(no subject)
So I've been thinking about this whole Amazon thing and have come to a couple conclusions:
1. I don't think this is as simple as a "glitch" or a simple buglet. As they say on "Seconds to Disaster", "Disasters don't just *HAPPEN*" . They're the result of a confluence of events and factors, which, each by themselves might be harmless, but when they coincide, create one huge monstro clusterfuck. If this is fundamentally an automation-related problem, it is large and systemic.
I don't know what Amazon's IT infrastructure is like, nor do I know anything about their software.
I do know that a poorly thought-out database schema(Or one that should have been revised) which maybe 1 or 2 people in the company understand, combined with some software interfacing into the database that an entirely other couple people only really understand, combined with a low-to-mid-level support tech who's just following a script or otherwise "doing what he was told" is a recipe for disaster. All they have to do is do what they're good in in their particular area and not think about anything else and it's possible to be a keystroke or two away from completely royally fucking things up. This is not a "glitch". This is a large-scale system problem.
A "glitch" would be "Oh, we're sorry, our support tech accidentally dropped $AUTHOR's works from our Best-Seller list. I'm sorry, we're correcting this right now".
To get an even better idea of what I'm talking about, go have a look at this Wikipedia article on Database Normal Forms .
I mention this because I've seen the results of trying to apply a set of properties to a large set of items in a very large database (hundreds of thousands of items selected from a database with several million entries) where the relationships and properties were not fully insulated or normalized from each other and when a certain query was run, it ended up dragging in hundreds if not thousands of wholly unrelated items.
2. This is at least as much a PR failure as it is a technical failure, probably more the former than the latter. If this HAS been going on for months, the question of WHY Amazon hasn't dealt with it up until now is a VERY good question. The number one rule of running and supporting applications that serve thousands or millions of customers: "Keep it like the weather channel: Constant Updates".
No, really, even a "We're still sucking 11 hours in and haven't fixed it yet" is way better than absolute silence or "No comment".
Not saying anything gives grist to the rumor mill and ammunition to people who'd *LOVE* for you to actually be doing it deliberately because it plays into their own conspiracy theories about how much of an overarching corporatist rat-bastard you really are.
I'm not buying the conspiracy theories or the mega-troll theories.
This is going to turn out to be a serious bug somewhere in their database infrastructure or application(s) that use it.
1. I don't think this is as simple as a "glitch" or a simple buglet. As they say on "Seconds to Disaster", "Disasters don't just *HAPPEN*" . They're the result of a confluence of events and factors, which, each by themselves might be harmless, but when they coincide, create one huge monstro clusterfuck. If this is fundamentally an automation-related problem, it is large and systemic.
I don't know what Amazon's IT infrastructure is like, nor do I know anything about their software.
I do know that a poorly thought-out database schema(Or one that should have been revised) which maybe 1 or 2 people in the company understand, combined with some software interfacing into the database that an entirely other couple people only really understand, combined with a low-to-mid-level support tech who's just following a script or otherwise "doing what he was told" is a recipe for disaster. All they have to do is do what they're good in in their particular area and not think about anything else and it's possible to be a keystroke or two away from completely royally fucking things up. This is not a "glitch". This is a large-scale system problem.
A "glitch" would be "Oh, we're sorry, our support tech accidentally dropped $AUTHOR's works from our Best-Seller list. I'm sorry, we're correcting this right now".
To get an even better idea of what I'm talking about, go have a look at this Wikipedia article on Database Normal Forms .
I mention this because I've seen the results of trying to apply a set of properties to a large set of items in a very large database (hundreds of thousands of items selected from a database with several million entries) where the relationships and properties were not fully insulated or normalized from each other and when a certain query was run, it ended up dragging in hundreds if not thousands of wholly unrelated items.
2. This is at least as much a PR failure as it is a technical failure, probably more the former than the latter. If this HAS been going on for months, the question of WHY Amazon hasn't dealt with it up until now is a VERY good question. The number one rule of running and supporting applications that serve thousands or millions of customers: "Keep it like the weather channel: Constant Updates".
No, really, even a "We're still sucking 11 hours in and haven't fixed it yet" is way better than absolute silence or "No comment".
Not saying anything gives grist to the rumor mill and ammunition to people who'd *LOVE* for you to actually be doing it deliberately because it plays into their own conspiracy theories about how much of an overarching corporatist rat-bastard you really are.
I'm not buying the conspiracy theories or the mega-troll theories.
This is going to turn out to be a serious bug somewhere in their database infrastructure or application(s) that use it.