First, a note to mods. I (obviously) think this is a topic meriting it's own thread. However, I know the Dominic craziness has you folks on alert to keep forum 3 sane. This is not a "down with google", "google is stupid", or "Dominic Craziness" thread. I'm trying to get at some specific reasons for Dominic being a "slow update". If you don't agree about this being a topic by itself and choose to take action, then I'd appreciate if you could merge this thread with another rather than kill it. Thanks.
While GG was nice enough to give us a heads up on what to expect with Dominic in terms of time frame, and what would happen when, I don't think we really got an idea of why this update is taking 2 months. And we can't expect to have GG just tell us G's secrets. But that doesn't mean we can't speculate.
Q1. PR Calculation?
A1. I don't think PR takes this long to calculate. Even if they have a completely different way of doing it. So that's out.
Q2. New algo's?
A2.Well why wouldn't they just test it on a test box and keep going like normal in the production environment?
Q3. New Spam filters?
A3. Same answer as 2.
Q4. New system of rolling updates?
A4. This was my assumption until today. But then I realized, the freshbot is acting a little more like a deepbot, true, but it hasn't really gone after new content and again, why couldn't they use a test environment for this?
Q5. Transitioning to new rolling updates, new PR calculation, new algos and new spam filters?
A5. Perhaps, but again, why not use a test environment?
Do you see a pattern to these answers? Correct. It all points to the question, Why isn't this whole update in a test environment?
I don't think G screwed up. It's intentional. If not, then GG wouldn't have warned us. I think, though it's purely speculation, that they have done testing in a test environment, they felt it worked well enough to go live. But they know the testing and tweaking could not be done completely and properly just with the test environment. They needed the resources of all those PCs to test with full data and full resources in order to get it right. They just couldn't go any further with their test resources. So we are essentially witnessing a live alpha/beta test for that reason.