote a piece called "Look at the sand" that was not taken well at all but I felt pointed out the glaring shortcomings of the game in its current state. Criticism is extremely healthy for making a game better. I hope to see you all in space in a game we can all enjoy soon!

They could just do what EVE did and make 100% player run economy
did i stir the hornets nest yet? :p

This would honest to God solve a lot of problems. The more content you can let your community create, the less development needs to be spent on creating scripted content.

Seeing how shallow the simulation is currently they could use all the help they can get.
 
Last edited:

Jex =TE=

Banned
I thought "its only beta" was a perfectly acceptable answer at the time too.
The main thing people kept bringing up was, as you say, release is in X number of weeks, you can't release the game like this.

I had the unpopular idea of having the devs give everyone money so they don't need to grind again to be able to test the python and orca, particularly the orca and its passenger missions. I thought very few people could earn enough money before release to find most of the passenger bugs.

So during the testing phase, they didn't test this out.

Pathetic, truly pathetic.
 
Is not a problem of capacity/technology, but of developing time.

Time spent on development is a huge factor, I agree - they 'released' this game way too early - it's basically an alpha/beta quality skeleton outline of very basic glued-together placeholder subsystems. This is being exascerbated by their choice of P2P archtecture.

Your average desktop computer today can run a 'simulation' multiple times more complex that what we got in-game right now, tracking billions of 'agents' and flows and with way more logic that what we got now, and all that in real-time.

So no, it's not a problem with processing power.

This is precisely my point. FDEV have potentially on tap a worldwide supercomputer able to maintain a persistent galaxy/universe by mere dint of the fact that at any one time thousands of game clients are online 24 hours a day.

Even when they bring the game down for an update, it is more than feasible to get each compute node to shut down gracefully, so there would be no concern there.
 

Jex =TE=

Banned
The OP makes some good points but as an amateur developer and a gamer I can explain at least some of this stuff to you.

Illusion Mechanics: The first is illusion mechanics. Despite what you may think 99% of everything you have ever played is based on this. That monster that is attacking you isn't really attacking you, he is following an orchestrated script that gives that appearance aligned with animations. It feels like he is attacking you and that's all that matters, but if you read the script and realized what that monster was actually doing, what the code was actually doing, how the code responds you would begin to see the flaws in it. What you would really be doing however is lifting the veil of the effect that illusion mechanics have and this is effectively what issue you are having.

NPC's don't go anywhere even if they appear to, they don't serve a purpose even if they appear to. I agree that Frontier could have done more to keep up the illusion in a few places, its particularly bad when your playing and you can spot a lack of intelligence in the code but if you are actually looking under the hood looking for how things work, I promise you that you will always be disappointed in all games. There is no magic or intellectual AI's, its just software doing what it's told and 99% of the time its doing it for nothing more than the illusion for the players benefit.

The important thing I'm here to tell you is that you need to stop looking under the hood if you want to enjoy this or any other game. Its become common place among gamers to dissect the inner workings and functionality of games, discover the illusions and complain about it. Your ruining your own experience. Stop analyzing and let the illusion trick you.

No, we get everything about the code, we understand how the game works but we want some cognitive dissonance - we at least want to see a world where the A.I. do things because it makes it more immersive for us. Most other games manage this fine, and we don't question it. The problem you're not understanding is that in ED, they do it horribly and it breaks immersion and it does need to be discussed because it's sh*t.
 

Jex =TE=

Banned
In all honesty, even if the game weren't based on peer-to-peer networking, I have serious doubts that the kind of persistent NPCs some posters seem to want would be practical.

I'm not sure how many inhabited star systems there are right now, but I have heard 100,000. Even if there were only 25 NPC ships per system on average (a rather low number, given the total population of the game world and the fact that the largest cargo ships can only carry 500t at once), to handle them all persistently would mean that the server would have to be running a 2.5 million AIs simultaneously.

You'd need incredibly powerful (and expensive) servers to handle that, and even then I'm not convinced it would be technically possible. I'm not aware on any persistent world online game that even tries something like this.

I completely agree with the OP. Also nice job writing all that down. But yeah, the players do not seem to have any sort of meanigful or even perceivable impact on the world. This is the part of the game that is the most disappointing to me. I sincerely hope that this is going to change in the future.

It wouldn't track all the AI in the 3d world - it would be done using data tables.

Falcon 4.0 came out on 12th December, 1998. In it you flew over Korea and it had around 30,000 units to keep track of, land, sea and air. It did this by placing a "bubble" around the players. Anything within 30 miles of the player was rendered and fought in real time, anything outside of that was fought or moved using tables.

