WIF_Killzone -> RE: What the **** is up with the AI Opponent? (7/4/2014 1:18:58 AM)
|
I'm not done yet. Take for example the original development and beta testing. In my world, I would have had a full set of requirements to develop to. In this case the rules as coded (or whatever the game rules are called ) actually constitute pretty good requirements, albeit complex as hell. You then have to come up with the system design, program it all, unit test, the functional test and then user acceptance test (UAT). Functional testing alone, with negative testing (the reverse of what the code should do) would constitute say 5000 test cases, but probably 20000 due to complexity. Generally you run 3 iterations of those test cases, with full regression testing after each release (that fixes issues found). Think for a moment the time and money just to write the test cases, let alone execute them. Think of the people required to do it in a somewhat timely manner. In this case there is no team, no big budget. A different approach is required but one that is also industry acceptable. That's called beta testing. Sooo, a callout for beta testers is asked. The community responds. (good on them). Are they professional testers, no, but they are willing to do their best, for the love of the game. Are there test cases written that test all those hundreds of thousands of code paths, not frigging likely. They do there best. Good on them. The system (game) generally has gone through due process, bugs are reported, bugs are fixed through multiple iterations. The bugs reach a level where it is generally acceptable to release the game (i.e. no critical bugs and bug count down to an acceptable level), is there risk, yes, always is, but there really is no choice, unless you never want it. So I say, we call out the beta testors (bo) and line them up and shoot them. :) Need I go on? Everyone involved is doing there best and I think the game is awesome. What an accomplishment!
|
|
|
|