Author Topic: Star Citizen - The E.L.E.  (Read 1137023 times)

dsmart

  • Supreme Cmdr
  • Administrator
  • Hero Member
  • *****
  • Posts: 4915
    • Smart Speak Blog
Re: Star Citizen - The E.L.E.
« Reply #75 on: November 24, 2016, 09:07:55 AM »
*snipped a bunch of rubbish*


I messaged a CIG developer on the RSI forums and asked outright, quoting your comment:

I hate to break it to you though, but that's pure - and utter - rubbish. You all know this.

64-Bit floating point precision map is NOT the same as 64-Bit positioning. In fact, on Nov 22, 2015, I wrote a missive explaining that.

But despite the fact that even the engineers have gone on the record debunking that nonsense, some of you still keep parroting it.

Basically, all the scenes are "stitched" together in order to have a cohesive whole that quantum travel sequence hides.


The dev was kind enough to respond to me. Here is part of the reply:

it's true that things get *authored* as isolated scenes. That's just good sense. But this whole "stitching" thing is just lack of imagination on his part, I think. Or a setup for him to backpedal and say that because assets get loaded as you approach them, he was right all along.
If all you want is a dev saying he's wrong, well, he's wrong.

I am a tier 1 engineer, I deal in absolutes - not semantics. There is literally NOBODY in the ENTIRE team over there, that has worked on ANY project of this type and/or scope. And the past 4 years has proven - without a doubt - that they're just winging it, spending a lot of time on R&D, tech demos etc - still no game. So excuse me if I scoff at anything that any "engineer" over there has to say.

The "stitching" thing is 100% factual and isn't even open to discussion because I am 100% correct. That's how it is being done, it's how to do it because it would be impossible to fit an entire world into memory. Why? Because it's not procedurally generated.

When you have two scenes (S1, S2) which are different, and part of the game "world", they have to be stitched somehow. It's not only efficient, but it also gives more options for scene and player management. As big as Elite Dangerous is, they do it. ALL my games do it. Including Universal Combat (see this ), which has the largest (until ED, it was the largest in gaming history) game world.

The key to having a "seamless" world is either i) procedural or ii) stitch it

If you go with (i) you use data values to generate the scenes in real-time

If you go with (ii) then that's where loading screens come from. When player moves from S1 to S2, you unload S1, then load S2. This is what ALL games with different scenes do.

If you want to give the illusion of a "seamless" world, then you have to come up with ways to stitch S1 + S2 in order to give that illusion as it being one cohesive whole. And there are several ways to do it. The two most popular ways are to either (i) stream load S2 when the player hits a trigger point in S1 or (ii) create a transit illusion when the player is committed to moving from S1 to S2.  The commit phase is why you use one of these to go from S1 to S2. You can't get out of it.

In ALL my space combat sim games (Battlecruiser/Universal Combat), I use option (i) to go from space to planet, and vice-versa - via an external camera transition when the player's ship breaches the planets gravitational pull, or when they climb up high enough and hit Escape Velocity to leave the planet. And I use option (ii) via a jump anomaly (jump gate, wormhole, fluxfield) to go from one space region to another.

This is what the segmented world of BC/UC looks like. It's not procedural; so each scene (space and planets are different) is completely separate. Each space scene is a simple small binary file with data that the top-level AI script parses in real time, to generate the scene, and pointers to everything in it. It's fast and efficient - hence no loading times.

This is what the data definition file for the Earth space region looks like:

