Elite Dangerous need to get serious about Cross Platform.

Well, you seem to be too sure of that. I do not know. I am not a netcode expert I am afraid. But I can remark that the player limit condition is very specific to sim-like games such as War Thunder, IL2 Sturmovik, Elite, DCS etc that probably have a much larger data realtime bandwidth requirement per player given they need to account for things like, speed, acceleration, relative attitudes, proyectile colision and trajectory, specific craft internal module spatial location etc that games like EVE or WoW do not need.
Yup. Hitscan twitch shooters are rather simple in comparison to sims.
 
And ED is not primarily designed as a PvP-focused game, especially on a massive scale. You can't put a saddle on a cow and complain about what a bad horse it is.

I never said it was designed as a PvP game. Unless you are referring to what the other guy said.
 
I'm a bit of a cheapskate. I'm more likely to pick a game because it has no monthly subscription than because it has good server architecture.

I'm just not a fan of subscriptions and have never paid one for a game and never will.

Having a C/S architecture doesn't require a monthly subscription, but it does increase operational expenses,so they have to get money for that from somewhere.
 
I never understood why Frontier forces every player to bring their own NPCs to things like stations and tourist spots. THIS is what kills performance more than anything, and it breaks the game in general when there are 100 NPC Belugas jostling to land. The number of NPCs should go DOWN as more players enter an instance. We are VIPs, and the streets need to be roped off and the rift-raff (NPCs) kept at a distance when a bunch of us enter town. In other words, instancing performance out in the middle of nowhere is significantly better than it is anywhere NPCs gather, especially stations, to the point where I've done CGs in solo because the NPC "clutter" and subsequent framerate drop griefed me more than any player could.

Well, i suppose they could reduce the number of NPCs in an instance based on the number of players. Maybe they already do that to some extent, i'm rarely in places with many other players around anyway.
 
And ED is not primarily designed as a PvP-focused game, especially on a massive scale. You can't put a saddle on a cow and complain about what a bad horse it is.
As I mentioned in a thread a week or two back, one of Elite's really big design decisions (giving both major problems and major advantages) is that it has exploration as a meaningful primary profession, and therefore needs a really big galaxy, and therefore has an expected modal and median player population of 1 per system (ignoring completely empty systems), and even many of the systems with 2-5 players in those players are so far apart that there's no point in trying to meaningfully treat them as being in the same place.

That instantly kills any (economically feasible) C/S model for inter-player messaging just because of the sheer number of mostly useless servers it would need. And it makes a hybrid of P2P for small instances and C/S for big ones also tricky because all big instances start out as small ones and there might be some interesting exploits around the different lag/latency behaviour in each mode.

Infinity Battlescape: one system and Dual Universe: a few systems can plan for thousands on each server because that's their baseline standard case and they can have a small number of very big servers ... but the cost is "no exploration" (and "nothing except combat" in IB's case).

I'd like to see a game go the other way and be based pretty much solely around multiplayer cooperative space exploration, with heavy logistics that make Fleet Carriers look like speedboats. But it'd be a big risk for a niche of a niche, and the "content" problem is always a tricky one.

Having a C/S architecture doesn't require a monthly subscription, but it does increase operational expenses,so they have to get money for that from somewhere.
Elite does of course have a C/S architecture for most of it anyway - it's just the inter-player messaging which is P2P.
 
As I mentioned in a thread a week or two back, one of Elite's really big design decisions (giving both major problems and major advantages) is that it has exploration as a meaningful primary profession, and therefore needs a really big galaxy, and therefore has an expected modal and median player population of 1 per system (ignoring completely empty systems), and even many of the systems with 2-5 players in those players are so far apart that there's no point in trying to meaningfully treat them as being in the same place.

That instantly kills any (economically feasible) C/S model for inter-player messaging just because of the sheer number of mostly useless servers it would need. And it makes a hybrid of P2P for small instances and C/S for big ones also tricky because all big instances start out as small ones and there might be some interesting exploits around the different lag/latency behaviour in each mode.

Infinity Battlescape: one system and Dual Universe: a few systems can plan for thousands on each server because that's their baseline standard case and they can have a small number of very big servers ... but the cost is "no exploration" (and "nothing except combat" in IB's case).

