Hacker News new | comments | show | ask | jobs | submit login
MySQL man pages license change is a bug (mysql.com)
228 points by tomaac on June 19, 2013 | hide | past | web | favorite | 56 comments

I prefer to imagine Larry Ellison wringing his hands saying "Curses! And I would have gotten away with it if it wasn't for those meddling kids!"

Also he's in an abandoned fairground for some reason.

In front of a sign saying "Scott McNealy's Open Source Wonderland"

Unfortunately, Oracle shut down the Wonderland project as well. http://en.wikipedia.org/wiki/Open_Wonderland

We've seen this twice in two days now. Facebook is blocking Tor except that maybe they're not. MySQL is now under a new license, except it's not. The alarmist headlines not waiting for a response from the newly-damned organization, instead issuing a retraction after everyone has freaked out.

Will the tech community ever get past the manufactured controversy and knee-jerk reaction we lambast the mainstream media for?

More like Facebook is blocking Tor except it was an accident, and MySQL is now under a new license except it was an accident. The idea of shaming people away from pointing out, and yes, speculating on the motives for these changes is a terrible idea for a forum. And if these things hadn't surfaced on HN, do you think they would have been corrected so quickly? What would the the "newly-damned organizations" be posting a response to?

Will HN ever get past all of the whiny meta about what other people are posting or upvoting?

My criticism is on the authors of the articles, not on the HN voters. If the headline reads "Facebook blocks Tor, putting activists at risk", that's quite a damning claim. Maybe they author should check with Facebook. The commentors here at HN saying "it doesn't surprise me that Facebook doesn't care about activists" might not be constructive, but I'm not criticizing it because it's a forum like you said.

I'm not sure I agree that these things necessarily needed to be brought up on HN in order for them to be corrected. Just reaching out to FB may have done it. The author did not do this, instead issuing a retraction after the fact. Did anyone contact the MySQL team to see if this was a mistake before lighting the torches and grabbing the pitchforks?

The problem isn't the people here on HN. The problem is the authors of these irresponsible articles; authors who know that the more sensationalist of headline they write, the more likely the community will join their frenzy without question. It's meta on the tech journalist community, not the HN community.

The article headlines are not irresponsible. "Facebook blocks Tor, putting activists at risk" is true. The fact that the block was automated (and with luck, temporary) is also true but little help to the activists using it. Same logic goes for the man pages.

If you run a project that enjoys the wide-scale usership of Facebook or MySQL then grow thick skin cause people are gonna raise flags fast when things look fishy. Because your project really fucking matters to people and mistakes hurt, deal with it.

I'd wager that this helps your product more often than it hurts it.

Kind of an aside, but still relevant to part of your post. Whining about the content and whining about the meta whining is part of any sufficiently old community. I can recall from communities I've managed, and from ones I've participated in there's /always/ metacommentary and metametacommentary. The best thing to do in these cases is not to engage it. But no, HN will never get past all the whiny meta and there's very little we can do about it except letting it slide :(

Quick Edit: This is one of the reasons why I love discourse, it encourages metadiscussion in a separate part, and while it doesn't keep it from contaminating, it seriously helps a lot to focus it and guide it.

That's one of the reasons I like Metafilter :).

benbeltrain: How does that work in Discourse?

lsiebert: How does that work on Metafilter?

Well, in discourse it's more convention than anything. But the first category created, is a `meta` category. But this explains it better: http://meta.discourse.org/t/what-is-meta/5249/2

You're completely right on the tor thing, but I think this MySQL one is different. I don't think this would have been considered a bug if there hadn't been negative backlash- if it had gone unnoticed or if no one had created a stink about it, then I don't think Oracle would have been nearly as eager to revert the change. Claiming it was just a build bug (that rearranged, and rewrote, their copyright and licensing info) seems more like a good way to back down from a bad change than anything else.

