Bazzy_505
-
Posts
150 -
Joined
-
Last visited
Posts posted by Bazzy_505
-
-
On 5/25/2021 at 2:16 PM, Cheith said:
True, but maybe - just maybe - one should stop trying to work around using the rocket engines you have. Either get enough thrust up with the atmos to keep you going until your space engines kick in OR us the rocket engines. I can agree the rocket engines need fixed but frankly I don't think you need a new type of rocket engine to do that.
Exactly, when done correctly space engines nicely light up and throttle up as atmos slowly die off. Of course give one calculates airfoils drop-off correctly ? I haven't had much trouble with it the moment i really start paying attatention to these things and trained corresponding piloting skills
-
4 hours ago, le_souriceau said:
A good example of the problem. Look at the number of likes, reactions and retweets. Kinda cheap posting.
I do get JC loves to retweet pretty builds, and that's all right. But that hardly something akin to community
engagement. Who is the a target audience for a tweet like this? Most of us who build, trade and script have
been hanging around ducreators from the beginning in one form or another. For anyone else it's just yeah
a ship. It looks great when you evaluate it with the knowledge it's done in voxels, but if you take that context
out of the equation the image is a dime a dozen. It the kind of tweet that get buried within minutes.
- Lethys, le_souriceau and JohnnyTazer
-
3
-
Small, regular, bite size blogs or vlogs would have done NQ a world of good in terms of community engagement. It has been a great service to those of sound names and obscure ones alike.
As an example, look at Tim Shafer, in many ways a brillant designer whose main weakness has always been his inability to get any of his project to get done on time or on original budget but has always been a master at community engagement. And that loyal and engaged community was instrumental in survival of his company.
Another example can be Eric Heimburg, ( some of you may know his as Citan from the early days of MMO's working on Asheron's call2 or Evercrack2) whose pet project "Project Grogon" didn't even manage to make the cut on KS, but thanks to persistance and regular blogging on development of his project managed to build enough interest to actually finish that overengineerd mmo to fairly favourable reviews and self publish it with as few as 3 people working on it full time. ( and still make time for blogging).
So really the argument that being small studio and working on groudbreaking tech is moot.
Maybe it's a problem of "not seeing the trees for a wood".
-
On 5/25/2021 at 12:13 PM, CptLoRes said:
Don't forget that NQ only just implemented voxel compression in the latest 0.25 patch which is insanity, since one of the main points for using voxels in the first place would be that the data structure compress very well. So now that there finally is compression, it is going to be interesting to see how this has impacted transfer rates.
It will surely help somewhat, but what would have helped more is if their engine were more content agnostic and could situationally also stream simple baked vetex arrays instead of large of chunks of voxel grid data. But at the end of the day, these are all just technicalities.
The true deep issue of both DU and Starbase is that neither one is much of a game.
Both are are quite good at catering to the Ladmark crowd, but like in Landmark, once you have built all you wanted to build, there's little reason to hang around. Sure you can trade or run an industry, but that too tied to premise that there is somone on the other end who has yet to build all they wanted to build. PVP is more of placeholder at the moment. PVE doesn't exist. There's no lore to give any context or structure to anything we're doing. Exploration could have been great way to extend gameplay, but procedural generation is rudimentary at best, no real flora or fauna, generic repetetive terrain no real points of interest other than markets.
Pre blueprint fiasco, DU at least had the magic first 20-30 hours, in which you learned how build your first ship, crashed it, than built it better up to the exhilarating moment when you reached the orbit of a planet for the first time. Where DU really fails is at what happens right after that.
To put fine point at it, if DU were a decent game, most of us would not have minded to buy into the most expensive gaming rigs just for the chance to play it.
