Roman Rokytskyy wrote:

>> What this confirms is that Monty is one unhappy camper, no
>> longer allowed to control the destiny of "his" project and
>> kicking out against those who do.
>
> Sun claims that it has nothing to do with it and the management of
> MySQL is quite independent.

I didn't mean to imply that it was Sun's management that annoyed
Monty - it was the management he and the investors hired after
the company grew beyond the skills and interest of its founders.

> Is that sign of some small revolution
> there? It would be interesting to see the share distribution before
> MySQL was sold to Sun. This information should be available somewhere,
> but would require some time to discover.

Perhaps even more than you imagine, since the founders' shares
are held by a number of trusts. Perhaps someone skilled in the
art could identify the owners of those trusts - which may be
more than one layer deep. Moreover, the control of the company
was not directly tied to the number of shares as the investors
had contractual rights and special categories of stock. Basically,
Monty ceded control of MySQL to Marten Mickos when Marten joined
the company.

>
> Another thought: The decision to release 5.1 RC happened at the same
> time when MySQL bought Netfrastructure (Jim left to MySQL in February,
> Monty writes about feature-freeze since February 2006). That were also
> the times shortly after Oracle took over the InnoDB and MySQL
> management was looking for an appropriate response on that threat. So,
> it truly looks like a package of pure management decisions (get RC out
> of the doors, buy Netfrastructure, etc), which at the end led to the
> situation where a company with investment of less than 100 Mio. was
> sold for 1 Bln. Not bad, I'd say.

Not bad at all. Feature freeze at MySQL happens before beta, so in
fact, 5.1 entered beta shortly after MySQL bought Netfrastructure,
which was shortly after Oracle bought InnoDB. Management is pushing
hard to have more frequent releases - release constipation is a
positive-feedback loop - the later the release, the more important
it is to get every feature and bug fix into it, because the next
chance could be years away.

>
> Anyway, I'm quite interested to see how things evolve over time (e.g.
> Monty leaving to Firebird, similar to Jim going to MySQL :))
>

Monty and Jim have a fair bit in common - don't tell either of them
I said that.

>
> PS. Interesting is that we have had our first RC published in May
> 2006, and it took us 5 RCs and 15 months to get to GA. Was it an
> "premature-RC"-influenza in 2006 floating around? :)
>

That may be, though a better explanation is that there is constant
stress between shipping releases and fixing all significant bugs,
and that stress was exacerbated at MySQL by their building a new
QA group after 5.1 entered beta. That QA group is really quite
good at finding obscure race conditions, usually fatal.

Cheers,

Ann