Here’s how the testing works.
First they test it in their dev environment. Pretty much to make sure the thing can even start up.
Next they push it to a beta-test environment which a few people help get some bugs reported and make suggestions.
Next they push it to testlive where the rest of the CE community can do their testing.
Finally they push it to live.
The dev environment isn’t going to find the bugs listed in this thread. There’s simply not enough people. For an update to come out bug free in this sense would take literal years between updates.
The beta as mentioned in the thread about Chosen of Asura, has a few dozen players, modders, and whatnot. Again, its not enough to fully test for an adequate amount of time. This would take months between updates and up to a little more than a year (and still wouldn’t catch everything).
Testlive is open to the entire population. Its opt in, and has the best chance of catching bugs and would take only weeks between updates to ensure all the bugs are hammered out.
Here’s the problem. That last one isn’t getting enough players. Let me explain why.
Its not enough to go ‘game broke fix it.’ Like many people in the forums do. Its not enough to say ‘this is bugged when I do this’ like even many reasonable people try to do.
What they need is reproduction steps. So that they can cause the bug in their test environment. If they can’t reproduce the bug. It effectively doesn’t exist. For example, if you have a crash when you craft an item. This is an example of bad reports:
I crash when I craft
I crash when I craft this item
I crash when I craft this item here.
You get the idea. They get progressively better. But we need to narrow it down. Try this:
I crash when crafting X item at Y station, but not Z station. I am playing on X map, at the coordinates x, y, z. When I play at Y map the crash still happens. My computer specs are xxxxx, running OS version Y.
That gives them a great starting point. It points to a particular station, and shows it is not map related. Also they get system specs so they can determine if its something with how the station interacts with the system. Say for example the report is an AMD card and they do it with an Nvidia and get no crash, but when they test it with AMD and see crashes, they know which part of the code to deal with.
But even with all that. Even with every information on reproduction, and the ability to see the crash happening. Sometimes the fix isn’t immediate. There’s no check boxes in the devkit that says ‘fix crashes’. If there were, they’d check them all. But no, they have to figure out solutions to the best of their ability. And they truly do their best. They want the game to be free of crashes, bugs, exploits, and other issues. They DO play the game too. And they DO want it to be a game people want to play.
And then there are times some bugs can’t get a fix. I’ve seen games that have been in active development for over 20 years with bugs that date back to 1999. Bugs older than some of the forum users here. Its not that people don’t care, some things just don’t have evident or even feasible solutions. It happens.