-
7 hours ago, HangerHangar said:
Yeah it’s all exit strategy with Starbase releasing next month.
though they’re not going to be able find any buyers/leasees for their software stack. Even the original devs can’t keep massive numbers of item duplication and access violation bugs out of their own game. Likewise they aren’t really using authoritative servers (which would be a selling point compared to SB) considering how many times I’ve seen other’s walking on air/in ground because they’ve had a different cache version. And their stack is absurdly resource hungry even after an hour on Thadis you’re still going to sit at 5MB/S transfer, where most publishers will essentially demand 100KB/S or so for a typical area.
While Starbase does a few things really well, they have their own technical can of worms they have yet to address. I do wish Starbase a long and happy life, they are very likely to bleed users faster than they gain new ones. In fact quite a few starbase playes were eyeing DU as possible alternative before NQ shot themselves in both of their feet with all that 0.24+ ish nonsense.
Most likely mid term effect of both SB and DU will be that no producer will want to touch a voxel based online project for quite a few years to come. Not that there was great interest from publishers to begin with. Esp not after the spectacular failure of Landmark/EQN.
You are completely spot on about the insane size of strings passsed between client and server for DU. Now that said 100KB/S is not a realistic figure for any voxel implementation, even in the most rudimentary implementation at least 12bytes for positional data in voxel grid and at minimum at 12bytes for surface normals, material type/color. And we're still talking about plain box voxel, now you need to implement some sort of transform, let's say we'll go with a variation index on of predermined transforms 1byte ( which is not that great but gives us a workable 256 possible predermined transform variations), you'll end up with 25bytes per voxel.
Let's say for surface 256x256 voxels (like a base of M construct) 1 voxel deep you'll end up with roughly 1.5mb to stream. Now even on ground you realistically need at least 2x as much even with modestly sized viewport, that brings you little over 3mb, excluding your ship, other flying garbage around you and all those fancy space elevators and prtocol overhead ( let's say UDP with MTU at 1500, max packet size 1518bytes -ethernet frame- IP header -UDP header gives you 1472 available for actual data (46bytes overhead)
Not defending NQ, just want to give you an idea of the size of dataset voxel engines work with.
- Cheith and le_souriceau
-
2
-
2 hours ago, Cheith said:
Apples ... oranges ... technically you cannot fairly compare this to EVE in 2003 - this is a lot more ambitious. It might also be worth pointing out that there was way more negative posting, acrimony and general whining in the early days of EVE. Also CCP with Dust might be a more appropriate comparison and that turned out well.
Actually the comparison is quite fair, CCP had to develop all their core technologies from scratch. NQ had the luxury of licensing most of them.
Also if you consider capabilities of period accurate hardware, the scope of challenges faced were quite similar (just consider how limited the bandwidth of system bus was even on server hardware just to pick one issue from the top of my head). It's also worth mentioning difference in capabilities IDE's now compared to those 20 years ago is difference between heaven and earth.
Having said that CCP is by no means an exemplary developer, they have had quite a few blunders under their hat in years that followed, But that's besides point. This is not a Marvel vs DC argument, it's about comparing two projects with similar ambition, same audience, and comparable development challenges in face of technical limitations.
But if you really want something relevant to challenges NQ is facing, Innes McKendrick had a really good presentation at gdc in 2017 https://www.gdcvault.com/play/1024265/Continuous_World_Generation_in__No_Man_s_Sky_
-
23 minutes ago, TONSCHUH said:
DU even shows me, how much worse Cyberpunk 2077 and Star Citizen could be and don't forget the latest ED DLC disaster, which I luckily didn't buy yet !
It's a bit sad to be honest, because DU has had a great potential feature-wise, but "I think it's dead, Jim !". ?
ED Odyssey is indeed a hot pile of mess, i have no idea how Braben plans to come back from. It's certainly no buy for now, if ever
Quite frankly CP2077 was not technically as bad as reviews would have you believe, at least not on PC with RTX, in about 170hrs in put into it, i haven't had a single crash or quest bug ( luckily). My main problem was that CP isn't a very good game gameplay wise. I've put as many hours as i have into it to make sure i somehow didn't miss getting into the good part, sadly there wasn't such, but at 100% completion i was at least certain i didn't miss it.
