With Cisco, you'll be on hold for 3 hours, until you read off your product serial number. Then they tell you you've reached EOL for the product in question, and offer to sell you an identical product whose only difference is the product number, at a vastly increased price. However, they won't tell you what the price is until you sign an NDA, because the gouge each customer differently.
They do this with pretty much every company they buy. Psionic and Riverhead come to mind quickly for me. The only reason they kept the Linksys brand was because they had no competing product at the time.
The real world is most users of MySQL don't care a damn about any of those. They care about which is easiest and cheapest to implement. So called MySQL experts are a dime a dozen. When you search Google for database software, you see MySQL on the first page of results, not Postgre, not MSSQL/SQL Server, and not Oracle. Lastly, other than standards zealots, who demands ACID compliance? In the real world, quality is often an afterthought.
I believe DLC is twofold. First, it means they can sell nearly empty games at full retail, and then charge you for "expansions" which just result in a full proper game (I'm looking at you Guitar Hero). Second, it means the game has virtually no resale value. It essentially guarantees a steady flow of income since you don't have to put any effort into a release product, and nobody will purchase said release product used. It's quite a clever con job, and it's legal and risk free. Cosa Nostra would be proud.
What makes you so sure he wasn't a victim of spam or the like? In the US, you have to show that the person knowingly and willingly sought out to download the images. If there's just 3, that's gonna be impossible to prove. If there's half a billion, then intent is easy to show. This was recently changed because people were spamming the hell out of other people with sick porn to try and get them in trouble. Distribution, on the other hand, ignores intent completely.
MySQL does support multi master replication, and it even has auto increment offsets. Not sure if older versions support point in time recovery, but with periodic backups (peh), replication, and query logging, you can achieve the goal quite easily.
A good DBA knows to back up his database via replication, not daily dumps.
Putting a proxy between the client and the server to handle the replication does not make Postgre horizontally scalable. Nor does doing a periodic table dump and copying it to the other machines. Postgre might be a ton more efficient than MySQL, but it is in no way scalable.
An adequate bootstrap is a contradiction in terms.