matt thompson wrote:
> As a recent ex-Sun person (who worked in the Cloud group) it was clear to me (and most of my team) that MySQL was the foundation of a strategy to make Sun much less reliant on Oracle while also appealing to the new generation of internet scale apps. From that perspective MySQL was a clear competitor to Oracle (and in the start up space, was winning in DB footprint).
>
During the acquisition I talked to Jonathan, had lunch with Greg
Papadopoulus, and talked extensively with Marten. After the
acquisition, I talked with probably a dozen Sun vice presidents. I can
honestly say that both then and now, I haven't a clue why Sun bought
MySQL or why they paid so much for it (I tend to subscribe to the
bidding war school of thought).
Nobody has the slightest idea of the number of MySQL installs --
nobody. Anyone with a number is making it up. MySQL tracks downloads,
but it's impossible to tell whether a download is an install, a failed
evaluation, a competitor, another guy added to a project, or whatever.
Yes, MySQL had a presence in the low end of technology startups. But
the "scale out" strategy was so warty that major investments in
infrastructure were required for anything web scale. Simply put, MySQL
replication wasn't ACID for updates on more than the head node. This
wasn't an issue with social networking, but for more demanding
applications working around this was manpower and time intensive.
Oh, if the $25M number is accurate, it is lower than MySQL revenues when
it was acquired, assuming apples to apples accounting systems. None of
the numbers are public, so it's very hard to tell. It is safe to say
that if Sun did in fact have a strategy, it doesn't seem to be working.
> If you do the math you'll come up with MySQL revenue (from services) as a bit higher than what is being discussed here, but certainly not an order of magnitude higher, so assuming revenue isn't Larry's motivation, then start thinking about this from the perspective that Facebook, Google, LinkedIn, Twitter, and many other of the new "internet scale" services don't use Oracle or IBM DB technology, but do use either the standard MySQL implemententation or a custom derivitive.
>
Since about all MySQL sells is services, it's probably inclusive. (The
MySQL Enterprise edition is largely GPLed, so only the monitoring tools
can be considered as licensing revenue).
> Sun bought MySQL because most of the new, interesting services on the net were built on it. Larry wants it for the same reason...
>
You're probably right. But it's probably as useful to say that Ellison
wants it because Ellison want it. Or maybe he didn't until somebody
tried to take it away from him....
> As for the losing $100M/month - as much as this likely bothers Larry, right now this is mostly on the cost side. A month after the deal closes, Sun's cost structure will be radically realigned (unfortunately by laying off a large number of very good people).
>
That's true -- and sad. There has been a great deal of value destroyed
by the EC's actions.