As far DU goes, it might not be dead yet, but it is sure on life support. And there are examples of games that came back from the brink. Just remember how sadly broken FFXIV or NMS were at release, and both are very much alive and great shape nowdays. Even Battlecruiser 3000AD, (another Derek Smart game) which released in completely broken state managed to pull through in the end, even if it had to be done on D.S. own unpaid time. In many ways it's all about commitment to your project and your paying customers that decides what side of cliff it lands on the end. And in case of DU, we should see either way before long.
-
16 hours ago, m0rrty said:
Hi guys,
When you build the ship, you want tot have windows, lights, all those wires and wire-corners, right?
BUTT!!1111!!1 But if you crash, most of the elements will be broken, and require repair. I have 250 elements on my M-ship, I am not talking about L-ship....... I have to walk around and fix every 50 hp point, 200 point, etc, elements, and it takes big part of my life, JUST to CLICK and fix red element, then click fix yellow, and wait of the fix....
What do you do? - You just don't place that much of elements?! Right? ----> we are loosing the esthetics, we are loosing part of the game without elements.
I have this problem. It is not a bug.
But I wand a feature request.
My suggestion: to have a Construction >> Right Mouse button >> Command "Repair", which may repair your ship for you, spending time, which you do whatever you want.
My Suggestion: Admin, make the glass and other small elements not breakable ?
Your ideas?
In my experience, the best practice, at least for the time being, is to build your ship like a russian cold war era, no bells, no wistles, no heating, no cushions, mugholders, fancy radars and chairs.
Once you have the hull, engines, airfoils, tanks ,adjustors and brakes laid out, test it and adjust the hell out of it untill the flight characteristics are rock solid. Make a conscious decision early between naturaly stable ( but less maneuvrable) or naturally unstable ( more maneuvrables) design.
Once your design behaves consistently with your design goals, than start adding all the windows, switches, force field ramps.. etc And even re-fly the ship often after adding any major element so you can backtrack effect of each change on design.
And last but not least, at all stages of tuning your design, always test with various sized dummy loads of cargo ( containers fill uniformely, rather they always fill as first linked first, last linked last) . This is particularly important when tanks and cargo containers are not placed along the center of lift/gravity. When using multipe fuel tanks, watch the changes in weight distribution as tanks are emptied. Not a huge concern on XS cores, but get more profound as the sizes increase.
And remember, first few ships you design will very likely get off the ground, they might even manage get out of atmosphere, but that doesn't mean they're not crap,
and that's okay, nobody get its right the first few times
-
1 hour ago, TONSCHUH said:
As the game is already in it's beta stage, it reminds me of the un-optimised mess called "Arma-3" and I ended-up un-installing DU.
?
Arma-3 was butter smooth compared to Universal Combat made by the one internet troll that shall not be named ?
-
On 5/20/2021 at 12:49 PM, Burble said:
Fair point there. I remember seeing youtube and other social media advertising a while back and thinking DU was getting to the main stage. Not so much now.
There was a fair bit of positive buzz going about DU sometime in 2018 when they released a new batch of videos, it even got covered by a few, what you can call mainstream, gaming portals, but since NQ had nothing more to stoke the fire with, it quickly fizzled out and fell back to obscurity by the end of the year.
After that it has gotten a few mentions when paid beta started, but no frontpage coverage.
-
I believe each and every customer has every right to express their concerns and grievances for the product they have paid for and in no mean resembles project advertised in kickstarter much less anything akin to what you see in videos and early pre-alpha previews. Additionally majority of the issues with NQ's product were well known and documented all the way back
in early alpha, and very little was done in terms of remedies to these issues. What's even more interesting to consider is the fact the NDA we had to agree to was never lifted, not even after beta launch. I've been in quite a few alpha programs, both paid and free and i have never ever encountered such draconian terms.