I'd like to see a game go the other way and be based pretty much solely around multiplayer cooperative space exploration, with heavy logistics that make Fleet Carriers look like speedboats. But it'd be a big risk for a niche of a niche, and the "content" problem is always a tricky one.


Elite does of course have a C/S architecture for most of it anyway - it's just the inter-player messaging which is P2P.

Yeah, which is why i generally refer to it as a hybrid architecture.
 
Infinity Battlescape? Potentially hundreds of thousands of players?


All time peak = 219

Ummm....

Easy to make claims of greatness when those claims are never tested.
Don't be an.... rediculous. I said hundreds to thousands. Not hundreds of thousands. There is a difference. Learn it.

The player numbers are because the game still is in very slow development (much complaining about this) and the fact that there is still no progression yet to keep people interesting in playing long term. So if you are going to use Steam player counts to make your case Agony Aunt (oh the irony... how many times I have seen you condemn others for using Steam player counts to make a case contrary to your white knight positions) at least be honest about it.

I know that you know this because we sparred on IB's forums together (I was under a different user name at the time). So stop playing the devil's advocate (to put it nicely).

Besides, we are talking about their amazing netcode. You are better than this Agony.
 
Last edited:
I agree with the sentiment in this post, including the fact FDEV can do better. Although I do not think this is a matter of apologism or lack thereof in this particular discussion. It is a matter of also being realistic and realizing that the player limit issue is not just a limitation in Elite but it is something quite widespread across most sim-like games out there, server based or not, including behemoths such as WarThunder, DCS or IL2.

Regarding Infinity Battlescape I have actually played it in early access since I am a backer, and it is true the battles there were spectacular. And something FDEV network devs may be interested in looking at. It is probably about the only sim-like game out there that I have seen so far manage a decent number of crafts in the same location.

Now having said that I never saw hundreds, nevermind thousands, of players in the same location or battle. At most the battles I took part in had a few tens of players at any given time, maybe 30-40 in the best of cases? (very smooth all of it though) Although maybe I missed the really big battles, so this is just anecdotal for me. Infinity Battlescape can host hundreds of players in the same scenario but most of those players were spread out quite far appart across the whole system and the netcode logic probably does not need to update your client as often for those ships far away not interacting with you at all etc (which is in essence no different from what Elite does). Also the ships themselves did not strike me as onerous as Elite´s or DCS´s in terms of quantity and complexity of systems and modules, which is info that I suspect also takes part in real time data exchanges during a battle etc.
Thank you for your honesty here and willingness to acknowledge a game that does their netcode better, way better IMO, than ED's.

There were hundreds to thousands (not hundreds of thousands) of ships spread out over their simulated solar system. Their solar system is all one instance with smooth transitions to atmospheric planet surfaces. I saw easily a hundred+ ships a few times in a single battle. This was before their releasing to Early Access on Steam (I am a long time backer).

Also, IB devs make this claim on Steam and it is correct as far as I could tell:

a massive, true-to-scale procedurally generated star system that is 100% seamless (no loading screens). Multiple corporate factions are fighting for control of a solar system in a war involving hundreds of players, potentially more than a thousand per-server ( without instancing ), that can last from hours to days

Anyway, I will reiterate (forum trolls and incurable white knights be damned), ED could very well improve their netcode to make the sort of conflicts, discussed here in this thread, possible or at the very least a whole lot better. There is no doubt about this AFAIAC.
 
Last edited:
Anyway, I will reiterate (forum trolls and incurable white knights be damned), ED could very well improve their netcode to make the sort of conflicts, discussed here in this thread, possible or at the very least a whole lot better. There is no doubt about this AFAIAC. The only people here who are not willing to at least acknowledge this are the ignorant and the die-hard white knights.
I couldn't resist it :)

Did the compression algorithm introduced recently have any visible effect on instancing? Only asking as I don't have the faintest as I never run into that many folk in the sticks...

(Hopefully not too White Knight :) )
 
In software development, there is always someone around to tell you they know better, you know nothing, your not doing a good job.

Normally it's your manager of your manager. Not your manager, he knows a bit about it.

Poor old Fdev have to put up with some of the people above telling them the above with absolutely zero knowledge of the problems involved
 
In software development, there is always someone around to tell you they know better, you know nothing, your not doing a good job.

Normally it's your manager of your manager. Not your manager, he knows a bit about it.

