How do you handle hardware testing with limited testing capacity?
We have some clients reporting crashes on specific hardware, but have no means to physically test it. How would you handle this?
Please log in or register to post a reply.
You can always ask the hardware vendor to do some testing for you. They probably have a variety of configurations available for their own internal testing and they might be willing to help you out.
its a pain for sure, a solution would be to use the most popular hardware, because it is a bad problem that happens with software.
At the last place we actually did PC game development, the QA department was huge.
We had 20 PC’s each with different motherboards. Each of them were fitted with a hard disk docking cradle and a set of hard disks. Each hard disk had a different operating system on it (or version of OS’s, so we might have three version of XP per machine as well as windows 2000, and several versions of Linux)
Then we would have a bunch of graphics cards.
Before release a game would have to pass QA on ALL combinations
This is why I hate PC games development, give me a platform where I know what hardware I have to run on please.
The only solution I have is make sure the end user has up to date drivers for all hardware, get him to send you a list of hardware and drivers.
Then either go to their site and have a look, or buy an equivalent setup. There is no @magic bullet@
There’s a magic bullet. Refund.
:> Ouch. Laughed so much it hurt.
Do you work for EA?
Thanks for the advice everyone!
@Reedbeta that’s a great idea! @Stainless That’s a very thorough method. One that requires many resources.
one solution is use tech you know everyone has. the previous generation of tech or just flat out old tech
Can you request a crash dump? That’s how programmers I have come in contact with have solved crashes that they themselves cannot reproduce, so the crash dump told them what function and condition crashed, they add some code to prevent that condition, then send back a build to the people on the specific crashing hardware. With some luck, the crash was then gone!