Now before you resort to pejorative expressions like "whining" let's put a few numbers into perspective and compare them with other past projects of the kind.
First of all NQ is not indie developer by any meaning of the word. The moment you have venture capital involved, you're not independent anymore. It's not necesarily a bad thing, quite the opposite. But even if you were to broaden the definition to anything self published on a dime 20mil+KS money+Alpha backer money gives you a number which is very much outside of realm of what is generally considered indie community backed project.
Now let consider past project that bears the closest semblance to DU and even JC's original pitch can be summarize as "that game" but bigger and in voxels.
CCP developed Eve Online in 2.5 years with a team of 35 developers on a buget of 2.6 mil euro in a small country of Iceland the pool of available developers of which is a fraction of what you'll find in Île-de-France alone and where everything is pretty much 2x as expensive as anywhere else in Europe.
With those 35 people and 2.6 mil in the pocket CCP developed their own 3D engine from scratch, developed the full server stack from scratch, built all 3D assets in house and has it run on their own metal server housed in datacenter in UK.
NQ with 35 people (if the number is to be believed) and 20.6 mil in the pocket , merely licensed Uniengine2, licensed core of the server tech, purchased most 3D assets, and leased server capacity in AWS.
Now granted, CCP initial development window was in the period between 2000-2003 so it would be just fair to adjust the budget for inflation which puts it just to little above 4mil in 2020 money.
I rest my case
As i've mentioned in my previous posts, i have no regrets about forking out for being alpha backer, i did get my money's worth out it. It is but a pity that it did not go much beyond that.
I still hang around DU for tiny slim off-chance NQ gets its house back on track,
But I will not be throwing any more money at DU unless NQ can, through their actions prove, that they can get out of that development ruth they've been stuck in for past 2 years and i would caution against newcommers commiting to paid beta access until they do. There simply isn't enough value on the table for the time being.
- Bobbie, Sabretooth, blazemonger and 3 others
-
6
-
21 hours ago, GraXXoR said:
Damn, run out of heart reacts for today...
Believe me, I am the master of minimising cross section: My daily driver is called the "SkySausage." ?i would say your defining trait would rather be your obsession with ailerons
but that's okay, you can never have too many xs ailerons ?
-
From what i can tell from my experience with ships i designed, the odd damage to tanks and such dissapeared as long i had an air gap between tank and voxel skin of the ship. Placing them near center of gravity seems to be a plus and has quite a few additional benefits when it comes to handling.
Also the damage calculation seems rather simplistic so it seems it's realistically just a few impact vectors calculated at the front the construct, and the inconsistencies are mostly due to float variables used instead of double variables. float has 23 bits of significand, 8 bits of exponent, and 1 sign bit coversely double has 52 bits of significand, 11 bits of exponent, and 1 sign bit.
calculations with doubles are twice as accurate, but also much much slower, and than there's question of what size is the actual grid on which these calculations are performed. You can see the occasinal drift even when hand placing voxels when the construct you're working on is sitting at an angle.
That said i don't see voxels soaking any damage, rather they affect where at what angle the impact happens.
I'm personally on the "brakes are pussies, real men flare up 30m from the ground" sides of things, and i've noticed i prevented most of the damage to front adjustors by tapering edges of the nose.
-
On 5/14/2021 at 10:33 PM, Revelcro said:
At some point in a relationship you sometimes have to recognize when it is becoming toxic due to the inactivity or inability to communicate from one or both parties. Once you are sure that a relationship has passed a certain point then usually there is only one outcome that is logical, and a lot of people here have already passed that threshold weeks ago. More are getting to that point. Daily it seems.
It hurts the most when you know the relationship could be salvaged or at least extended just a bit with better communication or a tiny bit of effort. Then your heart breaks when you never hear a word or see any of that effort put in and you may never know the whole story of why.