Poor old Fdev have to put up with some of the people above telling them the above with absolutely zero knowledge of the problems involved
No... ALL of the condenscending and toxic remarks aimed at the Developers are made by incredibly skilled, multi-talented developers who earn in a day what I used to earn in a year...

Or so they'll have you believe :ROFLMAO:
 
Anyway, I will reiterate (forum trolls and incurable white knights be damned), ED could very well improve their netcode to make the sort of conflicts, discussed here in this thread, possible or at the very least a whole lot better. There is no doubt about this AFAIAC.
You can point at anything and say "it could be better". It's criticism that is completely devoid of meaning and value. It's just criticism for the sake of criticism.
 
In software development, there is always someone around to tell you they know better, you know nothing, your not doing a good job.

Normally it's your manager of your manager. Not your manager, he knows a bit about it.

Poor old Fdev have to put up with some of the people above telling them the above with absolutely zero knowledge of the problems involved
I am a software developer with a bit of experience and I can tell you what's wrong with ED: technical debt.

They raised £1,578,316 in the campaign and if you divide it by £40,000 (what the average game developer salary is), you get a bit less than 40 monthly salaries for 1 developer. If you get 2 developers on board, you have money for only 20 months. Let's say that DB was really generous with this project and that his company invested the same amount. That's still 4 developers that have to bring the project to a successful end in 20 months. Can you imagine the amount of spaghetti code they wrote in that short period of time? Hint: they can't change cockpit colors, every bugfix creates more than one bug etc. etc.

I am very well aware why they need so much time to change anything, I am aware why things break and I am aware of companies that have better products because they had more time, more money and probably better software architects.
 
Don't be an.... rediculous. I said hundreds to thousands. Not hundreds of thousands. There is a difference. Learn it.

The player numbers are because the game still is in very slow development (much complaining about this) and the fact that there is still no progression yet to keep people interesting in playing long term. So if you are going to use Steam player counts to make your case Agony Aunt (oh the irony... how many times I have seen you condemn others for using Steam player counts to make a case contrary to your white knight positions) at least be honest about it.

I know that you know this because we sparred on IB's forums together (I was under a different user name at the time). So stop playing the devil's advocate (to put it nicely).

Besides, we are talking about their amazing netcode. You are better than this Agony.

I understood quite well what you said. Hundreds to thousands. ED can do a hundred under the right circumstances. Is that enough? Well, depends. For me, who is usually the only player in an instance, occsaionally a few others, then yes. But there again, i play GTA in solo lobbies or with just friends, because open sessions are pure toxic environments.

As an aside, Rockstar have just make GTA free to play via Epic. I dropped into a public sesssion today. First message i see "I killed you because you are a free to pay Epic player". Yeah, thanks, but no thanks. I don't need those sorts of players, and make no mistake, those players exist in ED.

Regarding Steam numbers, all you have to do is read back and pay attention to my comment about how I was not entirely happy about posting Steam stats, but with the lack of other information i had to go with them. And, as you rightly point out, it doesn't show the whole picture. But we can make some inferences from them, same as with ED. So when people say "only 5000 people playing ED" we know there are a lot more than that, at least double can be reasonably expected from those who purchased directly from FD and Occulus etc.

Now, take those Steam numbers and double, nay triple or quadruple them, hell, multiply them by 10x, and it still paints a very sorry picture, one that means its highly unlikely that that thousands of players in a single session is rarely, if at all, tested.

I did a search for someonse streaming thousands of players in a session, and couldn't find any. Maybe they exist, and if so, i'd love to see such a video, and check out the quality of it, see if there is lag/jitter or not. If you have such a video, please, do post!

I have no idea what you mean about us sparring on IB's forums. I've never been on IBs forums. You must be confusing me with someone else.

As for talk about their amazing netcode, well, indeed, but i've still got to see proof that their amazing netcode is anything more than talk, and talk is cheap.
 
I thought it's already known and accepted that indirect crossplay is there since ED launched on consoles.
Well, not by OP, but anyway he thinks he can win election by killing other players :unsure:
 
Might as well add my vote. I'd rather see cross platform playing than any other new features. The fleet carriers are neat, but basically just more of the same. There's no good reason I can see that they can't coordinate between the Xbox and PlayStation companies to do their later updates in a timely fashion.

More available players is better.
 
Back
Top Bottom