Apache what?

Sometimes I wonder who the worst enemies of Free Software are. And it occurs me that sometimes the worst enemies are ourselves, the proponents of it (or self-described as such).

The same thought struck me yesterday reading a press release calling for Mysql to be sold to a third party, separate from the prospect buyer of Sun Microsystem, Oracle. It says:

[Monty Widenius] believes the EU's antitrust regulator is “absolutely right to be concerned” and called on Oracle “to be constructive and commit to sell MySQL to a suitable third party, enabling an instant solution instead of letting Sun suffer much longer.

Speaking of conflict of interest, it appears annoyingly odd that Oracle is accused of  making Sun suffer through a prolonged merger procedure that should have been cleared outright, was it not for some ill-advised complaints, including some coming from the same source now calling for splitting up. But this is barely the point. 

The press release goes on pointing to a document which calls for re-enabling disruptive innovation in the market. Because of my position as one of the counsels to Oracle I have of course read this document before, but equally of course I was not at liberty to discuss it openly. Now that it has been published, I feel obliged to pinpoint some issues that disconcert me. Incidentally, I am entirely in favour of disruptive innovation, only I don't see how it is relevant in the discussion. Anyway…

This “disruptive” document seems to be written by an opponent, not by a proponent of Free Software. It goes as far as saying:

On the current basis, the fork vendor would be bound by the terms of the GPL, and the fork product would be affected by “copyleft“. Consequently, the users/customers of the fork would also be bound by the GPL, and any derivative works they would create (by embedding the code of the fork product) would equally be “infected“, i.e., subject to the requirement to publish the code of the derivative work under GPL terms.

Now wait a minute. You really meant to say “Infected“?

I thought this language is a remnant of a long-dead past, something that you only find in five-years old documents. There was a time when the GPL was called with all sort of names by you-know-who. That including “cancer” and “communism“. Yet nobody — as far as I can tell — has ever called a Free Software, copyleft license “infectious”. Indeed it is common to hear and read “viral” effect, but that naming is less crude, and it could be arguably used in good faith (like in “viral marketing”).

Just a matter of style? Not quite.

The consequence of such a premise cannot be other than suggesting a remedy that imposed to Oracle to release MySQL under a non-copyleft license like the Apache Public License, as a condition to get the green light to buy Sun:

An open source license meeting the above and other reasonable criteria is the Apache License, which is also used for certain projects by well-known IT companies including IBM and Sun Microsystems (Glassfish and other middleware products).

So what's the issue with Apache, isn't it a Free Software license? It is, but on one side there are issues concerning the compatibility with the GPL, on the other side, well, it is an entirely non-copyleft license. In other words, a “permissive” one. Have I ever said I am against this kind of licenses? “Liberal” might sound well, but it rhymes withproprietarization“. Because there is no copyleft effect, nothing prevents anybody from taking it and turning into a proprietary product. And that is not unintentional, as the rationale behind it is to allow for a dual licensing strategy, that is to say, to permit proprietary exploitation of the codebase. Anybody is free to do the math and see where we are aiming at.

This is apparently old stuff, as it would appear that the last call (as per the press release) is to divest MySQL, to sell it to another company. Now, Sun has bought MySQL for one billion dollars. Anybody around with a spare billion? Seriously, I don't see any suitable prospect investor which would be able both to pay the bill for this and to safeguard MySQL as Free Software more than Oracle is.

Let's face the facts. Oracle has bought InnoDB (the most performing transactional database engine used by MySQL). Anybody predicted that the sky would have fallen, it has not. Fact is that InnoDB is more often than not the storage engine preferred by power users over the “default” MyISAM. Many have predicted that Oracle would have used its patent portfolio to squash MySQL (and surely it could have done it way before meddling with a now controversial acquisition), it has not used them a single time, unlike others. Skeptical voices have predicted that the first move of Oracle will be to stop releasing the GPL version of MySQL, and the second to stop development of MySQL. No later than last week we have heard from the very voice of Larry Ellison two strong commitments to address both concerns once for all. Can we believe this? I think we can.

The track record of Oracle in defending an ethical, healthy and competitive ecosystem speaks much better of keeping the deal as is than all conceivable alternatives or remedies. Consider for example the battle to defend open standards or Oracle taking side defending the Commission in the browser case, both of which were not in the direct area of interest of it. I have a direct and insider experience of both of them. I have spent endless hours with its executives and lawyers, I think I have quite an extensive knowledge of what is the company's motivation. This has played an important role in convincing me that there is no better choice at sight. This is the better horse we can bet on. If it's not, let's happily fork the project, the odds are that this would even benefit it.

Update: as readers would perhaps notice, I have been writing this entry in pieces and in spare time of very busy days. Now Pamela Jones on Groklaw has been more effective than I was, and used more or less the same arguments. Does it mean that great minds think alike, sheer conincidence or the fact that we have the same information based on which we reach the same conclusion? I have the answer, do you?

Tipo di Entry: 

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.