Code: [Select]
Ted32 text format file id:191550422,1
Unit:3 (Km)
 group:MAIN
 size:0,origin:(0,0,0),extents:(0,0,0)(0,0,0)
 (
   points (11):
   pid 0:(72496,860087,393344)
   pid 1:(-413288301,-330238555,350381112)
   pid 2:(-65804045,-697187263,-524462638)
   pid 3:(-737764523,172603385,85956996)
   pid 4:(476547561,7329222,-30087638)
   pid 5:(-6221827,592913797,-389212126)
   pid 6:(654311881,647176854,-325243888)
   pid 7:(-436135255,370818784,-42587638)
   pid 8:(737466352,-598273156,225537362)
   pid 9:(565000000,145000000,435000000)
   pid 10:(410000000,-395000000,435000000)
   structures (10):
   [0] col:2 at:0,use:PLANET.3D:EARTH
   [0] col:11 at:1,Rx:85, use:JUMP.3D:JMP-01
   [0] col:11 at:2,Rx:90, use:JUMP.3D:JMP-02
   [0] col:11 at:3,Rx:94, use:JUMP.3D:JMP-05
   [0] col:11 at:4,Rx:94, use:JUMP.3D:JMP-04
   [0] col:11 at:5,Rx:89, use:JUMP.3D:JMP-10
   [0] col:11 at:6,Rx:90, use:JUMP.3D:JMP-12
   [0] col:11 at:7,Rx:89, use:JUMP.3D:JMP-11
   [0] col:11 at:8,Rx:92, use:JUMP.3D:JMP-03
   [0] col:7 at:10,use:MOON.3D:MOON
   subgroups (3):
   include:PLANET.3D
   include:JUMP.3D
   include:MOON.3D
 )

This is an excerpt from the master data file showing what the data definition for Mercury, Venus, Earth planet region looks like:

Code: [Select]
#
# SOL SYSTEM
#
."SOL","skybox08"
#
# MERCURY
#
:Mercury.3d,p,"MERCURY","SOL","PLANET 1/9"
:Mercuryz.3d,s,"TERRAN","SOL","SECTOR D9"
mercury,p,mercury,ptype01,ptype01,hot,TERRAN,4440,4878
jmp-10,j,earthz:TO..EARTH
jmp-15,j,venusz:TO..VENUS
flx-04,f,mercuryz,snv01z,blk01z
#
# VENUS
#
:Venus.3d,p,"VENUS","SOL","PLANET 2/9"
:Venusz.3d,s,"TERRAN","SOL","SECTOR D9"
Venus,p,venus,ptype02,ptype02,hot,TERRAN,25000,12100
jmp-02,j,earthz:TO..EARTH
jmp-14,j,plutoz:TO..PLUTO
jmp-15,j,mercuryz:TO..MERCURY
#
# EARTH
#
:Moonp.3d,p,"MOON","SOL","EARTH MOON"
:Earth.3d,p,"EARTH","SOL","PLANET 3/9"
:Earthz.3d,s,"TERRAN","SOL","SECTOR D9"
Earth,p,earth,earth,earth,temp,TERRAN,1440,12756
Moon,m,moonp,mtype07,mtype07,cold,TERRAN,38880,3476
jmp-10,j,mercuryz:TO..MERCURY
jmp-02,j,venusz:TO..VENUS
jmp-01,j,marsz:TO..MARS
jmp-04,j,jupiterz:TO..JUPITER
jmp-05,j,saturnz:TO..SATURN
jmp-03,j,uranusz:TO..URANUS
jmp-11,j,neptunez:TO..NEPTUNE
jmp-12,j,plutoz:TO..PLUTO

In both of the above, there are no scene "levels". It's all procedurally generated from pure data files. I wrote that back in the eighties, and have improved on it throughout the years, and is the tech still used in the upcoming Universal Combat - The Lyrius Conflict.

This is what the segmented world of LOD looks like.

The space (4 scenes), planetary (4 scenes), stations (4 scenes), carrier (3 scene decks) are all hand crafted in a scene/level editor because the engine is totally different from my previous engines, and was developed specifically to power this game. It fits the purpose because, though LOD takes place in a small (Syrius region) region of the IP world (see the UC map above), it was designed to be multiplayer, and so more control over the scene loading, population etc - were required due to the design of the Wide Span Global networking tech we built to handle both session and MMO based connectivity.

When the player goes from one scene to the next via a jump anomaly, a loading screen is required because it's both efficient in terms of memory, performance, but also in terms of determining pop count. The "stitching" of these scenes to give the illusion of a cohesive world, is standard fare. Nothing ground breaking about it.

