Slashdot Mirror
MySQL Reverses Decision On Closed Source
krow writes "I am very happy to be announcing that MySQL will be forgoing close sourcing portions of the MySQL Server. Kaj has the official statement in his blog. No portion of the server will be closed source including backup, encryption, or any storage engines we ship. To quote Kaj 'The encryption and compression backup features will be open source.' This is a change from what was previously posted here on Slashdot. I've posted some additional thoughts on my own blog concerning how we keep open source from becoming crippleware. Word has it that we will also have a panel at this year's OSCON discussing this topic. Contrary to the previous Slashdot discussion, this shows Sun's continued commitment to Open Source."
ZFS is open source, using Sun's CDDL license. the problem is that the CDDL isnt compatible with the GPL.
The MySQL software that was originally proposed to be closed source are portions of the online backup drivers. Each such driver has to be written in close cooperation with the developers of each storage engine. Well...
InnoDB already has an online backup tool, and even if/when they revise their tool to use this new API, it's still going to be theirs, open or closed, not the property of the MySQL Group.
Online backup of the engines for CSV, Blackhole, and Memcached doesn't even make sense. Archive already has a publicly available open source online backup tool.
Online backup makes sense for Maria, I don't see MontyW writing crippleware into his work.
How about MyISAM? I think that work is already done, but, the horse is already out of the barn, in that the online backup drivers for it are already publically available..
Looking even closer, the part that was going to be closed was not even the entire online backup driver set, but just compression and encryption. Any halfway competent developer would be able to hook in the necessary calls to azio, zlib, and openssl, and replicate the work.
So this is a big tempest over something that doesn't matter, and couldnt have happened anyway.
Plus, best practices for backup dont even use or want online backup. The Right Way to backup a real production MySQL instances is via filesystem snapshot, using something like LVM or ZFS.
As a small aside, the Slashdot headline of the original article was not entirely accurate. It wasn't the Sun executives who decided this. It was the MySQL executives. What that means, especially in light of the keynote speeches given by CEO Jonathan Schwartz and VP Rich Green, is interesting, and remains to be publically seen.
Good thing Sun was able to convince Sun to stick to Sun's official policy.
When it was announced that MySQL would be releasing some features in MySQL Enterprise and not in the community edition the original Slashdot headline was "Sun to close MySQL" or something similar.
Then Mickos (former CEO of MySQL AB and SVP of Sun Database group) comes here and says that it was MySQL's plan to do this before the acquisition by Sun and that it was in fact Sun who wanted them to release everything to the community. And if Sun had their way it would.
So now that Sun convinces Mickos to change his strategy the headline is "MySQL Reverses Decision On Closed Source"
HAHAHAHAHA
Open Source Java DAO Generator
What?
So you not only believe:
1. Sun (a corporation) makes decisions not based on what will bring in the most revenue, but based on what "fanatics" want;
You also apparently believe:
2. The Slashdot crowd has the ability to shape corporate policies to their whims.
I think a reality check is in order.
Sun/MySQL were considering a variety of licenses (including closed source ones). To the extent that comments made on Slashdot (and other online sources) made sense, they were probably taken into account. However, the final decision was undoubtedly what they thought would maximize profits. Yes, maintaining community good-will is probably part of their strategy, since it gives them free advertising (evangelism, etc.) and some free development (patch submissions, etc.).
Frankly I don't see how this isn't a victory for both open-source and MySQL. The community gets open-source code, MySQL gets development and exposure. Win-win.
Well said. Sun is not doing anything to try and keep OpenSolaris alive by locking up ZFS. Quite the contrary, BSD is picking up ZFS. Too many people want to sit around and cry about Sun "not allowing" ZFS to make its way into Linux, but at the end of the day its Linux that wants to force its terms upon everyone else.
Ok, this is nonsense. There is a license incompatibility, yes, but it is because BOTH licenses make requirements the other does not fullfill, not just the GPL. In other words , the license of ZFS does not permit using it in Linux because the GPL does not fullfill the requirements of the CDDL. SIMULATENOUSLY the GPL does not permit combining Linux with ZFS because the CDDL does not fullfill the requirements of the GPL.
There are a lot of trolls here who try to interpret this as the FSF the GPL being fanatic and Sun and the CDDL being more reasonable, the reality is that the the incompatibility arises from similar terms that exist in both licenses, namely that you cannot impose any further restrictions on derived works. Since the set of restrictions in two licenses differ they are incompatible. So basically, if you are going to consider this "a problem caused by teh GPL" then it is as much "a problem caused by the CDDL" and vice versa.
Of course bashing the GPL on slashdot is a lot more fun, but the boring reality is that both Sun and Linus have picked a license of their choice, and they turned out to be incompatible. It is either the fault of both parties or neither. You can't have your cake and eat it.