I dunno, to me its more of situation akin to such as when you break up with your crazy girlfriend and realize the next day,
that you left behind your favourite Phillies t-shirt you have had since highschool and she's been using it as PJ for the last 2 months and you really
wish you could find a way to get it back without a scene ?
-
codeword of the day is air-resistance (aka drag).
Which elements get damaged on chuck norris re-entries are affected by angle of attack, current trajectory, rotation of ship surfaces in relation to current movement momentum.
let's say you've been applying trust straight with -20 deg nose down nad cut thrust and rotate ship to the 10deg to right just before htting atmo you'll burn off adjustor on left side of nose and top side of left wing ( always extremities on the most out of misaligned axis in relation to trajectory)
-
20 hours ago, GraXXoR said:
@blazemonger I was thinking back to my first few days in game... when we didn't have capacity limits on our nanopacks.
What have we actually got since then?
A digit on our screen called quanta... Mining mining mining... slightly better looking scenery...
external containers... clunky external nanocrafters.
SCHEMATICS!!!! YAY!!!
randomly changing glass panels that seem to change colour every patch.
Clunky bot-infested markets... that can take minutes to interact with due to piled up junk from rage-quitted players.
Clunky pseudo-pvp... that nobody wants to play because it's clunky,
a few more Lua functions.
A one-size-fits-all warp drive and beacon.
shitty textures on our wooden furniture that make things look worn out.
Triangle panels that are TIER 2 that take longer to make than jet engines (because triangles are so much harder to make than squares goddamnit!)
Have I missed anything?You forgot to mention the most important change
MORE MINING!
-
I don't believe engines need universal bump to pre-nerf times. But having said there should really be more meaningful difference between different favors of the same size engines. As it works right now, there's no reason to bother with anything but military variant of any engine. More varied fuel-to-thrust-to-weight-to-fuel consuption would go to differentiate between the engine grades of the same size.
As for the engine wall problem, no matter what the effective thrust is, people will always fit as many engines as they can to haul as much as they can. The only true solution if to finally implement powergrid on cores and have each engine type and variant specific powergrid requirement to link. A size of core or a reactor size connected to core would determine max available grid to link all active elements. Shipbuilding would than become more of balancing game between different operational requirements and would have been good foundation for further game systems to implemented related to guns, defensive systems etc.
Wings, stabilizers and ailerons i feel are more of an immediate issue. Current models for wings are pure garbage. what's more you can't even properly add adjustors to its surfaces without insane amount of fiddling and attachment point issues are not even worth talking about. Rest of the airfoils also have a fair more quirks than they should have. Weight of ailerons vs their lift and size make the large sizes quite impractical to use. Also some interoperability and consistency between wings /stabilizers and ailerons wouldn't hurt either.
-
On 5/13/2021 at 11:40 PM, Merkor said:
Hmm maybe I'll try that. When you fly in space do you need to apply reverse thrust to slow down?
You certainly can, but flipping around to kill forward momentum when entering atmosphere can be very tricky to control.
Probably the easiest way is to use space brakes. And rough rule, for comfortable no frills maneuverability you want the
the braking force be around double of the forward thrust of your engines.
You can mount additional forward facing space engines to use as brakes, but compared to space brakes they're too heavy
for what they do and unlike space engines, space brakes need no fuel.
Now what directional space engines are very nice for is to get finer control on trajectory adjustment when entering atmosphere,
and get esp. important even more important when entering no atmosphere or low atmosphere planets/moons.
-
On 5/16/2021 at 10:08 AM, GraXXoR said:
I will NEVER buy a PC with application/OS storage on spinning rust (HDD) ever again again. Full Stop! Just don't punish yourself or your PC like that.
SSDs make a world of difference and can be well over x10 faster than HDD in some scenarios.
Without derailing too much from the original topic, there are still valid uses for spinning platter harddrives, and the choice is not just about classic harddrives offering larger sizes.