And it was built this way because my long term plan for LOD is to build the ENTIRE game world (that the IP is based on) via DLC. Doing it this way, is both efficient and cost effective. So one day, in addition to Sirius (already in LOD), we could end up with the adjacent systems (Sol, Barnard's Star, Alpha Centauri, Polaris, Tau Ceti, Omicron Eridani, Procyon) in the TERRAN quadrant. Then build the other quads over the years. And THAT'S why I spent all this resources on the game, but still started off small. And doing it this way then allows me to incorporate other game play mechanics such as trading and exploration, while bringing in assets such as all the transports and capital ships (cruisers, carriers) in the IP. The end result is a much accessible - and much simpler Battlecruiser/Universal Combat game - with multiplayer. And in the end, the FPS on planets and inside stations, will carry over to the capital ships - which is why the 3-deck Starguard carrier is in LOD as a test bed in which you can do all those things - besides fly (only devs can) it. I didn't have $130 million from a bunch of gullible fools to build it. I funded it 100% on my own.

Star Citizen is doing EXACTLY what LOD is doing. All their scenes are hand-crafted in CryEngine editor, and "demand loaded" when a player needs to transition from one place to another. In the case of going from one place to another in space, this "stitching" "hidden" behind a jump transition sequence. And if by some miracle they happen to create a planetary region, it too will be hand-crafted and demand loaded - and "stitched" - in the SAME EXACT WAY. Why? Because they DO NOT have procedural generated scenes (space or planets) because the terrain surface area is just a map they generated in the editor; then populated.

NOTE: The "dev" above, pretty much confirms what I said above, but uses semantics to deny that the entire world isn't cohesive, but rather, is "stitched".

"it's true that things get *authored* as isolated scenes. That's just good sense. But this whole "stitching" thing is just lack of imagination on his part, I think. Or a setup for him to backpedal and say that because assets get loaded as you approach them, he was right all along. If all you want is a dev saying he's wrong, well, he's wrong."

Using buzz words doesn't change the facts. And "buzz words" and "semantics" are what the Star Citizen devs excel (clearly, they can't build a game after 4 yrs and $135 million) at.

Anyone who thinks that this starmap of the game's universe is one cohesive whole that isn't "stitched", is a FOOL. And if you believe that they are actually going to build (LOL!!!) all of that, and you will be moving "seamlessly" from place to place - sans "stitching" - please, do go and buy an Idris while you wait.

And NONE of the above  has ANYTHING to do with a "64-Bit world" - which Star Citizen simply does NOT have. And they haven't stated anything to the contrary either, instead, relying on obfuscation to keep backers guessing and theory-crafting bullshit on-the-fly.

Quote
Here's a recent interview with Sean Tracey that seems to suggest you're wrong as well.

The moral of this story is very simple. Just because you don't think something is viable or possible doesn't mean it can't or shouldn't be done. They've done it. Claiming otherwise just makes you look even more ridiculous than ever.

Good thing that you're not a developer, just another clueless fool who i) believes what's coming from a band of KNOWN LIARS ii) doesn't understand ANY of what Brian said in that video

HINT: No, he didn't - in any way, shape or form, disprove what I've stated.

Here's a neat trick, go ahead and show your "developer contact" this post, and have him explain to you how they are building their worlds, that it's not "stitched", doing transitions etc. I'll wait - right here.

ps: While you're at it, see this OP where I discuss some of their "tech"
« Last Edit: November 24, 2016, 10:02:39 AM by dsmart »
Star Citizen isn't a game. It's a TV show about a bunch of characters making a game. It's basically "This is Spinal Tap" - except people think the band is real.

AP

  • Smarties
  • Jr. Member
  • ******
  • Posts: 69
Re: Star Citizen - The E.L.E.
« Reply #76 on: November 24, 2016, 09:16:02 AM »
The moral of this story is very simple.

Guys asks another guy if he's been an idiot to give him money, 2nd guys says no.

