Author Topic: Star Citizen General BS  (Read 2173056 times)

ecg

  • Newbie
  • *
  • Posts: 25
Re: Star Citizen - The Game
« Reply #975 on: August 31, 2017, 11:35:42 AM »
CR seems to use the "delay is due to release needing more polish" excuse  quite a bit.
The guys from the show Mythbusters proved at one point - "you can polish a turd"

Motto

  • Hero Member
  • *****
  • Posts: 1023
Re: Star Citizen - The Game
« Reply #976 on: August 31, 2017, 12:00:49 PM »
The problem with that if he eventually produces something, it's expected to be very shiny. So whatever they release as 3.0 must be in tip-top shape. Otherwise, what have they been polishing all that time with all those people? And then people realise there wasn't much to polish to start with, so Chris lied. Now, what else might he have been lying about. And there it goes....

StanTheMan

  • Hero Member
  • *****
  • Posts: 676
Re: Star Citizen - The Game
« Reply #977 on: August 31, 2017, 07:34:53 PM »
The problem with that if he eventually produces something, it's expected to be very shiny. So whatever they release as 3.0 must be in tip-top shape. Otherwise, what have they been polishing all that time with all those people? And then people realise there wasn't much to polish to start with, so Chris lied. Now, what else might he have been lying about. And there it goes....

Yes but we have been here before.

He can churn out crap and plenty of Backers will lap it up like chocolate milk shake on a hot summers day.

Plenty will retch of course but CRoberts will be hoping it can ride that out but this cant go on forever.

« Last Edit: August 31, 2017, 07:36:55 PM by StanTheMan »

dsmart

  • Supreme Cmdr
  • Administrator
  • Hero Member
  • *****
  • Posts: 4915
    • Smart Speak Blog
Re: Star Citizen - The Game
« Reply #978 on: September 01, 2017, 10:15:03 AM »
This exchange over at SA, confirms what I wrote in my article that they once again did a GC2017 specific build


Quote
Quote
-They found an animation glitch
-Fixed it by building a workaround ramp
-Which for unknown reasons made another 100% reproducible Glitch that nobody could figure out disappear
-They call it a happy accident and decide to show that to the backers as an example of bugfixing

If this really is how they work the 3.0 demo is amazingly stable.

It's also funny that it confirms the issue council bug tacker isn't actually used internally for tracking bugs, they use zendesk instead which explains why so many of the bug videos I found, even days after being logged, had 0 views on youtube.

Also they tracked the bug on zendesk, fixed it on zendesk and notice the comment at the bottom, someone went in and double checked all the fixed bugs to make sure they were resolved specifically for the gamescom build.  Indicating they've been working on a specific build for gamecom for weeks/months, not a 3.0 build for release that will just happen to be shown at gamescom.


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.

Backer42

  • Full Member
  • ***
  • Posts: 196
  • Refundian
Re: Star Citizen - The Game
« Reply #979 on: September 01, 2017, 10:17:45 AM »
Quote
It's also funny that it confirms the issue council bug tacker isn't actually used internally for tracking bugs, they use zendesk instead which explains why so many of the bug videos I found, even days after being logged, had 0 views on youtube.
Nothing public facing at CIG has anything to do with what happens internally. It's complete opaque with a 100 % PR shell.

the_wolfmann

  • Jr. Member
  • **
  • Posts: 55
Re: Star Citizen - The Game
« Reply #980 on: September 01, 2017, 01:06:18 PM »
I've been wondering ever since I started lurking about a year ago on one topic I feel has been left largely undiscussed. I consider it to be a tad bit more technical and that's probably why no one has picked it up yet - as it is a narrative that won't be understood by the general part of the SC audience. It might've had some traction in the past but lost momentum in which case I'd be grateful if you can point me to a place where it came to passing.

The topic is: Test automation. I'll get the fact that I'm a software developer but not a game developer right out of the way. "I don't understand game development™" that much is clear and I won't insist that business software and game software should be done the same way. However, games are still a piece of software and can benefit from good practices that the software engineering field provides.

With both disclaimers out of the way... questions time:
  • Do game engines allow unit testing of relatively simple pieces of code? As an example, maybe you want to test your physics related functions to see if for specific parameters they consistently return the same result. That way if someone accidentally breaks a calculation you'd expect to catch the error long before your character starts jittering uncontrollably during actual gameplay.
  • Do we have any insider knowledge of CIG leveraging from such kind of quality control? I've seen them using JIRA screenshots excessively as of late to show that they have proper management of their feature's development (scoring them some "faith" points). JIRA lets you integrate build results that run after code changes /commits/ on a specific feature and can show the automated tests' results as passing or failing. However in any screenshots I've encountered the most information we could make out is that they're using source control branches and not much else.

If CIG can only depend on debugging code and manual QAs I fear they're in a bit of a worse pickle in the longer term (if they make it that far). I'm not bashing on manual QAs and bugsmashers, they're the unsung heroes in any kind of software project but there's only so much a human can do when a project gets more than a 100 or even thousands of features in it. Also it's quite a moral crush if you have to test the same features over and over and over just to make sure the last 3 lines of code that were changed didn't break something that was working years ago.

