But is it even management? What different decision might they have done to fix the problem, given limited time and resources?
We have no idea what other issues were going on and needed fixing more than the relatively minor issues we're seeing now. It's easy to say bad QA, or bad coding, or bad management, but short of requiring delays until all bugs are fixed (which isn't gonna happen, ever) there's likely little else that could be done.
I get that people are angry because of bugs, and design decisions, but ultimately that's just par for the course for software. You can't wait until everything is fixed, because it never is. You just fix what you can in the time you have, and get the game in a state that can be released and will please most fans (which I bet is the case for DAI, including amongst PC users... forum noise is just that).
We're getting into QA software philosophy now, so move on if you're not interested. I worked at a company for 10 years that makes banking software for check sorting machines and balancing. The big companies we provided for were the Fed, Citibank, Wachovia, etc. National/international banks that process all their batches and stuff in the wee hours of the morning. If and when our software broke I was rousted out of bed at 1 in the morning and usually didn't get finished with the conference call until 6 am. The reason for this is that they had to process everything by a certain point or they would lose MILLIONS of dollars in interest overnight. We released updates for bug fixes at the rate of about once a quarter, I guess. The dev cycle was plotted out; on the tail end I had around three weeks to do QA. Smoke, regression, and bug testing; not to mention sev 1 issues that required the build to be recompiled. Three weeks was not enough time to do all of this, but frequently the builds did not get closed when they were supposed to and it crept into my QA time. So I did as much as I could with what I had. It had to go to the banks at a certain time because their staff had to test it and then they had to deploy it. My timeframe was what got et all up. We frequently paid for this with broken builds and then emergency fixes. I'd say about 50% of the time I would refuse to certify the build, which was about as useless as you can imagine. It went out the door anyways, because we promised the date. QA got ignored, builds went out, things fell apart.
No one wants to hear about a delay. The value decision is different for each industry and each audience; part of that judgement has to be which trust you want to tank: deadline or quality. Publishers are all about the former instead of the latter, yeah? The sooner they publish the sooner stakeholders make money and investors get a return. That's not a criticism, it's business. But here's the rub: the software company I worked for is about a quarter of the size it was and has maybe one or two customers left out of the dozen we had when I was there. We always made the release date, though.
QA in the gaming industry is in a bad state. Salary across the board is 40/50% lower than what it commands in the marketplace. Less clout, more hours, more crap to deal with. I never even thought about trying the gaming industry even as a gamer, I just wanted out. It's thankless, frustrating, and unappreciated.
IF companies committed to QA, they would commit to a test cycle that kept a timeframe that was reasonably scoped out. If your QA group says they need X time to test, you give them X time to test and don't let it slip. If you let it slip, that's your management call and your value judgement. That's not reality, though, because somewhere, somewhen everyone decided that the people building the stuff had way more priority/value over the people who make sure it works. Which is ridiculous and only happens to this extent in the software industry. There are bad apples in other areas, but in other industries: aircraft, automotive, machined stuff - the QA folks there are invested with the responsibility to make sure the products are not only safe, but work as designed and are quality goods (don't laugh at that okay you can laugh). But most developers and managers alike that I know would read this paragraph and scoff at the idea. "Testers!"* It's a bad mindset to have if you are concerned about quality over dates at all. Flag whatever part of that rant as unrealistic; but I stand by it as a philosophical statement about testing. I have strong opinions on it and I got out of the job (eventually) after beating my head against the brick wall for 15 years.
tldr; go to the next post
I just had eggnog, so I'm typing on a sugar rush.
*also I have zero idea what it's like at Bioware/EA, I am just speaking generally about the profession
Retour en haut