ConfusedMonkeh

  • Newbie
  • *
  • Posts: 39
  • Are we having fun yet?
Re: Star Citizen - The E.L.E.
« Reply #77 on: November 24, 2016, 09:23:26 AM »
Oh ffs. Why do I bother. I never said I didn't understand Derek. I couldn't work out to adequately debunk his theory so I asked a dev to help. I haven't posted the entire reply but let's just say I was convinced, even more, that Derek is wrong by the rest.

I'm not a drone believing whatever I read that seems to agree with my point of view. I did some research. I've followed along very closely. Both sides. I believe what I do after assessing information.

I said I wasn't replying to you again. I should have meant it.

Sorry. That was to jcr. I'm in work. I'll see what you've said  when I can Derek. BRB.
« Last Edit: November 24, 2016, 09:25:15 AM by ConfusedMonkeh »

dsmart

  • Supreme Cmdr
  • Administrator
  • Hero Member
  • *****
  • Posts: 4915
    • Smart Speak Blog
Re: Star Citizen - The E.L.E.
« Reply #78 on: November 24, 2016, 09:30:16 AM »
Oh ffs. Why do I bother. I never said I didn't understand Derek. I couldn't work out to adequately debunk his theory so I asked a dev to help. I haven't posted the entire reply but let's just say I was convinced, even more, that Derek is wrong by the rest.

I'm not a drone believing whatever I read that seems to agree with my point of view. I did some research. I've followed along very closely. Both sides. I believe what I do after assessing information.

I said I wasn't replying to you again. I should have meant it.

Sorry. That was to jcr. I'm in work. I'll see what you've said  when I can Derek. BRB.

Yet, here you are. And you haven't stated WHAT I am wrong about, and WHY. You know, words are meaningless without context. So please, by all means, post this information that proves me wrong. Why not? NOTHING you've said above debunks ANYTHING I've stated. And semantics simply aren't going to cut it.
Star Citizen isn't a game. It's a TV show about a bunch of characters making a game. It's basically "This is Spinal Tap" - except people think the band is real.

jcrg99

  • Newbie
  • *
  • Posts: 8
  • Do or do not. There is no try.
Re: Star Citizen - The E.L.E.
« Reply #79 on: November 24, 2016, 11:46:34 AM »
I'm not a drone believing whatever I read that seems to agree with my point of view, but I am a drone believing whatever I read that seems to agree with my point of view.

Fixed for you.

Rogerio

  • Smarties
  • Newbie
  • ******
  • Posts: 35
  • Tickle me, Sandi.
Re: Star Citizen - The E.L.E.
« Reply #80 on: November 24, 2016, 01:31:43 PM »
I'm not a drone believing whatever I read that seems to agree with my point of view, but I am a drone believing whatever I read that seems to agree with my point of view.

Fixed for you.

Bem vindo jcrg99!  :lesnick:

Welcome jcrg99!  :snoop:

ConfusedMonkeh

  • Newbie
  • *
  • Posts: 39
  • Are we having fun yet?
Re: Star Citizen - The E.L.E.
« Reply #81 on: November 24, 2016, 02:40:50 PM »

The "stitching" thing is 100% factual and isn't even open to discussion because I am 100% correct. That's how it is being done, it's how to do it because it would be impossible to fit an entire world into memory. Why? Because it's not procedurally generated.

When you have two scenes (S1, S2) which are different, and part of the game "world", they have to be stitched somehow. It's not only efficient, but it also gives more options for scene and player management. As big as Elite Dangerous is, they do it.

The 'entire world' doesn't need to fit into memory at the same time. The memory intensive POIs are only 'authored', via LODs, as players hit the trigger points. I presume the jump points between systems will be 'stitched' via loading screens, but inside each system, there is no loading screen when in quantum drive. You're moving through the large world space. I've seen a video that has caught it happening, a ship, QD screaming past another moving at SCM speeds.

I can, right now in the small Crusader map, QD at 0.2c ish for as long as it takes for me to hit the edge, in any direction. I don't need to go to a POI. I can fly at 1000m/s or so for an hour or four in the same direction, in an engine that used to have a limit on the size of it's maps of around 8x8 km

That's new for a CryEngine game.


