The European Commission has now got a lot of emails! It seems they even have to block some (at least GMail) just to keep their mail server alive. Thank you, all, for helping out, it is working.
As you can see, the emails already have had a small effect that Oracle had to backtrack and make some promises.
In this blog I will explain why these promises are not very helpful, even if they show that even Oracle is a little bit concerned now.
The first thing is of course that as far as I know, Oracle hasn't officially filed any remedies at all. They have only issued a press release. So whether you like their promises or not, officially they haven't promised anything to the EU. Lawyers working on the case tell me that this would be the first "horizontal merger case" ever, where the EU would accept a mere Press Release as "remedy". On the other hand, I can see how Oracle would like to set such a precedent, lowering Europe's standards for the next time...
From what I saw in Brussels last week, it is this kind of lack of respect to European law that makes this such a difficult case. From the Commission's perspective, they have no legal guarantee that Oracle will actually do what they say in a press release, so should they approve the deal or not? Though question. This is Oracle's style. They don't want to just win the case, they also want to embarrass the Commission.
Let's look at the ten promises Oracle made as a reaction to the public pressure you have all been generating.
The most significant thing is of course that everything is limited to 5 years. This is a very short time in the world of databases. (We had 1 major MySQL release in the last 4+ years.) So the storage engine vendors wouldn't be immediately "at the mercy" of Oracle, but I don't believe they would be
looking at investing heavily in technologies that compete with Oracle, since after 5 years the time is out for them.
The following "commitments" don't change anything from what MySQL does today:
7. Customer Advisory Board: This has existed for many years and is a standard method of obtaining customer feedback to guide development.
8. Storage Engine Advisory Board: Yearly conferences with storage engine developers are held (at MySQL User Conference and by invitation to Sun's internal MySQL developer meeting).
9. MySQL reference manual. The Commitment confirms that Oracle will not release this under any open source license, just continue what Sun is already doing. (And upholding one of the biggest non-free pieces of MySQL that are a hindrance to any fork.)
10. This is exactly as MySQL offers support contracts today. Maybe Oracle just wanted to add something to make it an even ten.
Excluded from the non-assertion offer
2. Non-assertion for storage engines.
The Storage engine API is pretty new and still not perfect. Most storage engine vendors in fact extend or work around the officially documented api in various ways. The way Oracle's promises are written, it wouldn't be of much help to the real world of storage engines, that actually hack around the current api in various ways.
The public commitments don't give similar non-assertion promises to applications embedding MySQL, which after all is the most typical reason to buy MySQL OEM licenses. (Storage engines are just a small particular type of OEM customer.) Given that Oracle has expressed several times that they don't believe such customers would need to buy licenses anyway, but could embed MySQL on a GPL basis... Maybe they didn't think such a promise is even necessary, but it sure would be reassuring for those MySQL customers to have a legally binding promise.
Finally, it seems Oracle would define the api(s) exempted by the non-assertion promise. So then we couldn't really enhance the api, which is badly needed.
And like I said, 5 years, or any amount of years, as a limit to such assurances is not satisfactory and customers and partners would immediately loose interest in MySQL with this promise. The only workable solution has to be perpetual and irrevocable promises.
The remaining "commitments", such as the promise to increase spending, doesn't really guarantee that the spending in engineering would be directed in areas where MySQL challenges Oracle's own database. (Such as ERP applications.)
The final thing, and this may be only that Oracle doesn't understand MySQL's business and hasn't during the past 9 months acquainted itself with it, but reading the promises for Enterprise and Community editions give a feeling Oracle is planning to start releasing the MySQL Enterprise editions with a proprietary license, where MySQL Enterprise Server today is licensed under the
GPL. (Even if not publicly available, the end customer receives it under a GPL license.)
So if you're still concerned, it is not too late to send your opinion to the Commission.