I read recently there's around 3000 systems populated and if that is the case, servers today, dedicated servers, not 486 pc's running at 66mhz Falcon was running on could easily handle the AI without breaking a sweat. I reckon 2.5 million wouldn't be an issue either.
 
I posted the following 18 months ago, during the height of frontiers touting for late backers, and a period where fantastic and very imaginative proposed game play features where rife on the forums, from players and devs alike. Braben's enthusiastic and regular dev diary videos were fresh in the memory back then too. There was a real energy and drive behind it all.

18 months later its plain to see how naieve those dev and player posts were, the one pasted below being a prime example. And now with hindsight I very much doubt this game was ever going to be even a fraction as deep as the dev proposals made it out to be during that fund raising period. I just don't see it in elite now. Its gone down a completely different path.

Anyway, the following background evolutionary gameplay event was based off the ambitious frontier videos and ddf blogs that were around back then...

June 2013...
I'm not sure if this type of thing is planned or even possible with ED, but its something I've wanted to see implemented into a gameworld since the days of Frontier and First Encounters. Now in 2013/14 maybe with today's technologies it's at least feasible?

Anyway, what I'd love to see in Elite Dangerous is an organic evolution of the environment and the political theater as opposed to a strictly scripted inception and outcome of pre-determined events. As a long time Eve player I found the mission arcs in that game where so limited in scope - almost completely scripted and had very little bearing on the wider world as a whole. They were predictable and repeating grinds that eventually became ISK faucets for large organised player-run corporations.

I'd hope Frontier Developments don't go down that path...

Here's just one single example of a mission (or lets say an in-game event or incident) that can organically grow into something far more interesting:

Stage 1 - Lets say you're out exploring the Frontier systems in your deep space scout ship. You enter an unsurveyed system and conduct a long-range scan. A few minutes later your scan reveals that the system is a binary and has a dozen planets. Now you switch to planetary scan mode.

A few minutes later you get a hit on a possible "goldilocks planet" i.e. a planet with enough mass to retain an atmosphere that orbits its sun at just the right distance that it could be temperate enough to sustain known life.

Now you want to conduct a closer inspection so you plot a course to take you into orbit around this world.

A few minutes later you're in low orbit. From low orbit you can see its a pretty desolate world, continent-wide deserts, large polar caps etc, not much cloud cover, no discernible vegetation.

You launch a probe (or in a later version of ED, you go planetside and fly into the atmosphere, fly across the surface, or even land and get out of your ship to take rock & soil samples :cool:)

After a few minutes the probe sends back its analysis and full break down of what prime resources this world has. But more importantly it concludes that the world could be made habitable through Terraforming. Bingo! You just hit the jackpot.

You now have in your possession the complete digital readout of a possible habital planet with untapped resources. This item (a datachip for example), can now be sold to a prospecting-type NPC corporation back in the core systems.

The game would prevent resurveying of this particular world for a set time to give you time to report your findings and for other reasons mentioned later*

You set off for home where you sell your datachip to the highest NPC bidder and here ends our scout pilots role in this saga.


Stage 2 - Some time later the corporation / businessmen who purchased the data chip makes it known via bulletin boards that they're hiring planetary surveyors for a classified mission. Human players OR NPC's (if no human players are interested**) are hired as the actual surveying team or their escorts. Their role is to survey the planet further and gather information on what types of terraforming equipment will be needed (again this is done by sending down advanced probes, and analyzing atmospheric conditions etc). A full in-depth survey is now complete and the info is relayed back to the funding NPC corporation. Here ends our surveying teams role in this saga.


Stage 3 - Some time later the NPC corporation begins advertising for freighter pilots and escorts to deliver classified machinery to a classified destination (again Human players or NPC's are hired for the roles via bulletin boards).

On successful delivery of the cargo to the destination world, you are paid accordingly. This marks the end of our freighter pilot & escort roles in this saga.


Stage 4 - This stage is all NPC driven. Deployment of the terraforming equipment is undertaken and adequate automated defence systems for that equipment is put on line.

Terraforming can take some substantial time to complete - lets say from weeks to months real time.


Stage 5 - The planet is now habitable and the NPC corporation begins sending NPC colonists and hires (human or NPC) escorts to take them to the new world. The NPC corp is now given full mineral rights & ownership of this world once the first colony base (NPC built) is on line.

This ends the first chapter of this saga...


Stage 6 - By now the NPC corp has begun building spaceports in orbit and has a permanent NPC presence in the system. Human player or NPC pilots are being hired throughout the Core systems to deliver equipment and food to this fledgling system. NPC prospectors are moving to this system enmass in order to exploit its untapped resources. The high traffic attracts pirates - and bounty hunters (Human players and/or NPC's).