Quote from: dsmart
*SOME CODE STUFF*
I wrote that back in the eighties, and have improved on it throughout the years

It saddens me that you can't believe anyone could better an improved version of some code you wrote 30+ years ago.



Quote from: dsmart

Anyone who thinks that this starmap of the game's universe is one cohesive whole that isn't "stitched", is a FOOL. And if you believe that they are actually going to build (LOL!!!) all of that, and you will be moving "seamlessly" from place to place - sans "stitching" - please, do go and buy an Idris while you wait.

Agreed. But nobody does think that. That's a preposterous proposition. Each system will be 'stitched' via jump points, but inside each system is a single 'map' millions of km wide.

Quote from: dsmart

Good thing that you're not a developer, just another clueless fool who i) believes what's coming from a band of KNOWN LIARS ii) doesn't understand ANY of what Brian said in that video

HINT: No, he didn't - in any way, shape or form, disprove what I've stated.

Here's a neat trick, go ahead and show your "developer contact" this post, and have him explain to you how they are building their worlds, that it's not "stitched", doing transitions etc. I'll wait - right here.

I don't have a 'developer contact'. I messaged a developer and they replied. I don't believe I am a clueless fool. Opinions on that will vary I'm sure. Que sera sera. Brian Sean doesn't need to disprove what you've stated in that video, he's just discussing what they've done. It's there to be played, right now, in 2.5. A CryEngine map, hundreds of thousands of km wide. In the engine that can only create 8x8 or so.

Quote from: dsmart
I am a tier 1 engineer, I deal in absolutes - not semantics.

A tier one engineer must have a long list of popular and critically acclaimed games to their name! Cool.  Let me go have a look at the absolute Metacritic scores for your back catalogue. Be right back. I could do with a new game to play.
« Last Edit: November 24, 2016, 02:57:33 PM by ConfusedMonkeh »

J How

  • Smarties
  • Newbie
  • ******
  • Posts: 7
Re: Star Citizen - The E.L.E.
« Reply #82 on: November 24, 2016, 03:05:52 PM »

The "stitching" thing is 100% factual and isn't even open to discussion because I am 100% correct. That's how it is being done, it's how to do it because it would be impossible to fit an entire world into memory. Why? Because it's not procedurally generated.

When you have two scenes (S1, S2) which are different, and part of the game "world", they have to be stitched somehow. It's not only efficient, but it also gives more options for scene and player management. As big as Elite Dangerous is, they do it.

The 'entire world' doesn't need to fit into memory at the same time. The memory intensive POIs are only 'authored', via LODs, as players hit the trigger points. I presume the jump points between systems will be 'stitched' via loading screens, but inside each system, there is no loading screen when in quantum drive. You're moving through the large world space. I've seen a video that has caught it happening, a ship, QD screaming past another moving at SCM speeds.

I can, right now in the small Crusader map, QD at 0.2c ish for as long as it takes for me to hit the edge, in any direction. I don't need to go to a POI. I can fly at 1000m/s or so for an hour or four in the same direction, in an engine that used to have a limit on the size of it's maps of around 8x8 km

That's new for a CryEngine game.


Quote from: dsmart
*SOME CODE STUFF*
I wrote that back in the eighties, and have improved on it throughout the years

It saddens me that you can't believe anyone could better an improved version of some code you wrote 30+ years ago.



Quote from: dsmart

Anyone who thinks that this starmap of the game's universe is one cohesive whole that isn't "stitched", is a FOOL. And if you believe that they are actually going to build (LOL!!!) all of that, and you will be moving "seamlessly" from place to place - sans "stitching" - please, do go and buy an Idris while you wait.

Agreed. But nobody does think that. That's a preposterous proposition. Each system will be 'stitched' via jump points, but inside each system is a single 'map' millions of km wide.

Quote from: dsmart

Good thing that you're not a developer, just another clueless fool who i) believes what's coming from a band of KNOWN LIARS ii) doesn't understand ANY of what Brian said in that video

HINT: No, he didn't - in any way, shape or form, disprove what I've stated.

Here's a neat trick, go ahead and show your "developer contact" this post, and have him explain to you how they are building their worlds, that it's not "stitched", doing transitions etc. I'll wait - right here.

