Game Discussions Star Citizen Discussion Thread v12

Any competant developer should be creating tools to check the validity and integrity of both game environment and entity data, as well as code outcomes, from the outset and especially if you have a large complex environment and population of entity types (see my post above about the number of bugs found in the product I ran the development on). If CIG are only now doing this after 10 years, then better late than never, but points to larger problems...
It is good to know why stuff is happening and why the algorithm puts out a weird value. I don't even think it is unique to professional developers.
 
Being in (late?) stage of pre-alpha after 10 years & $ 500 million does not indicate a problem. But it does indicate a whole bunch of problems.
Nobody said they were no problem with CIG. They were a lot in the past. But for this one, calling it a problem is false.
 
To create complex autotests that emulate human QA for core functionnalities, you need to have the core functionnalities almost done and stable if you don't want to redo your autotest tools at each patch.
Doing it at late stage of pre-alpha does not indicate a problem, it's the right moment to create them indeed.
If you think that is an acceptable position after 10 years of development and $0.5bil, then clearly you don't understand game development.
 
To create complex autotests that emulate human QA for core functionnalities, you need to have the core functionnalities almost done and stable if you don't want to redo your autotest tools at each patch.
Doing it at late stage of pre-alpha does not indicate a problem, it's the right moment to create them indeed.
First, you ask a question suggesting you have no experience in how it is done in the game industry.

Two posts later, you claim to know when it makes sense to develop this type of tooling in the context of game development.

Learning fast, I can see.
 
If you think that is an acceptable position after 10 years of development and $0.5bil, then clearly you don't understand game development.
Lol

First, you ask a question suggesting you have no experience in how it is done in the game industry.

Two posts later, you claim to know when it makes sense to develop this type of tooling in the context of game development.

Learning fast, I can see.
Never said I have zero knowledge in how it's made in the game industry.
And more generally, everyone can give his opinion about subjects with limited knowledge, free to anybody to correct if the opinion is erronous. If you think I'm wrong on this subject, please correct me
 
Lol


Never said I have zero knowledge in how it's made in the game industry.
And more generally, everyone can give his opinion about subjects with limited knowledge, free to anybody to correct if the opinion is erronous. If you think I'm wrong on this subject, please correct me
I think you're not only wrong but also wilfully misrepresent how things are.
 
Never said I have zero knowledge in how it's made in the game industry.
Have you worked on QA tooling at a major publisher? You mentioned AAA.
And more generally, everyone can give his opinion about subjects with limited knowledge, free to anybody to correct if the opinion is erronous. If you think I'm wrong on this subject, please correct me
The tests suite should grow with the features being implemented (or even before), not after they are "ready". If they are not tested, they are not ready. Whether the suite is manual or automatic, is another manner. Only companies awash with inexpensive cash can postpone aggressive automation.
 
So is the "hidden" core technology never going to be released, or does it just not exist?
I don't understand your question. Just look at the progress tracker to see if you find enough guys working on core tech for you. You find for example a "Subsumption Service/Server Mission Logic" task finished that had/will never had a card on the release tracker.
 
Fool of me to have not been more precise here...
As someone who uses automated tests in my work on a daily basis, I am not asking about the use of automated test. I am asking for the creation of an automated test with what appears to be a real map with NPCs/PCs interacting with objects. This type of test is normally done manually by QA testers right?

It was a bit silly of you :D

Regarding this specific test, no idea, because a) its not clear for me how they are doing this. Is this some sort of simulated test in a simulation mode or is it a test checking pass/fail in engine with some sort of script moving an actor around? Its just words without real information as to how they are doing it. Have other companies automated similar testing? No idea, there are a lot of gaming companies out there who make FPS games with interactable objects.

I can't really make a claim as to whether other game devs use manual or automated tests for this, or whether its something unique, and i doubt you can either.

EDIT:

Also, someone who uses automated tests, you know one of their weaknesses is when things change in some way, it can break your automated tests, which is why often you stick with manual testing for a lot of things, because it just makes more sense to use manual over automated. Especially in areas that are not finished, that might be changed a lot... like Tier 0 stuff, like most of what CIG have.

Have CIG jumped the gun trying to automate the testing for something that shouldn't be automated, or at least not yet?

Did someone buy Chris a book on QA Automation for Christmas and he ran into the office... realized it was empty, ran out, got on Skype and told everyone about this great idea he had to automate testing?
 
Last edited:
Thanks, that's precisely the subject. So what's your opinion about "the creation of an automated test with what appears to be a real map with NPCs/PCs interacting with objects" ?
I have no opinion on it, apart from it being one part of an incredibly padded and useless report. These updates are jargon dumps that communicate no vauable information whatsoever, particualrly when it comes to understanding how far along both games are. For example, if we were to take them at face value, CIG have been finalising the bartender for over three years now:

So how many more years until the bartender is ready? Nobody knows. Certainly not CIG.
 
Back
Top Bottom