I would be willing to believe that they are planning on this change and just happened to accidentally release the change too early (and therefore can legitimately call it a bug). I would doubt they would go through all of the legal and organizational red tape just to revert the change when a couple hundred customers complained. Oracle can't and wouldn't back down that quickly.

News at 11, citizen journalists begin acting more like real journalists by jumping the gun on stories!

You heard it here first!

Isn't it more likely that in both cases, large companies changed course in response to overwhelming user feedback and are trying to cover their butts with "oops, just a bug"?

It's possible, but I would hate to assume that. I would guess that an article on HN wouldn't really be "overwheling user feedback" anyway, at least not for Facebook. During that event, it also came out that several other organizations did the same thing (Google and Hacker News, to name two) and that this was a known event. Surely reaching out to Facebook before setting their headquarters on fire would have been reasonable.

So no, I wouldn't say it's more likely. Possible, yes, but the question of 'more likely' depends on how quickly you think large companies like Facebook and Oracle are willing to change course based on a very minor PR event. I doubt it would affect their bottom line at all. The more likely event seems to be that it was us that got it wrong, not the other way around.

I found the jump to assume that Oracle decided to stick it to the MySql community telling of the community's trust of Oracle's commitment to MySql as an open source database (and not just from the echo chamber that Hacker News can be sometimes ... I was greeted with this news upon starting work this morning from two different colleagues, neither of which frequent this site).

My sense is this originates from a few perceptions of Larry Ellison and Oracle: (1) Oracle does give a whip about Open Source software. (2) Oracle charges so much for their primary RDBMS technology that it would only be natural for them to undermine MySql as an open source database technology (a view I, personally, share despite evidence to the contrary). (3) As a tech, in a past life, I've had to support Java on the client side for some really poorly written Java apps. Aside from the experience being utterly awful due to this horrible app, I have little trust for a company that installs toolbars alongside critical patches and then defends the practice when called out.

I'm finding my view of Oracle is so negative (maybe I've been supporting Oracle software for way too long?) that I'm not going to be fully convinced this was a bug until the "wrong set of copyright headers" is fixed. That's pretty sad since it makes little sense that Oracle would relicense just the man pages in some sort of conspiracy to squeeze a dime out of its customers.

Either way, it's always a good reminder of my personal favorite variation of Hanlon's razor: Where there are gaps of understanding, people jump to the worst possible conclusion.

When you combine Hanlon's razor (Never attribute to malice that which can adequately be explained by stupidity) with Clarke's law (Any sufficiently advanced technology is indistinguishable from magic) you get Grey' corollary (Any sufficiently advanced stupidity is indistinguishable from malice).

Anyone who has sufficient exposure to any large bureaucracy will have examples of the last to share. Which then raises the possibility that (as is happening currently with Bank of America) the possibility of advanced stupidity is used to cover actual malice.

...which leads us right back to Hanlon's razor (it's probably not malice, just stupidity).

But this time with the "Never" somewhat qualified.

The "Never" isn't literal in the original razor.

Yes, it is literal.

Hanlon's razor is advice about what predictions you should make (or which hypotheses you should advocate, if you like). It says that when you don't know whether something is malice or stupidity, you should never predict/theorize the former.

If the 'never' is literal, then it's attempting to establish a universal truth, and is entirely useless as that universal truth is incorrect.

If it's not literal, then it's argumentative hyperbole and should be interpreted as "usually" or "most often". In this case the truth of the statement is intact and can be a useful guide.

So either you're right and the saying is worthless, or I'm right and the saying retains useful meaning.

> If the 'never' is literal, then it's attempting to establish a universal truth

No, its not.

As a statement about predictions, its about likelihood. "Never assume malice when incompetence is plausible", even if taken literally, need not be an attempt to establish that it is universally true that when both are plausible, malice is always the correct prediction; it is a sufficient premise that there are no circumstances in which both explanatiosn are plausible where malice is more likely.