I don't have a 'developer contact'. I messaged a developer and they replied. I don't believe I am a clueless fool. Opinions on that will vary I'm sure. Que sera sera. Brian Sean doesn't need to disprove what you've stated in that video, he's just discussing what they've done. It's there to be played, right now, in 2.5. A CryEngine map, hundreds of thousands of km wide. In the engine that can only create 8x8 or so.

Quote from: dsmart
I am a tier 1 engineer, I deal in absolutes - not semantics.

A tier one engineer must have a long list of popular and critically acclaimed games to their name!  Let me go have a look at the absolute Metacritic scores for your back catalogue. Be right back. I could do with a new game to play.

Sorry but in no way could that be held in memory, even without assuming overhead of the engine or resource pools of loading "millions of km" in the engine, it is simply impossible without causing resource exhaustion on the server or even the client. Limitations on Cryengine (Star Citizen frankenengine) aside it would be nigh on impossible dynamically update that many objects in 3D space at that size without streaming data from the server to the client as a form of "stitching" by procedurally generating the environment around the player based on X,Y,Z.

Let's say the dev is right (doubtful), then why isn't CIG licensing out their breakthrough engine and technology to fund the game? There are game companies screaming for that sort of technology - myself included however it is outside the realms of possibility on a practical standpoint.

Source:  Two cloud architects who work on MMOs. A lone Sys Admin is holding his hands to his face and shaking his head.
« Last Edit: November 24, 2016, 03:07:31 PM by J How »

AP

  • Smarties
  • Jr. Member
  • ******
  • Posts: 69
Re: Star Citizen - The E.L.E.
« Reply #83 on: November 24, 2016, 03:44:36 PM »
You can't even fly a spaceship in Star Citizen, your avatar becomes a spaceship.  How can anyone look at 2.5 and think it's anything more than a complete joke?

Kyrt

  • Full Member
  • ***
  • Posts: 176
Re: Star Citizen - The E.L.E.
« Reply #84 on: November 24, 2016, 04:42:49 PM »
Guys asks another guy if he's been an idiot to give him money, 2nd guys says no.

I think Derek Smart brings up some good points and a number of potential technical issues that the Star Citizen developers might be or have faced.

However....I would take his assertions of 100% certainty with a pinch of salt. CIG might have developed solutions to the issues he sees, or developed their StarEngine to address them or like FD does, might be using technical or mathematical tricks to work around the issues.

I personally think development of star citizen is going better than Derek Smart suggests...but also a lot worse than CIG are letting on.

ConfusedMonkeh

  • Newbie
  • *
  • Posts: 39
  • Are we having fun yet?
Re: Star Citizen - The E.L.E.
« Reply #85 on: November 24, 2016, 04:55:07 PM »
Sorry but in no way could that be held in memory, even without assuming overhead of the engine or resource pools of loading "millions of km" in the engine, it is simply impossible without causing resource exhaustion on the server or even the client. Limitations on Cryengine (Star Citizen frankenengine) aside it would be nigh on impossible dynamically update that many objects in 3D space at that size without streaming data from the server to the client as a form of "stitching" by procedurally generating the environment around the player based on X,Y,Z.

Let's say the dev is right (doubtful), then why isn't CIG licensing out their breakthrough engine to fund the game? There are game companies screaming for that sort of technology - myself included however it is outside the realms of possibility on a practical standpoint.

Source:  Two cloud architects who work on MMOs. A lone Sys Admin is holding his hands to his face and shaking his head.

What can't be held in memory? Millions of kilometres of nothing with some POI markers and a few low LOD images of planets and a star? I think you're wrong. The 2.5 alpha which does exactly that, only over a smaller space, (which is still around five factors of ten larger than current maximum CryEngine map), thinks that you're wrong, Elite Dangerous, that does exactly that, thinks that you're wrong.

Source: Star Ctien alpha 2.5 and Elite Dangerous, both showing that it would, and does, fit in memory.