One can go on in much further details as to why automated testing with good coverage is all round beneficial and not a waste of time but the gist of it is already in this post. No need to post full blown chapters of books that cover the matter.
« Last Edit: September 01, 2017, 01:09:58 PM by the_wolfmann »

dsmart

  • Supreme Cmdr
  • Administrator
  • Hero Member
  • *****
  • Posts: 4915
    • Smart Speak Blog
Re: Star Citizen - The Game
« Reply #981 on: September 01, 2017, 02:34:00 PM »
Oh great. I just got the latest Star Citizen newsletter. The highlights? I'm shocked. Completely.



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.

helimoth

  • Full Member
  • ***
  • Posts: 120
Re: Star Citizen - The Game
« Reply #982 on: September 01, 2017, 02:41:05 PM »
Don't forget their latest Bar Citizen video. If you ever wanted to have a truly shit night out I've got just the place.  :lesnick:


dsmart

  • Supreme Cmdr
  • Administrator
  • Hero Member
  • *****
  • Posts: 4915
    • Smart Speak Blog
Re: Star Citizen - The Game
« Reply #983 on: September 01, 2017, 03:14:56 PM »
I've been wondering ever since I started lurking about a year ago on one topic I feel has been left largely undiscussed. I consider it to be a tad bit more technical and that's probably why no one has picked it up yet - as it is a narrative that won't be understood by the general part of the SC audience. It might've had some traction in the past but lost momentum in which case I'd be grateful if you can point me to a place where it came to passing.

The topic is: Test automation. I'll get the fact that I'm a software developer but not a game developer right out of the way. "I don't understand game development™" that much is clear and I won't insist that business software and game software should be done the same way. However, games are still a piece of software and can benefit from good practices that the software engineering field provides.

With both disclaimers out of the way... questions time:
  • Do game engines allow unit testing of relatively simple pieces of code? As an example, maybe you want to test your physics related functions to see if for specific parameters they consistently return the same result. That way if someone accidentally breaks a calculation you'd expect to catch the error long before your character starts jittering uncontrollably during actual gameplay.
  • Do we have any insider knowledge of CIG leveraging from such kind of quality control? I've seen them using JIRA screenshots excessively as of late to show that they have proper management of their feature's development (scoring them some "faith" points). JIRA lets you integrate build results that run after code changes /commits/ on a specific feature and can show the automated tests' results as passing or failing. However in any screenshots I've encountered the most information we could make out is that they're using source control branches and not much else.

If CIG can only depend on debugging code and manual QAs I fear they're in a bit of a worse pickle in the longer term (if they make it that far). I'm not bashing on manual QAs and bugsmashers, they're the unsung heroes in any kind of software project but there's only so much a human can do when a project gets more than a 100 or even thousands of features in it. Also it's quite a moral crush if you have to test the same features over and over and over just to make sure the last 3 lines of code that were changed didn't break something that was working years ago.

One can go on in much further details as to why automated testing with good coverage is all round beneficial and not a waste of time but the gist of it is already in this post. No need to post full blown chapters of books that cover the matter.

Unit testing is used in gamedev for testing small bits of code. It's not called unit testing because that's a misnomer. e.g. in my case, to test vehicle physics, I would implement, then test it on a vehicle in an empty map. Once it's all good, it goes off to testing with different parameters for the various vehicles.

Aside from devs testing their own code, CIG does have a dedicated QA + playtest team. And they use Jira. We have no idea if they use Jira to generate/run build results.



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.

dsmart

  • Supreme Cmdr
  • Administrator
  • Hero Member
  • *****
  • Posts: 4915
    • Smart Speak Blog
Re: Star Citizen - The Game
« Reply #984 on: September 01, 2017, 03:15:30 PM »
Well, the latest schedule it up. Analysis coming soon.

https://robertsspaceindustries.com/schedule-report

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.

Backer42

  • Full Member
  • ***
  • Posts: 196
  • Refundian
Re: Star Citizen - The Game
« Reply #985 on: September 01, 2017, 03:37:08 PM »
The topic is: Test automation. I'll get the fact that I'm a software developer but not a game developer right out of the way. "I don't understand game development™" that much is clear and I won't insist that business software and game software should be done the same way. However, games are still a piece of software and can benefit from good practices that the software engineering field provides.
My impression is that game development has the worst standards of software engineering. Which is understandable. Video games are not "mission-critical". If your product crashes, at most someone gets upset, but no planes fall out of the sky and no million-dollar factory stops manufacturing. OTOH "pushing the limits" (esp. performance) is key, so having correct programs without "hacky" shortcuts is less of a priority.

Talent-wise I see some negative selection bias due to the working conditions in the industry. Everyone who can applies elsewhere, only doing grunt coding work at EA & Co as a last resort. The only people who seem to get actually paid well are the voice-actors for some reason.

