On Thursday it became public that Carlo Piana has joined the Oracle team as co-counsel in the EU investigation on Oracle acquiring Sun. The short introduction of Mr Piana is that he has been (still is?) legal counsel for the Free Software Foundation Europe, in particular representing the Samba project in the Microsoft anti-trust case (which we/he won!).
Carlo has blogged at length about his involvement and motives. Readers of my blog will know that I have been involved on behalf of Monty Program, and I think technically we are classified as a competitor to both Oracle and MySQL when we answer the EU questionnaires. Unlike Carlo, we have been keeping a low-key public profile, and until the case is over we will continue to do so. Except:
Carlo, I would like to warmly welcome you to the case, an experienced Free Software lawyer is a good addition to Oracle's team. You say that your motive is to help get clearance of the Sun acquisition as fast as possible. I know that during the past week you've been reading through my correspondance with
the Commission, so you should by now be aware that this is also something we are trying to achieve. If there is anything we can do to help you, you know how to contact us.
- Log in to post comments
- 14048 views
really?
Henrik,
Not that it matters much, but I don't think you have been keeping a low profile on this issue. You have written much more about this topic than Mr. Piana. But his post has a few interesting things:
1) Ingres competes with Oracle, not MySQL. And he knows this because the folks at Ingres told him. Meanwhile, this is on ingres.com -- "An Alternative to MySQL and why you should make the move now!". I don't recall MySQL ever pitching to Ingres customers.
2) Dual-licensing is moot (according to Mr. Piana). Does this mean we shouldn't sign the SCA and only worry about pushing our code to MariaDB?
Low profile...
Yes, you're right. "Keeping a low profile" is debatable.
We feel that educating the press is something that needs to be done, and in one case a journalist decided to quote me by name... so much for the low profile. Even so, getting facts right was more important in that incident.
Since I have many former collagues who's careers are influenced by this case, I do make short blogs or tweets on issues like when there is new questions from the EU, and when the deadline is. However, note that we have not in public discussed the actual arguments we submit to Brussels. (The kind of open communicaiton that would be normal in the open source ecosystem.) We feel that we need to answer honestly when the EU asks, but there is no reason to make a public mess about it.
henrik
quite interesting
Henrik, Mark,
I have been reading a lot on the Sun/ORACLE deal, and it has taken quite a few interesting turns. As an old SUN and MySQL user I would hate to see Sun and the related stuff disappear. It is quite obvious that Sun will not survive if broken up or if left alone. There could have been better options than ORACLE, but none of the publicized alternatives really were appealing. HP/EDS? Yikes. IBM? Very little incentive there besides converting a few Sun boxes to Big Blue. Microsoft? Well go figure.
From my perspective, a quick closing of the deal is essential. So Carlo, go make it happen.
However, the dual licensing issue raised by Carlo and cited by Mark is interesting. We have been tinkering with the idea of a dual licensing scheme for our BlackRay project. After reading a lot and learning about the subtle differences in European and American copyright and IP laws, the requirement for copyright assignments and other issues, I believe we will simply stick to a GPL only license.
In fact, there are many reasons to follow Carlos reasoning on dual licensing. Is the overall hassle worth the license revenue? Won't professional services and other revenue schemes provide a much better way of enabling a sustainable business model? Scalability of course is limited. Service contracts? Better choice probably.
Thoughts and comments greatly appreciated.
A much belated reply
Hi Felix
I just found this old comment of yours while cleaning up some spam on the site. In 2009 I have chosen not to reply to you since we tried to not discuss certain things in public while the EU investigation was ongoing.
The truth is that for the MySQL business the so called OEM revenue (which was the dual licensing revenue) was always a significant part. In the first years from switching to GPL, it was essentially all of the revenue. Back then it was still possible that someone would buy a proprietary license "out of habit", but that quickly faded away and for most of the 21st century that revenue was from customers that specifically wanted to avoid some GPL requirements when re-distributing MySQL as part of their own software.
The other main revenue stream for MySQL was, and still is, the so called MySQL Enterprise subscriptions. In this product the customer buys an annual support subscription but uses the MySQL software under the GPL. A commercial license is available to those who specifically ask for it. I think Oracle has reversed this so that they primarily want subscription customers to run the proprietary licensed MySQL binaries as the only supported software. This impacts your ability to discontinue the subscription, since you would then have to uninstall all those binaries (and replace with the identical GPL licensed ones) in case you don't renew the subscription. If I remember correctly Oracle softened this requirements so that support also covers GPL licensed binaries. (Which makes sense, since that is what a new customer will already be running in production.)
To "add value" to the Enterprise subscription, MySQL adopted - as one of the first practitioners - the so called open core model, where a vendor adds some closed source functionality to their open source software. The idea behind this is that when selling support alone, it is difficult to drive significant revenue from large customers - say 7 or even 8 figure deals - when even a large customer might file only a dozen or less support tickets over a year. Adding a few closed source bits was supposed to turn the selling of subscriptions back towards traditional proprietary software sales. Note that with dual licensing, the sales dynamic is already very similar to traditional proprietary software sales, since most customers considered the GPL requirements a non-option for themselves.
In 2009, when you have asked this question, both of these models were bringing in roughly half of the revenue. Obviously since the Enterprise subscriptions were the newer model, they were showing faster growth, but in fact the OEM business was also growing rapidly thanks to adoption of MySQL Cluster in the telco equipment vendor space. Still, the executives were of the belief that going forward the market for embedded databases would stagnate while SaaS and other models would take over, so growth expectations were mainly on the side of the Enterprise subscriptions business. When looking at the industry in general they have clearly been correct. Unfortunately I lack insight into how much of that has materialized into MySQL Enterprise subscription revenue.
Professional services was only a small slice, like 10% or less. You could easily increase that, but professional services have much lower margins, so while that increases revenue, there's no margin left to cover the actual R&D of the product. In the worst case professional services could then cannibalize the subscription business, if a customer prefers to buy consulting by the hour instead of support for an annual fee.