You're trying to tell me that the 2.0 SC alpha never happened. Well I'm sorry but it did.

jcrg99

  • Newbie
  • *
  • Posts: 8
  • Do or do not. There is no try.
Re: Star Citizen - The E.L.E.
« Reply #86 on: November 24, 2016, 05:22:57 PM »
What can't be held in memory? Millions of kilometres of nothing with some POI markers and a few low LOD images of planets and a star? I think you're wrong. The 2.5 alpha which does exactly that, only over a smaller space, (which is still around five factors of ten larger than current maximum CryEngine map), thinks that you're wrong, Elite Dangerous, that does exactly that, thinks that you're wrong.

Source: Star Ctien alpha 2.5 and Elite Dangerous, both showing that it would, and does, fit in memory.

You're trying to tell me that the 2.0 SC alpha never happened. Well I'm sorry but it did.

Fair enough. So, how went in your rig the epic Fleet Battles with capital ships and fighter squadrons, seamless, full PvP, space combat ship to ship and FPS combat going on inside the ships, boarding parties, pupil to planet, etc, all that in high fidelity, 10x more details than any AAA out there and in an universe with millions of players logged in at the same time? Org battles, etc etc, in this that it was the promised level of fun and fidelity that the game released by the end of 2014 should be.

Last that I heard, nobody pledged for "Star Citizen 2.0 Alpha" with whatever fancy tech that the devs are not able to even be clear if they managed to do this or that, considering that everything that runs over it is broken and messy as hell, and have not even 5% of the features or tech necessary to run the scenario described in the previous paragraph.

Don't you think that you are discussing irrelevant stuff? Who cares if the Star Citizen developers were able to build an engine capable to lead the humanity to Mars. This is not what people pledged for. Where is the tech for Star Citizen that supports the promised scenario? Tip: It does not exist.

AP

  • Smarties
  • Jr. Member
  • ******
  • Posts: 69
Re: Star Citizen - The E.L.E.
« Reply #87 on: November 25, 2016, 02:15:59 AM »
Guys asks another guy if he's been an idiot to give him money, 2nd guys says no.

I think Derek Smart brings up some good points and a number of potential technical issues that the Star Citizen developers might be or have faced.

However....I would take his assertions of 100% certainty with a pinch of salt. CIG might have developed solutions to the issues he sees, or developed their StarEngine to address them or like FD does, might be using technical or mathematical tricks to work around the issues.

I personally think development of star citizen is going better than Derek Smart suggests...but also a lot worse than CIG are letting on.

I've followed Star Citizen closely since being a backer in 2013 and starting to get my money out before the end of December 2013.  At the point of the delay of Dog Fighting Module in December 2013 it's become obvious that CIG were deceiving people to receive funding.  They didn't tell people the DFM was delayed until after the "end of LTI" sale was over when they obviously knew months in advance they wouldn't be able to meet the end of 2013 estimate.  Check on their store page today in November 2016 and LTI is still for sale, not that it means anything as there'll never be a game.

I think Derek Smart is completely accurate as to the true nature of what's going on.  2 years after the original estimated release date for the whole game, CIG are completely rewriting all the network code and dropping hints of maybe thousands of players being online in the same areas.  It's all fantasy & dreams to hype sales.

I hope when all this is over people at least have the guts to come back and say sorry to Derek for having doubted him.
« Last Edit: November 25, 2016, 02:37:12 AM by AP »

Kyrt

  • Full Member
  • ***
  • Posts: 176
Re: Star Citizen - The E.L.E.
« Reply #88 on: November 25, 2016, 02:18:31 AM »
What can't be held in memory? Millions of kilometres of nothing with some POI markers and a few low LOD images of planets and a star? I think you're wrong. The 2.5 alpha which does exactly that, only over a smaller space, (which is still around five factors of ten larger than current maximum CryEngine map), thinks that you're wrong, Elite Dangerous, that does exactly that, thinks that you're wrong.

Source: Star Ctien alpha 2.5 and Elite Dangerous, both showing that it would, and does, fit in memory.

You're trying to tell me that the 2.0 SC alpha never happened. Well I'm sorry but it did.