In the end it all boils down to cost: proper QA costs money, while the video game media don't care much about software quality. Fancy graphics are more important. Star Citizen is a perfect example.

Serendipity

  • Full Member
  • ***
  • Posts: 240
Re: Star Citizen - The Game
« Reply #986 on: September 01, 2017, 04:26:27 PM »
So what do we think CR's, 'dont show the loading screen', was about? What could possibly have been unshowable on a convention only build, loading screen?

I'm intrigued.

Aya Reiko

  • Full Member
  • ***
  • Posts: 125
Re: Star Citizen - The Game
« Reply #987 on: September 01, 2017, 11:04:26 PM »
So what do we think CR's, 'dont show the loading screen', was about? What could possibly have been unshowable on a convention only build, loading screen?

I'm intrigued.
Because the whole thing was a CryEngine level and not representative of 3.0 or any build of SC.  Showing the loading screen would've tipped everyone off.

Motto

  • Hero Member
  • *****
  • Posts: 1023
Re: Star Citizen - The Game
« Reply #988 on: September 01, 2017, 11:04:53 PM »
^ That. Or showing the loading screen would give away that is was a scripted video right from the start I'd say. I can imagine he didn't want that to show.

Kyrt

  • Full Member
  • ***
  • Posts: 176
Re: Star Citizen - The Game
« Reply #989 on: September 02, 2017, 01:01:43 AM »
I can't show you Squadron 42 because it isn't finished yet. You know this. If you can wait until Citizencon, then apparently we should get to see a lot of Squadron information. You probably also know this too.

We were supposed to get a lot of S42 info last year as well. As I is, if the S42 ingame graphics are as good as that picture, then that'll be great.

HOWEVER - while you might think Miles looks "pretty good", IMO that would have been correct 3 or 4 years ago. Right now? Today? Compared with other games?  He's getting to the point of "average". And lets not go in to the player models. Maybe I'm spoiled...

I will grant that the Idris landing did give a "WoW" factor. And I liked how the planets/moons looked. But quality wise, the characters and ships weren't special IMO and the inevitable glitches and bugs only emphasised that. The jerkiness also raises the question as to whether the current level of detailing is sustainable or if optimisation will be enough - I would assume those PCs running the demos were fairly powerful and they still struggled at times.

Quote
What about the render player faces on video screen technology they showed too? The pilot on screen in the other player's ship? That looked impressive and very cool

I think that was the point - it SOUNDS impressive. But it's also technology that has been in used since 2012. And lets be honest here...the technology demo here was NOT kind to FOIP. The character models are not detailed enough or flexible to get away with it. It was technically impressive but those gormless grins were ugly and only emphasised how artificial the entire system is.

That's also ignoring the fact that while it has the potential to be impressive, right now it simply underlines that CIG are working on cosmetic features that have no impact on gameplay or immersion and have done so without having implemented a finished game engine or netcode. CIG are going to look incredibly stupid if they have to cancel this because their netcode isn't up to the challenge.

And it might not...CIG is already spending bandwidth like water and not everyone has a gigabit link. Good quality VOIP alone would be a challenge for some players but CR also touted a VOIP system for in game communication. All that on top of his "1000 player instances" and now FOIP.

Maybe I'm getting it mixed up, but I would have assumed that getting a working engine and netcode into play would be of prime importance. That way you have some idea of what your engine can do. I understand the idea of setting the limits high and then designing the engine around that, but aiming high can't get around physical limits. There is only so much that can be done...only so much bandwidth available, only so many CPU cycles.

I've disagreed with Derek Smart plenty in the past. And I think he is too invested in seeing himself right. I've objected to the use of the word "scam" for example because that implies a degree of dishonesty which I do not believe is present. I think CIG are working hard to make the BDSSE.

BUT - just because I don't believe it is a scam doesn't mean I don't think that CIG are skewing the development priorities to maximise their current money flow. Hence, the focus is more on pushing ship sales and hyping up cosmetic features and technologies rather than game development.

I have no objection to FOIP and VOIP in principle. BUT - the engine and netcode SHOULD take precedence. Even if you wanted to build in the hooks for such technologies now, these are the type of features that should be hyped up and finalised in the Beta phase. Not the pre-Alpha. Certainly not before the netcode is in place and you have an idea of how much bandwidth you have to play around with.

But the FOIP right now is not impressive and merely underlined how much work CIG still have to do on their models. You might think it impressive but all I could think of when watching the stream was how UGLY and old those models looked.

To put it in prespective...as far as I can see, CIG worked on FOIP when they still don't have their game engine finalised. Still don't have the netcode in place. Still don't know what their bandwidths needs for gameplay will be. Still need to finalise their instancing technology. Still need work on the flight engine - what they showed last week didn't look great. What they appear to be doing is pre-Alpha, Alpha and Beta development at the same time. Working on the engine, ingame assets and game mechanics, alongside cosmetic features and then going back to rework everything when something fundamental changes.

« Last Edit: September 02, 2017, 01:17:20 AM by Kyrt »

 

SMF spam blocked by CleanTalk