The system information is updated on the galactic map. Its now a thriving and developing colony and the Frontier boundaries are pushed a little further out from the Core.

Weeks / months / a year goes by and now due to the changing political landscape (or the discovery of an ultra rare element within the system) it becomes a disputed system between two rival factions. Both factions begin hiring mercenaries to defend/attack the colonists and a whole war could break out over this one world.

-------------------------------------------------------------------

In conclusion, these event arcs should be organic and at any stage could go off in a completely different tangent. For example - *if the original guy who scouted the planet is blown up on his way back to the core systems, the data chip could be lost or captured by a completely different entity that has no interest in doing anything with it. That world would remain unsurveyable (therefore undeveloped) until the datachip timer expires - be it months in real time later. The story would end there.

Or if the surveying team or subsequent colonists are constantly destroyed by pirates enroute, the NPC corp may postpone the venture until hired gun bounty hunters (NPC or Human) have eliminated the threat, or the corp may even abandon the venture altogether if it became too costly or the corp found more lucrative incomes elsewhere through other overlapping world events. If the latter, again the story would end there.

The point is a whole host of unrelated circumstances could crop up that delay or even prevent this evolutionary arc in the gameworld from developing. None of the stages are a given. Anything could happen that prevents this whole situation from occurring. The main thing is its completely unscripted. One stage could lead to another if the right conditions occur..

This is how universal events could grow and evolve from one single incident - a lone pilot out in the wilderness who took the time to survey a few distant worlds. His impact on the gameworld may have been insignificant at the beginning but as time and events unfolded, he can look back and see just how much his actions have evolved into something far bigger.

** I think its important that these types of events should be independent of human interaction because to feel fully immersed in a game on this scale I need to know that I'm just a part of the gameworld and that the gameworld doesn't revolve around me. Yes my actions have a bearing (like the fictional surveyor pilot), but the universe will evolve whether I just sit and watch it from my cockpit window outside Lave, or whether I get directly involved. I think this is where the biggest challenge is for the game developers. Is this the kind of 'evolution' David Braben mentions in one of the vids, or is this something still a few years away development-wise?

Someone once said on here that he feared elite would have the scope of a vast seamless ocean, but the depth of a sidewalk puddle as developers never deliver a fraction of what they talk about once the monies been banked. So far he was right.
 
This is precisely my point. FDEV have potentially on tap a worldwide supercomputer able to maintain a persistent galaxy/universe by mere dint of the fact that at any one time thousands of game clients are online 24 hours a day.

Even when they bring the game down for an update, it is more than feasible to get each compute node to shut down gracefully, so there would be no concern there.

Well, the problem is that it only brings more complexity and overhead for the already poorly performing P2P code. Not saying what you propose has no merits, but the BG would first have to justify such architectural design choice. Right now it does not, as it can be run on my freaking smartphone probably (that's how complex it is).

So first improve what we got with the current architecture, and in the future if there is a need for it they could decide to use the P2P networking to their advantage.
 
Well, the problem is that it only brings more complexity and overhead for the already poorly performing P2P code. Not saying what you propose has no merits, but the BG would first have to justify such architectural design choice. Right now it does not, as it can be run on my freaking smartphone probably (that's how complex it is).

So first improve what we got with the current architecture, and in the future if there is a need for it they could decide to use the P2P networking to their advantage.

I agree with the above - I also did say it would probably be a daunting task to implement such a thing ;)

And yes - you probably could run the BG sim on a modern mobile phone (though I'm now waiting on a dev to pop in and say otherwise ;) )
 
Well, the problem is that it only brings more complexity and overhead for the already poorly performing P2P code. Not saying what you propose has no merits, but the BG would first have to justify such architectural design choice. Right now it does not, as it can be run on my freaking smartphone probably (that's how complex it is).

So first improve what we got with the current architecture, and in the future if there is a need for it they could decide to use the P2P networking to their advantage.

p2p is a waste of code...its infinitely hackable inefficient and and a very poor interface realtime interactions that are time sensitive.....the only advantage of p2p is that they didnt have to spend money on servers...ie they went ultra cheap and used a substandard antiquated multiplayer interface

they dont have to justify the need for a better interface simply fire the ...erm fellow who had that bright idea
 
Last edited:
p2p is a waste of code...its infinitely hackable inefficient and and a very poor interface realtime interactions that are time sensitive.....the only advantage of p2p is that they didnt have to spend money on servers...ie they went ultra cheap and used a substandard antiquated multiplayer interface

That's the compromise I wouldn't have imagined FD to make. Appears to be a very poor move.
I love the retro feel of the game in design but in the engine it's a pain.
I wonder if someone builds a house and believes it's supposed to last long, why not using the latest and most potent materials?
Or if money is an issue why not build a smaller one with the same stabilty? (I don't know how it's organized financially in FD - building up a reliable core engine for the background simulation may be more expensive ONCE but then it carries the game all along. Now I can only shoot wild guesses about how much will it cost to keep the "manual content injector" people in service all along the supposedly long period the game will be supported... Is that really a deal?)
 
