Agreed on both points. I think the free Oracle offering is designed to encourage adoption and ease upgrade. So you start with the free Express edition, and build up a nice little business, but then performance or space becomes an issue and so the easiest course is to upgrade to Standard Edition for $$. And the developers working for you are all now familiar with Oracle and PL/SQL, so big business and government that run Enterprise edition for $$$$$ have plenty of developers to choose from.

I'm not sure what datatypes you're referring to. Enum and Set are kind of neat, but other than that both MySQL and Oracle seem to stick to the datatypes defined in the SQL standards.

Funny you should mention Java - were you aware that Oracle databases provide Java integration? It's kind of like MS SQL Server's dotNet integration. So you can do stuff like add your own Java libraries, or store Java objects directly in the database. I've never used it, but I suppose that would qualify as a "rich data type", right?

I am a DBA, and FYI there are multiple editions of Oracle. I'm not sure what use cases you were thinking of, but if you're looking for a free edition there's always Oracle Express Edition. Free to download, use and distribute, and allows databases up to 11GB. I've worked at companies that run bigger MySQL installations, but I would venture that they are less than 1% of the MySQL user base. The majority of MySQL installations are small ones to back websites, such as Wordpress installations. You could easily replace them with Oracle Express. For other use cases, there's Oracle's NoSQL database, or Oracle's In Memory database (called TimesTen for some obscure reason), and they used to market Oracle Database Lite for mobile apps.

So in summary, Oracle has a bunch of products that would compete with MySQL, and we can't understand why they don't just give MySQL away to Apache or some other foundation. Maybe they have support contracts that actually bring in some money.

versus

You're forgetting drag. Just like flying through air, flying through parts of the solar system results in drag from dust. The dust density is expected to increase when the probe reaches the inner Oort cloud, unless Voyager 1's path has angled enough above the ecliptic that it manages to miss it (I thought 35 degrees was high enough, my colleague disagrees.) If dust density increases, the drag will provide a small but continuous slowing effect. Once past the inner Oort cloud dust density will likely decrease, though no one I've worked with has a great guess of the dust density in the outer Oort cloud. It will still be non-zero though, and Voyager can't avoid the outer Oort. Added to the small but still present force of gravity from the sun (which is what keeps the Oort objects from drifting away), you have continuous drag on the craft.

We can't calculate the effect of that drag without knowing the dust density, and our estimates of the size of the Oort clouds are still rough (on the order of +-100AU last paper I read), which is why that NASA paper estimated crossing the outer edge of the Oort cloud in a range from 14K to 28K years. 14K if the Oort cloud is small and fairly dust-free, twice that long if our worst-case estimates of the density and size are correct.