As far as speed goes, best (MLC) SATA SSD's like 860pro are about 2-2.5times faster on reading/writing compared to HDDs in similar price bracket ( we're not talking all the budget Blues, Reds etc.) But aside from Samsung, nobody makes SLC or MLC drives for consumer market anymore (in industrial/enterprise space they're still a thing), and when you look TLC or QLC drives, which are the most common drives sold these days, the write speed very much dives off the cliff once the fast area of SSD is filled.
Most TLC and QLC drives have an area reserved that is treated by controller as if it were a SLC cells to which data flows first once dram cache is filled and than later rewritted to final destination on drive. But in large transfers, once both DRAM and fast area is filled mid transfer, the write speed dives off the cliff. If you're transfering 20gb data you will initially see the 400-450mb/s at the beginning, but about half way thru you will see it drop (depending on controller and actual flash used) as low as 160mb/s. And on writing side of things you can beat it quite easily with WD Gold ( formerly branded as Re) enterprise class drives, more so if you're running them in RAID10. Granted, these are by no means cheap, are quite noisy and run quite hot.
Having said that, when it comes to pure gaming rig, there's little reason to go with classic HDD, even as secondary drive for to offload porn collection to. And even the crappiest SSD's like Crucial MX-500 or the whole Adata range will easily outlive the useful life of the system ( as long as it's reasonably sized). And true enough with an old rigs from before the times when SSD's were afordable ( say 2-3rd gen Core times) it feels like getting a new PC.
Now for workstation/mixed use, HDDS are very much still a thing, Especially if you're a coder working with larger repos where there's a large amout of small sequential writes happening all the time when debugging, profiling etc.
On PC a use for work, i have 2TB mvme for system volume and 2x 2TB (128MB cache) WD Re in Raid10 to store project files so i pretty much get the best of both worlds.
-
too bad it's absolutelly the worst time to buy SSD right now due to the bloody chia coin craze.
most worth-a-while SSD's are either hopelessly sold out or selling for insane prices.
example: Corsair Force MP510 (nvme) 1.9TB drive which until recently had very much one of the best price/size/tbw/speed ratios
which i used to buy for about 240E a pop till late april, now sells for 440E... same situation with 1TB+ samsungs
-
after so many "we told you so" moments we have in alpha and through whatever this is now,
the main point of ideas forum is for us to have means to emulate the genius loci we had gradually lost over the course of last year.
a true cynic might insinuate it to be the means to bide time till that juicy 1.8 trillion wad of taxpayer's money from post covid EU rebuild fun washes over private and public sector alike?
-
well besides all that it was meant to be and isn't,
sanct still has realatively uncluttered markets with convenient route to the lagville to grab an odd missing component on foot. In short it's still a very decent place to build ships and due its smaller atmosphere fairly fuel cheap access to space.
-
internal voice "...........yay another, markets landing zone full of garbage threads"
...........silenty adds another notch to the sizeable collection of notches on even more sizeable wall of a very spacious prison cell, many have gone thru before.
Yup! exclaimed a farmer not agreeing with the police
-
as it has been discussed a few times before, DU's combat would have greatly benefitted from more defensive and ewar systems. But from the gameplay perspective radar jammer as you have described, would have removed interaction from the game rather than add to it. And for an online game, it would have been quite a fair bit contraproductive.
having said that, i can imagine few radar jamming implementations could be interesting. For instance, pulsing jammer that would not make the the opponent blind to your presence but within either increase lock time, when taget isn't locked yet or had % chance to break lock each jammer cycle when already target locked. To conterballance it, jammer could have been limited to application on target locked opponent, rather than having passive aoe effect. And finally there would be energy requirements to consider to make its use situational.
Repair Unit 100% broken.
in General Discussions
Posted
I think the proper name should be "very obtuse relativistic automated diagnostic unit" VORAD .. even the acronym has a nice military ring to it ? . Easy to use even when cursing on public chat ?