Last edited:

Jex =TE=

Banned
I cringe everytime I hear something about "planetary landings" in a game where I can't even interact with placeholder NPCs.

Planetary landings are a big white elephant. This is not going to fix anything. It's not a part of gameplay because you can land on planets. It's cosmetic - just another place to land. Does anybody think landing on space stations is really cool now we've all done it a hundred times? The stations are amazing, they look great but they're all the same and once you've landed on them so many times, I bet people hardly bother looking at them now. The same will go for PL's - they'll be great looking no doubt but are not going to fix anything because they are not a game play objective. There's no "Elite" for the guy with the most and best planetary landings.

Also, in Frontier, the Eagle and wheels on it. If they made the landings actual airplane style landings then that would be really good but I doubt it. FD need to work on the content - I don't know why everyone is getting excited about PL's - what are they going to offer to the lacking game play??
 

Jex =TE=

Banned
As far as the 400 billion stars - for playable game content, and by playable I mean something more involved than a place to visit and eyecandy - a working fully interactive system - it's 399,999,999,000 too many - I don't think Eve O has even 1,000 systems and everyone seems happy enough 11 years later. It's been pointed out several times that it's not even possible to visit all of them in any human lifespan - and apportioned down it would take thousands of players with separate lists an entire lifetime to visit just their list.

It'll take you over 30 years just to count to a billion ;)
 
The OP makes some good points but as an amateur developer and a gamer I can explain at least some of this stuff to you.

Illusion Mechanics: The first is illusion mechanics. Despite what you may think 99% of everything you have ever played is based on this. That monster that is attacking you isn't really attacking you, he is following an orchestrated script that gives that appearance aligned with animations. It feels like he is attacking you and that's all that matters, but if you read the script and realized what that monster was actually doing, what the code was actually doing, how the code responds you would begin to see the flaws in it. What you would really be doing however is lifting the veil of the effect that illusion mechanics have and this is effectively what issue you are having.

NPC's don't go anywhere even if they appear to, they don't serve a purpose even if they appear to. I agree that Frontier could have done more to keep up the illusion in a few places, its particularly bad when your playing and you can spot a lack of intelligence in the code but if you are actually looking under the hood looking for how things work, I promise you that you will always be disappointed in all games. There is no magic or intellectual AI's, its just software doing what it's told and 99% of the time its doing it for nothing more than the illusion for the players benefit.

The important thing I'm here to tell you is that you need to stop looking under the hood if you want to enjoy this or any other game. Its become common place among gamers to dissect the inner workings and functionality of games, discover the illusions and complain about it. Your ruining your own experience. Stop analyzing and let the illusion trick you.

Well, Oolite has got this right: every ship has a purpose. Perhaps FDEV should have a look at their code, it is open source...

Heck, even if it is nothing but an illusion: Archimedes Elite (1992) did better in that respect than E: D.
 
The OP makes some good points but as an amateur developer and a gamer I can explain at least some of this stuff to you.

Illusion Mechanics: The first is illusion mechanics. Despite what you may think 99% of everything you have ever played is based on this. That monster that is attacking you isn't really attacking you, he is following an orchestrated script that gives that appearance aligned with animations. It feels like he is attacking you and that's all that matters, but if you read the script and realized what that monster was actually doing, what the code was actually doing, how the code responds you would begin to see the flaws in it. What you would really be doing however is lifting the veil of the effect that illusion mechanics have and this is effectively what issue you are having.

NPC's don't go anywhere even if they appear to, they don't serve a purpose even if they appear to. I agree that Frontier could have done more to keep up the illusion in a few places, its particularly bad when your playing and you can spot a lack of intelligence in the code but if you are actually looking under the hood looking for how things work, I promise you that you will always be disappointed in all games. There is no magic or intellectual AI's, its just software doing what it's told and 99% of the time its doing it for nothing more than the illusion for the players benefit.

The important thing I'm here to tell you is that you need to stop looking under the hood if you want to enjoy this or any other game. Its become common place among gamers to dissect the inner workings and functionality of games, discover the illusions and complain about it. Your ruining your own experience. Stop analyzing and let the illusion trick you.

Please don't ask me to lobotomise myself in order to enjoy this game.
 
Back
Top Bottom