IF you CAN attribute something to stupidity, THEN YOU MUST. That's what it'd be saying if you took it literally. It's not a statement about prediction or tendency.

It does NOT say "malice is plausible" it says "malice is IMPOSSIBLE, when you CAN attribute an act to stupidity."

I'd attribute this to Heinlein's Razor.

That is the most amazing banner image I've seen on any blog ever.

It's interesting that in order to avoid the guy's head leaning over at exactly the same angle in every picture, they flipped the image on the left horizontally.

I can't decide if he's doing the Blue Steel look or Le Tigre.

Yep, the buttons on the shirt are reversed on the left image compared to the other two...

I just wanted to verify how amazing the banner image is. Hilarious.

I think you and I differ on our definitions of 'amazing.'


haha, that was a good one :). I guess everyone on HN can keep away the swords and pitch forks now !

This is typical Monty. He cries foul how awful $mysql_owner is and how they do such bad things. He is the one that sold MySQL to Sun. If anyone is to blame for the current state of MySQL affairs, it is undoubtedly him.

This report has nothing to do with Monty

And if you watch his presentations about MySQL you can see he fully accepts that he made a mistake with the way MySQL grew and he is rectifying that with MariaDB in which he has invested a significant amount personally

> started pulling in man pages with the wrong set of copyright headers

And how has the wrong set of copyright headers ended up on said pages?

The MySQL build team has a script that updates the license header in files that require it. Since the "different license" is actually the license of the documentation, from which the man pages are generated, it is possible that the script failed to update the licence text in the man pages.

Source: Ex-Oracle MySQL engineer.

Applying Hanlon's razor:

Never attribute to malice that which is adequately explained by stupidity.

I've always liked Hanlon's razor, I think it's woefully forgotten in the tech community who often reach for the pitchforks without thinking.

Witty. Now please demonstrate how it looks applied to this case.

Replace "malice" with "conspiracy to eradicate open-source MySQL" and "stupidity" with "bugs in the packaging code".

Good, good, almost there. Now elaborate on "bugs in the packaging code" part.

What does "the wrong set of copyright headers" do in a vicinity of the packaging pipeline and how it came to exist in the first place. It's easy to throw around flashy one-liners. It's a bit more work to actually try and read into the facts presented.

As has already been stated on this same page, there's a community edition and a commercial edition. They're likely part of the same packaging pipeline.

Do not be condescending on Hacker News when someone is responding to your question. It's easy to jump to conclusions. It's a bit more to actually try and understand what happened - that is the point of Hanlon's Razor. This may have been intentional, yes. But it may also have been a bug.

> Do not be condescending on Hacker News when someone is responding to your question.

Do not dispense unsolicited advice on the Internets. That wasn't a reply. It was a highbrow remark that was as generic as it was barely applicable. A proper reply was the sibling comment that pointed at two editions.

The context was implied and definitely applicable. This is now becoming pedantic so let's invoke Godwin's law and stop valuable electrons being pushed around for no reason.

The nazis did it.

Everyone happy now? :)

it's even more work to edit your package management system to include a set of documents that have nothing to do with the anything released ever before it. I would love to the commit comment.

The non-community release, maybe?

No word yet on why the source code for the MySQL documentation is no longer publicly available. Was recently and silently made unavailable. This is much worse than a few man pages.

You can download the documentation here: http://dev.mysql.com/doc/ Or are you referring to the scripts that generate it?

I'm referring to the source code that is used to build and generate the documentation available in that link. The MySQL documentation is written using the DocBook XML format.

Here is a link to an older version of the page where the documentation was available: http://web.archive.org/web/20121030130559/http://dev.mysql.c...

See "Documentation Repositories".

the only thing any mysql/mariadb news does for me is highlight how the average (and depressingly top voted) hacker news/proggit/slashdot commenter knows fuck all about either of them.

At least this served as a reminder of the possibility of such action from Oracle.

I made that comment on a phone. I mostly wish I had not.

Applications are open for YC Summer 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact