Mordgier
-
Posts
459 -
Joined
-
Last visited
Posts posted by Mordgier
-
-
1 hour ago, CptLoRes said:
The reason why NQ stopped publishing roadmaps, was probably because they never once came close to actually following them..
I don't think NQ ever had the staff to actually align their vision with their resources.
I think the roadmap was JC putting a list of things he wants into a game on a timeline without any clear understanding of what was actually required or if the feature was even possible without anyone asking "How would this work?" and "How many dev hours is that?"
So as the vision met reality everything began to slide and slide and slide.
Planetary warfare is a prime example of a JC vision roadmap item - it sounds great -but he never had ANY idea of how it would actually work.
If you don't even know how it will work or if it's even possible how can you put it on a roadmap? You have way way too many unknown unknows at that point.
-
On 11/25/2022 at 1:58 PM, Kezzle said:
Not just the devs. The entire staff should be playing.
Please have some respect for the devs. Coding DU is bad enough, being required to actually play it too feels like just kicking someone when they are already down.
I know how many of us feel about NQ, but please find some mercy in you for the devs - they've suffered enough without having to play DU.
- CptLoRes, Novean-61657, Vilifier and 3 others
-
6
-
23 hours ago, Cergorach said:
It also matter greatly how much prestige the person doing the negotiating is bringing to the table, the use case as you mentioned, the potential future growth and at what point in the lifecycle the product being negotiated is. I've seen folks making incredible deals that just blow away the list prices and others with far, far bigger companies just getting a 20-25% discount (which isn't insignificant, but not blowing away the list price). It seems a bit like how much decision powers does the negotiating person have (how quickly they can make and approve the deal) makes a lot of difference. Also having a large and well connected 'network' helps you hook in higher into a company and just get better deals overall. I have none of those skills (nor do I want to pursue those), but I've seen people in action that do and it's incredible seeing that contract signed! 😉
Well back in the day NQ did manage to get Intel to brag about how good a multicore CPU is for DU - https://www.intel.com/content/www/us/en/developer/articles/technical/intel-core-i9-processors-reduce-open-world-load-times.html
Guillaume Gris is still at NQ so I'm not sure it's totally fair to say that they don't have the talent to pull off some fancy overhyped whitepaper for a vendor that could be used as leverage for a discount - but given that DU has squandered it's potential I doubt any vendor today would be eager to have a NQ logo on their customers page with a case study attached to it....
-
2 minutes ago, blundertwink said:
That also depends on what you mean by "signficant user" -- I'm skeptical NQ is actually big enough to leverage much of a break.
For example, an adtech company I worked for had well over $50,000 a month in AWS usage. The only break they gave was in flexibility of payment terms...that was helpful, but hardly game changing.
The only other discount they pointed us to was in reserved instances, which is very significant but also comes with very significant up-front costs. Other services (like bandwidth charges for CF) have built-in discounts that scale based on usage, already.
I can't emphasize enough the world of difference between on-demand and reserved instance pricing -- the more you can afford to pay up-front, the more affordable (and less flexible) AWS is.
Otherwise, the only other discount I know of is for their enterprise level plans -- e.g. companies with more than $5 million in spend per year start to get discounts, with more committed spend leading to more discounts.
Honestly, I was under the impression that it's really rare for Amazon to give special pricing, except through very high levels of enterprise spend or existing discount programs -- that's the experience I've had with firms that likely outspend NQ, but I'm hardly an expert with the biz side of AWS.
Either way, with the list of the top AWS customers being mostly billion-dollar entities, I don't think NQ would be a "significant user", relatively speaking.
Also...I wouldn't underestimate their CPU usage!
CF might push huge loads of voxel data, but there's still plenty of server-side processing that relies on traditional infra, likely including RDS as well. We can only guess on their usage, but I wouldn't lay a bunch of bets on their stack being efficient.
Fair points - our cloud hosting bill is about 25M per year so given our scale we've never paid list for anything but I have no idea how low the bar goes for "You pay list and you don't" - I admit my general mindset is that "list price isn't real price" but I might be totally out of touch here...
Org scale is not the only thing, sometimes "unique' use case that can serve as whitepapers/case studies for the vendor or proofs of concept that they can use to secure other clients can be used to reduce pricing at least in the short term. AWS going around pointing at NQ as a K8 cluster success story in the gaming world given Amazon's push into game hosting via GameLift might be enough for them to cut them a deal - but who knows.
You're also right about the compute costs not being 'insignificant' but I just have no way to know how intensive or not their backend is because none of the data points are visible to us. All we do know for a fact is how much data DU pushes to clients by looking at our cache size.
I'd love to have more insight into the DU backend - especially given that DU is a fairly unique concept - but doubt it'll ever be public data.
-
1 hour ago, Jinxed said:
I have never used AWS but I believe there are two basic repeating charges.
number of CPUs used and peak CPUs.
Amount of data transferred.
Lets say If we double the number of players, the we generate double the cpu usage.
this in and of itself is no issue because we’re doubling income but salaries and office space rent won’t change so they’ll get more profit.
the problem comes with the fact that from day 1, the amount of constructs and data served per players will generally increase.
at some point in the future, you could envisage a time when players have so many constructs that NQ can’t afford to have them downloaded to all the players.
So the cost of DU is not really CPU cycles but CloudFront - DU uses Cloudfront to push out all the voxel data. Any construct or terrain change you make is stored in cloudfront and pushed out to other clients as needed and then stored as cache by your client, at least till the cache is invalidated. Just how much data is pushed is easily tracked by checking your cache size folder. For the estimated costs it's pretty easy - https://aws.amazon.com/cloudfront/pricing/
You can confirm this easily by the way -
Keep in mind that NQ likely has a contract with AWS that is for negotiated pricing, it's pretty rare for any significant users to just pay the list price.
That's why mining had to go and that's why constructs have to get destroyed if not tied to a paid account - else they are forever sent out to other clients and NQ pays for it.
Tracking millions of autominers is just a bunch of rows in a table - it's not a significant cost really.
As to why NQ made the AM timer 24 hours instead for 20h like the "daily" timers in many games - it's because they have a consistent history of failure to learn from the mistakes made by other games.
"Daily" timers being either set to a specific reset time or less than 24h to avoid the constant 'shift right' for the players has been standard practice for some time in other games.
I had previously posted a breakdown of what I estimated NQ to be paying for hosting and how many players they needed to cover their staffing costs with the profit assuming France average DEV costs + taxes - and that post got deleted and I took a 3m ban for it sooo uhh not doing that again - you're welcome to draw your own conclusions though...admittedly I did also state that you shouldn't buy into DU because of that data because well...uhh...DOA.
- huschhusch, Duragon, Novean-61657 and 3 others
-
6
-
-
1 and 4 are not possible.
Please at least learn how DU works in the backend before making grand plans to save it. It is fundamentally tied to running on an AWS k8 cluster and cloudfront.
3 is way more complex than you think because DU as is suffers scaling issues and would need a ton of work to look decent in VR.
5 has been planned already
6 is not possible because 1 and 4 isn't possible.
The End.
-
-
23 hours ago, Haunty said:
Well they've already had 2 previous releases, approaching 4k logged in players in the early months of beta with relative stability, so there shouldn't be anything catastrophic, but I'd expect a few days of occasional reboots and hotfixes at least.
I think you may have a very different concept of "relative stability" from others - because what I recall is people being unable to log in and unable to finish the tutorial because of the crippling lag...
-
On 4/10/2022 at 9:44 PM, NoxDragon said:
The game appears to be completely broken. Is this the case?
Hmm. Guys this man might be on to something....
-
On 2/16/2022 at 5:05 AM, ELX987 said:
i dont care if its 10 players, 100 or whatever, i think this game will come back!
This is cute.
Imagine thinking that a game with 100 players is viable.
NQ has 68 employees according to LinkedIn. It's safe to assume that they are burning through ~5m a year - in short they need 50k paid accounts to break even.
-
Wow it looks like all 3 remaining builders are outraged...
-
I can't help but laugh at the MU implementation. I was begging for MUs because mining was just such a nightmare:
So I hoped MU's would help remove a lot of the tedium from the game and actually make it fun to play....
Except when we got MUs it was totally not what I had hoped for....I wanted to remove tedium - not add more!
-
-
1 hour ago, CptLoRes said:
And this is what we tried telling NQ since 2017, but they never even acknowledged that there might be a problem until suddenly Demeter.
- We think you are underestimating how much people are going to dig, how will you handle the persistence server load? Silence...
- How will you deal with fixed resources (ore and tiles) making it so that new players will have a harder and harder time staring the game since all the best spots will be taken? Silence..
- This game is really difficult to learn for new players and the building system requires lots of resources to learn proper. How will you solve this? Silence...
- Can this game really function without PVE? Silence..
- How will you limit people from advancing to quickly but still have fun building? Silence...
- The lag city stress test (pre-alpha thing before we had District 6 to highlight the same issue) shows that this game will not scale anywhere close to what you promised it would. Is this going to be fixed before release? Silence...
- Why is there so little progress being made with regard to improved game loops? Deafening silence...
Ha - I wish it were just 'silence' - half the time we'd get talked down to like we're a bunch of chicken little's .... "You don't understand just how much ore there is to mine! It'd take years!"
Every time you did get a dev response on any of the concerns, you'd get told you were just wrong...
- CptLoRes and blundertwink
-
2
-
1 hour ago, Jake Arver said:
I think you know full well what I am saying. Running the game on any other platform than Windows is not supported, it may work, it may not.
That Unigine supports Linux does not mean the game does or should. And Unigine is really just a small part of the game relatively. Several other parts of the game do not support Linux.
It would really be a waste of resources for NQ to support Linux. Not only is there really no userbase to speak of, they also do not have the manpower or financial emeans to even consider it. I get that may not be what some woudl like to hear, but those are the simple facts.Literally the only part of the game that does not support Linux is EQU8.
You are welcome to tell me what these 'other' parts of the game that do not run on Linux are. DU ran 100% fine on Linux before EQU8 was added.
DU ran on Linux fine - albeit unsupported - prior to EQU8.
Once EQU8 got added, firing up DU in a linux container died with it.
I don't game on Linux, but no matter how you twist it - NQ made a decision to murder any hope of linux support with their choice of anticheat.
DU was runnable in a linux container until late summer 2020.
-
23 minutes ago, blundertwink said:
NQ wasn't led by a game developer or even someone with gaming experience, so they decided to go the "faster" route and rent servers from AWS and use an existing engine.
You're wrong here - DU uses Unigine for rendering on the client side but their server tech is proprietary. There is nothing wrong with running on AWS. AWS offers phenomenal scaling capabilities and is a fantastic solution for game hosting where player counts vary through the day and life cycle of the game. It is frequently cheaper and better performance to host via AWS than in house if you are expecting a major difference between your high and low loads.
It would have cost NQ far more to buy whatever silly number of physical hosts they would have needed to host the game when it launched - and then what? The player count is likely less than a third of what it was a year ago.
The game engine and hosting were far from the primary failure points of DU. The lack of a game and the complete and total failure to listen to the community played a bigger factor. JC literally played the game with infinite resourced and complained that mining was too easy despite never having mined.
-
13 minutes ago, CptLoRes said:
And it is only now at the end of this year they have started to admit publicly to be struggling with the server stuff, and start making changes (some good, some shockingly bad).
Remember when they conned us into kickstarting this game by implying that their server tech was totally sorted out and could scale basically infinitely to be a true massively multiplayer game?
This was the vid that largely sold me on DU. Having burned out on the unbearable lag in Empyrion - this seemed like a game that'd solve the unsolvable problem of Empyrion lag.
The server stuff was supposed to be the sorted out solid foundation that they built this game on. So many other open world games failed due to poor server tech - I genuinely believed that this would manage to be the exception. The game built by an actual engineer, from the ground up - focused on a solid technical foundation of a scalable and high performance massively multiplayer server engine. After all - the engine failures are what made Empyrion and Wolds Adrift awful in combat - surely DU would get it right by starting with the server tech...
So much for that....
-
2 hours ago, Jake Arver said:
That may be so @zyziux, but DU has always been a windows only game and is wel ldocumented as such. It is you personal choice to risk buying in to the game when not on Windows.
There is always the option of a Shadow PC though..
No. DU was not a windows only game until the addition of EQU8 anitcheat. Prior to that the game ran fine though a KVM on Linux etc.
DU is built on Unigine which runs fine on Linux.
The current limitations are entirely the result of choosing to use EQU8.
-
2 hours ago, Underhook said:
Interesting. Cutting costs and increasing profits on a product that isnt finished. Thats typically what you do when you want to sell.
Nobody is going to buy NQ in it's current state. The more likely scenario is trying to position themselves for more loans/investments by extending the time at which funding will run out.
What they should do is turn blueprints into NFTs and watch investors throw money at it...I mean come on lets be frank it won't make this game any worse....
-
-
On 6/4/2021 at 12:50 PM, blazemonger said:
The "devblogs" were published little over a month ago..
My expectations
PVP will not see it's rework until maybe.. possible.. end of this year, early 2022
PVE.. well.. we'll get scripted events called asteroids mining
power/energy management is put on hold
There is no actual information on anything right now.. it's all deliberately vague and non descriptive. I still suspect that there is no certainty about the future of the company and at best I can see a 3-4 month budget while the powers that be try and work out to either do an investment deal or turn off the lights.. I doubt we will hear anything structural about he future of DU until after we hear anything on the future of the company.
Finally.. here's a good example of providing informative updates, focusing on what you _can do_ instead of putting the effort in excuses for what you are not doing .. ANNOUNCEMENT Elite Dangerous | Odyssey: Roadmap | Frontier Forums
It's sad to me that the unfinished release of Odyssey and the communication after can be used as an example for NQ to learn from.
-
The subscription cost is nowhere near the elephant in the room.
It's not even in the same room.
I wouldn't play DU for free.
- JohnnyTazer and Dhara
-
2
-
Lack of roadmap kills motivation
in General Discussions
Posted
Right because it was a wish list of things JC wanted in a game - without ever sitting down with real game designers to sort out the mechanics for these and then engineers to see what it would take for these to work as described - and then to do a POC and then road map it publicly.
I'm willing to bet you that a lot of these roadmap items were placed there with no feedback from the dev team on feasibility.
I've worked in places exactly like that - where client provided list of deliverables had items on them that any of the company engineers would have said were not feasible with our infrastructure and could not be provided with the resources we have. For example a place I used to work promised a client that we would improve our guaranteed uptime from 99.9 to 99.999 because sales folks doing the contracts "didn't think it'd be a big deal" for us to do that - and we found out that we were on the hook AFTER the client was presented the new contract....needless to say we failed to deliver. Of course IT guys got the blame for failing to meet the SLA....
The DU map is no different - it is full of things that cannot be delivered regardless of dates because nobody sat down and asked "How would this even work?" before throwing it on the map.