I'm not quite sure what exactly you are trying to argueng
As I understand Derek Smarts argument, he is saying that the game cannot be "seamless"...that it is stitched together from a number of instances and zones. That it is not one big map...because having just one big map means having to devote server resourves to tracking each and every object within it.

That could be players...it could be ships...it could be planets and moons and stations and it could be every single asteroid in an asteroid belt.

It isn't just POIs that a server needs to keep track of...it would be every single object within the instance.

I would assume clever programming tricks could reduce this number to some degree, but one of those tricks would appear likely to be instances and zoning so that the load can be shared on the servers. And of so, then there would need to be some way to stitch these instances together...two of which Derek Smart outlined.

And we have evidence that CIG does uses instances as it creates dedicated orbit instances and battle instances...so it musr have a way to stitch these together into a whole.

Either that or its developed a revolutionary new system that does away with the need for instancing...in which case it can easily license that technology for additional funding.

So I'm not ecactly sure what point you are trying to make since Toberts own description of how SC works tallies pretty much with what I umderstand Derek Smart is saying


Kyrt

  • Full Member
  • ***
  • Posts: 176
Re: Star Citizen - The E.L.E.
« Reply #89 on: November 25, 2016, 02:48:45 AM »
I've followed Star Citizen closely since being a backer in 2013 and starting to get my money out before the end of December 2013.  At the point of the delay of Dog Fighting Module delay in December 2013 it's become obvious that CIG were deceiving people to receive funding.  They didn't tell people the DFM was delayed until after the "end of LTI" sale was over when they obviously knew months in advance they wouldn't be able to meet the end of 2013 estimate.  Check on their store page today in November 2016 and LTI is still for sale, not that it means anything as there'll never be a game.

I think Derek Smart is completely accurate as to the true nature of what's going on.  2 years after the original estimated release date for the whole game, CIG are completely rewriting all the network code and dropping hints of maybe thousands of players being online in the same areas.  It's all fantasy & dreams to hype sales.

I hope when all this is over people at least have the guts to come back and say sorry to Derek for having doubted him.

I would say rather that CIG have handled the situation badly. And their current sales model is not one that encourages them to a quick development process and release date. If they can bring in money via ship sales and will lose a lot of their revenue stream on launch.....well, that's an obvious DISincentive to release the game.

Of course, CIG would say it gives them the freedom to expand the game, to develop it until it is perfect and can deliver the experience that they promised their backers.

Other publishers...Frontier for example...have made their own mistakes. EDs Season Pass for example.

The issue at hand is....
Are CIG developing a game? I believe so
Will they release this game? Again I believe so.

I do not think they are trying to "scam" anyone. But I do think that the luxury of having no investors, the lack of a must meet deadline and a steady income stream during development has removed the urgency of meeting that deadline. There is no incentive to publish the game soon, none to curtail costs or rein in ambition. It means they have the luxury of pushing the game back till everything is "right".

As for Derek Smart...I think he is too caught up in this. I think he makes some very good points about development wrt Star Citizen. But I also think he can be his own worst enemy in this debate. Too certain he is correct. Too certain that if he can't do something, noone can and therefore CIG must be wrong.

I think Derek Smart is correct when he suggests CIG is doing worse than many suppose. That they are having problems. That they may even have funding issues.

But I don't think things are as bad as he is portraying.

But neither do I think things are going as well for CIG as they say. Their new sales seem to indicate a grab for as much money as they can get. That isn't indicative of a business rolling in cash. They've had numerous delays, they've split SQ42 off in a blatant attempt to create a new revenue stream, the demos they've shown have had issues, they've allowed expectations and hype to get out of hand, they've had poor communications with their backers, we still have seen next to nothing of SQ42 - worrisome for a game supposedly due to be released in 2017 - and so on.

But i do believe they are developing the game they want...and I think they'll get the $150 million Derek Smart suggested it would take to realise their vision.

What I don't know is what the game they release will be like, what features it will have or how big it will be. I suspect they will go the MVP route and we will see a lot of stuff patch in later.


 

SMF spam blocked by CleanTalk