Hacker News new | comments | show | ask | jobs | submit login
MySQL man pages silently relicensed away from GPL (mariadb.org)
398 points by endijs on June 18, 2013 | hide | past | web | favorite | 181 comments

Given the environment, you probably already know that there's a great talk called 'Fork Yeah! The Rise and Development of illumos' at https://www.youtube.com/watch?v=-zRN7XLCRhc .

The reason why it's relevant here is that this is not the first time that Oracle has closed-sourced an open-source project bought up from Sun. Oracle seems almost paraconsistent towards open-source in a very interesting way: they seem to view the GPL as a good way to get other companies to help them out without cheating (e.g. OpenJDK), but if they're going to do most of the development in-house anyway it seems they're much more willing to close the source (OpenSolaris, now apparently MySQL).

Could anyone comment on whether there's a larger pattern that I'm missing there? And are there other open-source technologies which were obtained from the Sun buyout which are also in danger which we should know about?

Sun Grid Engine used to be open source, but now Oracle Grid Engine isn't. But what's far worse, in my opinion as a user, is that Oracle made many of the existing documentation and community support resources hosted by Sun disappear.

Oracle make a huge chunk of money from support contracts. I don't have any figures to hand, but when I used to work heavily with Oracle, the support contracts far exceeded the licence costs. And then they have the epic certification fees, on-site consultants and developers, and so on.

So as much as it pains me to say this, I can totally understand why Oracle have removed those resources because it's flies in the face of their business practices:

1. build something massively complex and counter intuitive to developer for and administrate

2. employ the best salesmen to encourage potential customers that your product is the most powerful and flexible tool out there

3. then charge your confused customers ridiculous support fees because they can't manage the solutions they've just bought

This is why my heart sank when Oracle bought Sun. The two businesses couldn't have a more different software culture (Sun Enterprises were, in my opinion, were one of the best for releasing open, well documented and well designed software)

If it makes you feel any better, most of the big companies I work with are moving full-stop from Solaris on SPARC to Linux on x86-64.

People who make the purchasing decisions are finally fed up with Oracle's incompetence and utter contempt for the rest of the industry. It's a shame, but I won't really miss Sun when they're finally, mercifully put to death. I hate Oracle for what they did to that company and I wish nothing but ill toward them.

Their support is atrocious and their pricing is outrageous. We maintain a lot of Oracle Sun hardware, and while it is great gear, it isn't worth the headaches Oracle puts us through.

We've happily defined Solaris as deprecated in our environment and are shifting to Linux wherever possible.

This is near universal in all of the financial world except for some legacy systems

The problem with wishing the company death is that we know what it will do in the final times. It'll become another SCO generally making life inconvenient for everyone for many years before we finally see the back of it.

We've already seen elements of this trait with their case against Google regarding Dalvik.

What do you suggest? They are making life inconvenient for everyone now and will continue to do so. Sometimes you just have to suck it up and put up with a one time pain and get it over with.

Ideal world (i.e. never going to happen): let evil people keep buying their stuff and getting screwed by them, but keep the rest of us out of it!

Didn't the financial crisis teach you that we can't be kept out of it?

because these sorts of resources are worth money, especially if instead of self-help, you had to pay a "professional services company" to help you.

By removing it, Oracle can lay claim to more professional services clients.

Or, it could just be incompentance on their end in maintaining the server...

Interestingly enough, the response seems to be people moving towards Slurm [1], at least for newly deployed systems.

Anyone know of other open-source resource managers out there?

[1] https://computing.llnl.gov/linux/slurm/

There are several I can think of: Open Grid Scheduler, TORQUE, Condor (ugh), and Slurm.

There is DTrace software (http://en.wikipedia.org/wiki/DTrace) whose documentation also disappeared. My friend pointed out while exploring DTrace to profile ruby apps.

Open source that I know of:

OpenSolaris - OpenIndiana

OpenOffice - LibreOffice

Ksplice - Discontinued support for non Oracle linux

MySql - MariaDB

Ksplice still supports many flavors Linux and, if you use desktop Ubuntu or Fedora, you are welcome to use it for free. What you can't do is purchase new licenses for non-Oracle Linux if you aren't a grandfathered customer.

Disclaimer/Source: I work on the Ksplice team.

So, it's not really working anymore on non-Oracle Linux.

As cool as Ksplice is (and as stupid as Red Hat were not buying it), I don't know anybody that will use it now that Oracle bought it.

Hudson CI - Jenkins CI


Hopefully DotCloud isn't following the same pattern

The pattern you are missing is that Oracle, formerly Sun, paid Monty a billion dollars for MySQL, which he happily accepted. And let's not forget that MySQL was ALWAYS dual-licensed.

Sun didn't become Oracle, Oracle purchased Sun, so it wasn't Oracle who paid Monty. When Monty sold MySQL to Sun he believed them to be a good fit for the project, and they would keep it going in the spirit he hoped for. He could never have foreseen it would be sold to Oracle.

And yes, it has always been dual-licensed, but as Monty himself says: "The basic idea for our dual-licensing was this: if you bought a license then we waived the GPL restriction that you have to redistribute your code as GPL. You could change, modify, extend, distribute, and redistribute the copy in any way you wanted (but of course not change the license of the MySQL code). The license was for any version and usage of MySQL, for now and forever." [1]

[1] http://monty-says.blogspot.co.uk/2009/08/thoughts-about-dual...

The writing was on the wall for Sun at the time. Maybe Monty couldn't have predicted it would be Oracle specifically, but he should certainly have seen that the continuity of Sun's operation could not be relied upon.

You should really consider that at no time would Monty himself have received a wire transfer for a billion dollars. This "Monty got a billion dollars, what more does he want?!" thing is repeated all the time, but it's just not true. The MySQL companies were bought by Sun for a billion dollars, but Monty did not own 100% of the companies (or even close) -- MySQL was substantially owned by a bunch of VC companies and other executives, all of which got the lion's share of the money.

Since tax records in Finland are public, you can see how much Monty made from this acquisition (approximately) here:


(Ulf Michael Widenius income for 2008, ~16.8M EUR)

If they hold all the copyright on the documentation, they're free to do as they please. It was always expected that Oracle would kill the project.

MariaDB is a fork of MySQL started by the founder of MySQL after the Oracle acquisition. This is good news for them because it gives people more reason to use the fork.


I just switched from MySQL to MariaDB on a ubuntu image and it was completely painless. Took the 30 seconds to enter the commands. No config editing, and it imported my old databases. Props to the MariaDB team.

Probably why this article is on the MariaDB blog.

Someone having the freedom to do something doesn't mean they get to be free from criticism.

I wonder if that statement is actually true. Has not a single distro made fixes to the man pages throughout the years? Not a single Debian patch with spell fixes or corrections gone upstream?

Many projects require copyright assignment in case they want to relicense. I wonder if they did that here?

Yes, and this whole debacle is a great reminder of precisely why people should be so leery of Canonical demanding copyright assignment for their GPL projects.

MySQL always did, so that they could sell the commercial version.

> It was always expected that Oracle would kill the project.

Oh, Oracle does. Very, very slowly.

Monty could fix this any time he wanted... By giving back the $1bn Sun paid him for MySQL. He took the money happily enough. All this whining shows is that he lacks class.

I was at a talk Monty gave last year in Oxford, UK. He was perfectly happy with the deal with Sun as he trusted them with the project. He's less keen on Oracle, however.

I used to work for Oracle. Something I heard several times around different parts of the company was that Larry believed that open source was a good way to build products, not businesses. He was more than happy to take advantage of open source, but not willing to contribute to it. Oracle's attitude toward open source was one of many reasons I left, and one of the reasons I gave for why I was leaving.

It would be ironic if you typed that on a mac or iphone.

That's... not very fair. Apple contributes a fair amount to many random open source projects and also runs a number of important ones (Webkit, Clang, LLVM, ...)

The kernel to OS X/iOS is open source but this is not of any serious utility as far as I know :P

Aaaand the AppStore basically destroys Open Source since it doesn't allow distribution of GPL programs. If that's not FOSS hostile, I don't know what is...

GPL is not the only open source license out there.

I guarantee that practically every app you use incorporates open source code. Blocking GPL doesn't seem to have hampered the amount of open source code written for the platforms, either. Look how many libs/projects target iOS!

Hell, I consider GPL to be hostile party in that arena. As an open source author, I want people to use my code; that's it. I don't give a damn if they open source their own code base or fork. That's their choice.

I have interacted with more than enough contributors on the "more loose" licenses. Plus, by going GPL you're effectively cutting off contributions from large companies that play in the OSS space (like Google)

Apple refused to update to any GPL-3 software, including GCC and GDB. Clang/LLVM is great, but it's not a complete replacement given that it's lacking commonly-needed things like OpenMP, thread-local storage, Fortran, and others. Valgrind chronically does not work properly on OSX because the Valgrind developers don't use it and Apple hasn't invested a modicum of effort in helping.

I use BSD-2/MIT for my own work so I have to deal with not depending on libraries that are GPL-encumbered, but I still respect the GPL and think it's poor form for a platform to discriminate against it.

Wait, are you saying that Apple is hostile towards the GPL because it contributed developer resources towards GPLv2 projects and shipped hardware with preinstalled GPLv2 binaries, and refuses to spend developer resources on contributing to GPLv3 projects? And are you blaming Apple for not devoting resources to Valgrind?

From my perspective, GCC was becoming a tangled mess of code due to politically-motivated sabotage by the FSF. Nothing like LLVM bitcode would ever have made it into GCC because it would make GCC too modular, and because the FSF not only supports open-source software but opposes closed-source software, it is against modularity (since someone could potentially distribute closed-source frontend to GCC, if it were modular). This is not mere speculation, people have tried to fix architectural issues in GCC in the past and been shot down for explicitly this reason. You could even say this behavior is hostile.

But this is nothing special. This is just another iteration of the conflict between BSD and GPL.

> Apple refused to update to any GPL-3 software, including GCC and GDB. Clang/LLVM is great

To be fair the problem with GCC isn't just GPL, v3 or otherwise. Apparently its design is deliberately convoluted to reduce modularity, making it harder for people to use parts rather than the whole, and the architecture also makes dynamic compilation much more difficult to implement than LLVM's tool chain (this was one of the key drivers, rather than the GPL thing).

That explains GCC, but why the refusal to use up-to-date versions of Samba?

Apple is no more responsible for contributing to Valgrind than you or I am. I do wish Valgrind worked properly on Mac OS X, but I don't have the right to demand someone make it happen for me, especially for free.

Same with GCC and GDB. If the GPLv3 is an unacceptable contract to Apple, they have a right not to agree to it. It's not exactly an uncontroversial license.

Apple did not create the need for the App Store any more than locksmiths created the need for us all to live in locked homes. The App Store was created by a small group of Malware writers, virus creators and Spammers who exploited their freedoms to make computers a dangerous, confusing environment for the majority of its users who are also the less technically able members of society.

The fact that many geeks seem to think that the App Store reduces their freedom in any way beggars belief. You have the freedom to use other devices or create new devices that live outside this eco-system: your Grandparents and young children do not.

Get of your high-legged blinkered horse and take a look at the real users of the technology you no-doubt help create. They are exploited, patronised, ridiculed and blamed for merely wanting to share the most important technology of the last 200 years with us. Rather than try to include them we ostracise them and call them idiots.

Once again I was recently reminded why we need the App Store. I spent some of my weekend helping my ex and our 10-year-old son discover why his innocent use of his mothers' work laptop to look at a flash gaming site has caused her PhD thesis to become infected with a virus and all her emails to be sent back to her as containing malware.

Is it my ex's fault for letting our son use her work laptop? Is it his fault for not understanding the dangers of the internet? Is it her fault for not having the correct settings set or the correct anti-virus protection? Is it her department's IT guys' fault for not enforcing a stricter policy?

It's the fault of the fucking arseholes who write malware and deliberately target children and other computing novices by presenting them with ten choices of download button and flashing click-me-now popups that look like system controls.

Either that or it's my fault for not buying my sons an iPad.

I wanted to say that your answer really doesn't address what the parent was saying, but I see your point now...

Integral part of GPL is the ability for users to change their software, and Apple doesn't allow that (both for the reasons you stated and because... well, a work "monopoly" comes to mind ;).

Nicely put.

Ha. Yes, maybe I went slightly tangential due to my weekend experience but I think the core is relevant:

Some geeks accuse the App Store of eroding our freedoms (and the freedom of free software). I say that truly free software is a luxury that we geeks afford ourselves but make no real effort to share with the less-technical while we intellectually discount the mountainous problem of malicious software that makes non-geeks lives' a misery.

It is the latter that erodes the freedoms our our software not the the App Store.

Weight & measures, environmental health departments, competition commissions, ombudsmen.. etc. all erode some freedoms while protecting others. This is, sadly, a necessary trade-off.

Re-reading this thread it looks like I had originally meant to reply to the comment further down - this one:

  The App Store is openly hostile to freedom, let alone FOSS.

The App Store is openly hostile to freedom, let alone FOSS.

So is the GPL, imo.

As ridiculous as it sounds, that depends on your definition of freedom.

BSD and such allow people the freedom to do almost anything with the code. Close it, leave it open, whatever. BSD grants you the freedom to deny a freedom.

The GPL attempts to preserve that freedom for people other than the current recipient. GPL denies you the freedom to deny a freedom.

Yep, GPL preserves freedom in exactly the same way modern societies preserve freedom for their citizens. You are free, everyone who lives in the country is also free and you're not allowed to enslave other people. Of course, someone could make the argument that preventing you from owning slaves denies you some freedom. And from a narrow and purely technical side he would be right. However, which society would you say has more freedom, the one in which slavery is forbidden, or the other? I think the same argument applies to FOSS.

I just want to be absolutely clear...did you compare non-GPL licenses to slavery?

No he didn't. he compared the historical objections that was made by slavers when their ability to own slaves became illegal.

The south has even songs about the "good ol' days" where slaves was working in the field. Plantation owners clearly saw the change as oppression, and a limitation on their freedom to operate ones house and businesses as one pleases.

It's called analogy. I think the same reasoning and conclusions apply in both situations, which doesn't mean they are equivalent. It's a useful example when all this "which kind of license offers more freedom" arises.

The societies we live in have answered that denying a little personal freedom in order to protect the collective offers more freedom than the one in which that doesn't happen. There's no debate about that. Then, why are we still debating this?

"the freedom to deny a freedom" is not freedom, it's power.

Yes, and the GPL is only as powerful as the copyright system that supports it. Some people publish under public domain in the US and something like the MIT license for other countries where that license is less restrictive than their particular public domain.

The freedom to have a power is also a freedom. We as a society often want to restrict such powers though.

your opinion is based on misunderstanding. The GPL says one thing: no restricting others.

You can say, "that takes away my freedom to restrict others" but that's like saying, "your law takes away my freedom to have slaves"

The GPL doesn't remove freedom except in one way, which is an unintended side-effect: incompatibility.

What the GPL actively restricts is POWER. What you want that the GPL blocks is not freedom, it's power.

  > "your law takes away my freedom to have slaves"
This does not work that way in software. More permissible licenses allow me to do whatever I want with my copy of the code. The original codes is still here, I cannot take it away.

I thought that the GPL restricted your redistribution of the code, as far as I know it still lets you do whatever you want with your own copy of the code.

Depends on whether you consider redistribution under a different license a part of "whatever you want".

What you do with your own copy, may that be adding to it, removing parts, or giving your one copy to someone else is generally not considered redistribution.

Only the creation of more copies which are then distrusted out are considered as redistribution, and to be covered by copyright.

The GPL says nothing at all about what YOU can do with YOUR copy. It says that you can't enslave OTHERS by limiting what THEY do with THEIR copies.

No, their motivation is to preserve your personal freedoms with your own software and hardware.

As a dev, I view the GPL's primary motivation to be about _forcing_ downstream developers to "contribute back". The additional freedom for a user to tinker with that software is a byproduct that isn't very compelling.

Note that I'm not trying to suggest that I don't find open source software very compelling - just that the trade off doesn't make sense to me.

I'd rather be generally trusting of downstream devs; it pays more dividends in terms of goodwill (and subjectively, contributions).

It happens that GPL encourages continued sharing, but that's not it's motivation. The motivation is strictly to prevent people from using the software in a way to restrict others.

You are perfectly free to keep your contributions to yourself and never distribute anything. GPL only jumps in when you are distributing code further downstream, as not distributing code along would refuses those users the same access and freedom to developing the program as you got.

Note that it only forces you to contribute back in the event that you not contributing back would restrict another's freedoms with their hardware or software.

Why do you think GPL is hostile to freedom? Could you elaborate please?

Apple doesn't destroy OSS. It DOES however make the GPL code unusable.

I'm okay with that: I don't really like the gpl. There is a very vibrant non-gpl OSS community around OS X and iOS.

Actually it is more like GPL does not allow software to be distributed under AppStore EULA.

It's not either specifically blocking the other, it's both. The two rules specify restrictions that are incompatible with the other. GPL says: no taking away certain freedoms. AppStore EULA says: no allowing certain freedoms.

Make your own GPL app store. People can sideload all they want on their own devices if they're really interested.

Besides, the iOS App store is pretty crappy anyways.

Isn't restricting the use of open source software to those who are "really interested" part way to destroying it?

I guess I'm just a pragmatist, and I don't believe in forcing Apple to do anything with their own properties. If they want to conduct their business in a way that is unethical but legal, so be it.

While we might bemoan the walled garden model that Apple has picked and most people's seeming readiness to go along with it, you might be surprised at how many people would be interested in a store specifically "curated" towards GPL apps.

You can simultaneously believe that something is justifiable/natural and that it is destroying open source. I think the majority of people would not jailbreak to use such a store, thus limiting the potential audience and reducing the incentive for developers to release to it.

No, they can't, unless they are willing to pay Apple a $100 a year tax for the privilege.

An infinitesimal subset of all iOS apps that use open source components: https://www.cocoacontrols.com/apps

i dont understand the rationale for not allowing GPL'ed apps. Can someone explain it to me?

The TL;DR is that the app store EULA is incompatible with the GPL.

The main issue is that the GPL stipulates that the licensor cannot restrict distribution above and beyond the GPL's terms. Because apps must also be distributed under Apple's ToS, the app author would violate the GPL (e.g. due to DRM and regional restrictions).

Basically, this is Apple protecting their asses as well as the developer. Any downstream dev could sue you for using their GPL library (or at least force you off the app store). This is why build of VLC was yanked.

The FSF has a good article on it: http://www.fsf.org/blogs/licensing/more-about-the-app-store-...

I'd like to hear more about this. Can you distribute on the official Google Play store and incorporate GPL code? Is this allowed because there is an alternate means of installing apps (side loading) which you must make available to all users?

I'm curious if a developer can incorporate GPL code in their app and make it available only on Google Play or the Amazon App Store while supplying the relevant source code (or object files) — but not making the unsigned/unprotected APK available to the public.

GPL != FOSS. And if you think AppStore is FOSS hostile, you really don't know much.

Open Source? Who cares about open source? What we're talking about is free software. The bulk of the code they write is all-rights-reserved.

Well, CUPS still did not suffered what Oracle's is doing to MySQL.

That's a wrong comparison. I think the author is complaining about Oracle taking advantage of Open Source and not giving back. I don't think he's worried about companies developing closed-source software for profit.

Correct. Anything that had some kind of open source license (particularly for acquired products in spaces where Oracle had little actual experience) required justification for why the license had to be that way to maximize the business value to Oracle. Not a single penny could leak from the purse, even on speculation that it would generate good will and lead to a better market position etc. It was solely about the money, and it felt like we were reminded of this every single day. This was in product management, so that probably had something to do with it.

Or a Dell or a Samsung or a Lenovo, any hardware, obviously, or Windows...

Dell and lenovo does not manufacture anything that requires even a driver. And the stuff they do manufacture they choose providers that actually are pretty good on providing linux support.

So who holds the copyright on these man pages? Did MySQL get copyright assignments from all contributers before Oracle bought them? If not, how do they have the authority to relicense the content?

> So who holds the copyright on these man pages? Did MySQL get copyright assignments from all contributers before Oracle bought them? If not, how do they have the authority to relicense the content?

As I recall, MySQL had a firm requirement for copyright assignment for contributions to MySQL because they dual-licensed it, using both GPL and commercial licenses, and without assignment they wouldn't have the freedom to do that without owning the copyright.

So, yeah, I would expect that they own the copyrights on everything in the man pages and can release them exclusively under closed licenses if they choose.

But of course you're also free to fork the last published open source version of the documentation.

If memory serves right, anyone contributing to MySQL (before the acquisition, probably still so) had to sign a contract giving up any copyright to them or it wouldn't even get considered. This is rather common in corporate open source projects, but I think the Free Software Foundation has similar rules for contributing, so it's not necessarily a corporate thing. Still kinda dodgy though.

It seems kind of impractical to do it any other way.

If the maintainer (whether an individual, for-profit company or a not-for-profit foundation like the ASF or FSF) doesn't hold a copyright to all of the contributed code and documentation, they couldn't release it under an updated license (like the ASL2 or GPL3) or dual-license it for compatibility with other projects without getting clearance from each and every contributor.

If this seems "dodgy", bear in mind a couple of points:

1. You're not "giving up" the rights to your contribution, you are sharing the copyright with the organization. That is, your work is your work and you're still free to do whatever you want with it. You're just giving the organization the same rights, so that work that was once "copyright andyroid" is now (independently) "copyright andyroid" and "copyright whatever-project-you-contributed-to". You don't give up any rights, you're just giving some rights to others.

2. Re-licensing a project doesn't change the license for the already released work. I.e., if Oracle decides to only use a proprietary license for all future MySQL releases, that doesn't post-facto change the license on the previous releases. You (or anyone else) remain free to take the last GPL'ed version and run with it (under the terms of the license under which it was released). As crappy as that re-licensing decision might be, it really only impacts the stuff you use from them or contribute to them after the re-licensing occurs.

> 1. You're not "giving up" the rights to your contribution, you are sharing the copyright with the organization. That is, your work is your work and you're still free to do whatever you want with it.

Sorry, but - as a general rule, this is false, at least in common law jurisdictions. Copyright is a transferable property right. If I assign copyright on a work to you, I don't have it any more. (That's a bit of a simplification, ignoring moral rights and glossing over more complex situations like if we're joint authors, but the gist is true).

Now, the FSF agreement (and most other CLAs) has a 'grant-back' clause: after you assign them your copyright, they grant to you a wide perpetual licence to use, copy, relicense, etc. etc. the software. So you can still do most[1] of what you could do before.

But that's a function of that clause, not something inherent in assigning copyright.

[1] Most -- but not all. There's one thing that having the copyright lets you do that a licence never can, and that's enforce it against third parties. You can't sue someone for breach of your copyright after you've signed it away, only the copyright owner can, and that ain't you any more.

>> 1. You're not "giving up" the rights to your contribution, you are sharing the copyright...

> Sorry, but - as a general rule, this is false, at least in common law jurisdictions.

It can work either way: It depends on the particular Contributor License Agreement being used, and what it says.

Some actually transfer copyright, some just grant the project entity a perpetual non-exclusive license to all rights.

> they couldn't release it under an updated license (like the > ASL2 or GPL3) or dual-license it for compatibility with > other projects without getting clearance from each and > every contributor.

IMHO, this is a huge advantage: it guarantees that the maintainer of the project can't run off and so something crazy. The linux kernel allows every author to retain copyright, and there's some security in that: since it is basically impossible to get clearance from even the most significant subset of copyright holders in the kernel, it will forever be GPLv2.

Look at it this way: if Torvalds were evil, he could require everybody to assign him copyright, and then take everybody's work and let corporations pay him to allow them to use it closed-source enviorments. Or even un-GPL the whole thing and sell it to somebody. (IANAL, obviously - not sure about that last one)

Point is, there is security in nobody having controlling ownership of a project,

If the maintainer (...) doesn't hold a copyright to all of the contributed code and documentation, they couldn't release it under an updated license

I'm not sure if it would hold in court, but it's fairly common for GPL projects to include "or later", specifically for that reason.

or dual-license it for compatibility with other projects without getting clearance from each and every contributor.

If you're releasing under a non-copyleft license, it shouldn't be needed, and if you're releasing under the GPL, that's (arguably) a feature, not a bug ;)

All versions of MySQL since 3.23.19 have been released under the GPL, but AFAIK the documentation has always retained its copyright under the ownership of MySQL AB, Sun and now Oracle. At no point was it ever distributed under the Creative Commons or some other free-ish documentation license.

When I refer to "the documentation" I'm mostly talking about the web docs at dev.mysql.com, which are are far more important than the man pages.

If the man pages are covered by the GPL'd may be a thorny issue. The GPL is not generally applied to documentation.

Its actually quite common that man pages are used under the GPL. Looking at Debian, almost every packages comes with man pages included, and aren't the package commonly under a single license? However, those projects who uses more complex form of documentation might put more effort in picking a suitable license for documentation.

I'm sure it even matters. GPL is copyleft. Doesn't that mean that they can not relicense the same or derived work under a more restrictive license?

No, that means you can't relicence it. The copyright owners aren't affected by the license and may do as they please.

Not being able to relicense under a more restrictive license is actually the main element of copyleft.

With an MIT or BSD licensed work, you can add more restrictive conditions, e.g. make a derived work proprietary, or add a "noncommercial use only" clause to your derived work, or whatever else you want. But with a GPL-licensed work, you cannot add additional conditions to the derived work, i.e. you must give all downstream reusers the same permissions you received. That's what's sometimes referred to as the "viral" property.

Unless, as here, you are the owner of the copyrights, in which case you can apply any license you want to the work.

> Not being able to relicense under a more restrictive license is actually the main element of copyleft.

Not being able to relicense under a license not acceptable to the authors of the particular copyleft license at issue is a main element of copyleft, but copyleft licenses often allow redistributing under more restrictive licenses that remain acceptable to the copyleft license author (e.g., the FSF encourages the use of "or any later version" language in GPL licenses, and GPLv3 is in many respects more restrictive than GPLv2; further, GPLv3 allows relicensing with specified kinds of additional terms that may in some ways be more restrictive.)

Some more examples: the LGPL is copyleft and you can convert it to the GPL. Sometimes the GPL comes with a special exception, which can also be removed. AGPLv3 is more restrictive than GPLv3, and again you can convert GPLv3 to AGPLv3.

If you consider what developers are allowed to do with a piece of software, the main element of copyleft is actually not being able to relicense under a less restrictive license on the spectrum of FOSS licenses. The least restrictive "license" of all is no license, no copyright, i.e. public domain (in the US).

Because you can convert MIT/BSD to a closed source license, but you cannot convert GPL to a closed source license, in this sense the GPL is more restrictive than MIT/BSD. You can convert MIT/BSD to GPL (less to more) but not GPL to MIT/BSD (more to less).

I understand that closed source licenses are the most restrictive of all, this is just an alternative perspective on software freedom with regard to developers.

Well they are Oracle. Perhaps they will rename it to 'TheirSQL' :-)

I know, I'm kind of a party pooper, but My, Maria and Max are the names of Monty's children. Do you see the pattern? MySQL, MariaDB, MaxDB.

Are you sure Mongo wasn't one of them? :-)

No, he rode off into the sunset after sheriff Bart freed him from Hedley Lamarr

That poor horse.

It was a donkey! ;)

Speaking of beasts of burden, I've been atop my elephant (Postgres) for about 3+ years now and I feel it's harder and harder to deal with idiosyncrasies of MySQL.

MariaDB is better, much better. And at work, when we do have clients already on InnoDB, the first choice for upgrades is to move them off MySQL onto MariaDB. All fresh work starts with Postgres. Being BSD licensed, docs and all, it's the least painful legally and technically.


So... Monty kind of sold his firstborn child to Oracle?

Thank you, I wasn't aware of that fact. I guess one should then pronounce the my part as the name is actually pronounced rather than the word my.

Oracle is killing MySQL. On purpose. This has been going on for a while, and they're not doing it because they're stupid, it's in their best interests.

Thankfully Maria is such a nice clean replacement.

So Oracle acquired Sun to stifle MySQL develop for fear that it would encroach on their own DB.

However, by blatantly damaging the product, more and more users moved (and continue to move) to better alternatives. Thus damaging the company's reputation.

However, MySQL is hardly a competitor to Oracle. Oracle 11g is meant for use at completely different type of company. I don't get it.

Time to donate to Apache. Sound familiar?

I have used MySQL and Oracle and I have found MySQL to be more reliable, more performant, easier to configure, and easier to develop against.

I believe that in 80-90+% of the cases where Oracle is deployed, it is only still being used because business executives are out of touch and don't want to admit that the massive investments they committed their companies to are no longer required.

>I have used MySQL and Oracle and I have found MySQL to be more reliable, more performant, easier to configure, and easier to develop against.

Then you're simply not an enterprise level developer. Oracle is, to this day, one of the most powerful databases out there. Granted, they have absolutely awful interface tools (which is why most people are reduced to using a buggy 3rd party app to do common DB work on Oracle), but the Database itself is world class. MySQL isn't really a competitor. It competes better with things like SQLight.

MySQL's query optimizer is a joke compared to Oracle's. If you don't do joins and just use it as a key-value store, that doesn't matter. MySQL also doesn't enforce most SQL constraints like "can't store a string in an int field" etc.

Wow, that's interesting if the SQL constraints are all completely ignored.. seems doubtful.

My experience with Oracle's query optimization was that it was non-intuitive to configure to work correctly especially with the way that the optimizer statistics config works.

MySQL will do quite useful query optimization (for joins) if you have appropriate indexes. So I think that in 80% of cases that is what matters in both Oracle and MySQL -- creating appropriate indexes.

Honestly I don't think you know what MySQL is capable of.

So Oracle acquired Sun to stifle MySQL develop for fear that it would encroach on their own DB.

Which is pretty silly, because MySQL doesn't really hold a candle to Oracle's RDBMS.

There is one metric where MySQL far surpasses Oracle's RDBMS: Developer mindshare in open-source projects. If I'm starting out learning PHP, you can bet that I'm going to be learning Maria/MySQL and not Oracle. This has the side effect that not many open source projects support Oracle. (Wordpress? Drupal? PhpBB?) In fact, I'm not aware of any PHP projects that do support Oracle's RDBMS.

There is another metric where MySQL shines: Ease of installation. On any server I come across, I can install and startup a new MySQL/MariaDB instance by just asking for `apt-get install mysql`, or whichever package manager my distribution is using. You can't install Oracle in a single command like that. I'm not aware of any distributions that carry Oracle in their repositories, making it a pain to update and manage.

Most likely explanation is that they acquired Sun for their hardware line to directly compete against system integrators like HP. MySQL, Java, etc. just came as part of the package.

Fact is that Oracle bought Sun to have control of Java. As many know, all the middleware and development tools that Oracle sells for its database are written in Java. They even tried to replace with some success Pro*C with some sort of Java from hell in the late 90's.

Interesting, thanks.

Is this even legal? they released the previous documentation under GPL v2 which has a clause stating any modified versions have to be released under the same license. This new documentation is clearly just a modification of the previous documentation. It would be interesting if it came of they did not have the right to do this. If even even one line of the documentation was written by someone else and they did not get a transfer of copyright they could get sued for doing this.

The copyright owner can license under whatever license they like. As many have said even before it was sold to Sun all contributions needed a copyright assignment before they were accepted (so they could sell the option to make closed source modifications amongst other things) so Oracle can now license as they wish.

What they can't do is revoke the existing GPL licenses so you can still develop and modify the old docs.

Having said that this is Oracle and you should probably just run for the hills (MariaDB) before Oracle make it hard to make it fully compatible with future version of MySQL by doing something else.

According to mysql bug 69512 [1] this particular issue was the result of a build process error and was not intentional.

[1]: http://bugs.mysql.com/bug.php?id=69512

This is indeed the given explanation, but why on earth was there a separate set of man pages with incorrect copyright headers?

And just in case they hadn't made their point, section 38 from the pastebin is particularly friendly:

"This documentation is NOT distributed under a GPL license."

I have been really thinking about switching my DBs from MySQL to MariaDB. Wondering if someone could help me out with a few questions..

Is it a really a drop-in replacement? Any gotchas I need to be aware off? How does MariaDB compare with MySQL performance and memory wise? Did anyone face any issues with replication?

Thank you in advance.

It all depends. When you're talking about MySQL, do you mean MySQL 5.5 (the penultimate and most stable version) or the brand new Oracle released 5.6? If you're not one to jump on the 5.6 bandwagon there isn't a lot of difference between MariaDB, XtraDB (Percona) and MySQL 5.5. To make it even more confusing, MariaDB includes Percona's XtraDB (a patched InnoDB) so in essence with MariaDB you're also getting Percona, although perhaps not the latest offerings from Percona.

Oracle MySQL 5.6 has a number of interesting new features and performance improvements that I don't believe have yet made it into either XtraDB or MariaDB.

So if you're talking about MySQL 5.5, MariaDB and XtraDB are really drop in replacements, they're all based on the same codebase, and hardly anything differs except the InnoDB engine. MariaDB and XtraDB have as few experimental InnoDB bug-fixes and performance tweaks that vanilla 5.5 does not.

> How does MariaDB compare with MySQL performance

Slower last time I tested with my app on my HW. But you should test now with your app now on your HW of course to be sure :)

What I have heard is if you have lots of CPUs MariaDB doesn't scale up as well. If you scale out, that might not matter.

Why would this be so? MariaDB started from the same source tree that MySQL had. In order for performance to drop, a Maria developer would have to intentionally commit changes that lower performance.

We're not talking about performance loss but a relative performance difference. It's at least possible that MySQL might've introduced performance gains that MariaDB didn't get. I have no idea whether this is the case.

Regressions can be accidental. Did you see how the parent stressed testing on your specific hardware with your specific code?

I haven't done any really extensive work, but I have replaced it, and it is "drop in" because it's a fork of MySQL. None of my applications had any problems interfacing and the performance is reported to be better, though honestly I don't push it to the limits to say whether that's the case for me.

Just switch to postgres. Don't take any chances.

What is Wordpress going to do if MySQL goes closed source? Switch to MariaDB, or finally start supporting more databases, such as PostgreSQL?

Wordpress doesn't have to do anything, when they people say MariaDB is a drop-in-replacement for MySQL they mean it. On the functionality front, except for details that only gearheads and MySQL DBAs care about you can view them as identical.

MariaDB certainly seems like the path of least resistance in this case.

Which means Monty has screwed everyone again. He should have let MySQL die in Oracle's cradle so the open source world could move over to a better database: PostgreSQL.

I'm already using LibreOffice. Is there a fork of VirtualBox that I should be aware of too? If I actually used any databases, I'd be switching MySQL for MariaDB right now.

There's no free fork of VirtualBox, though various Linux distributions do clean up the upstream source themselves. However, KVM has gotten quite good these days; you might give it a try.

sadly, it looks (from the website) as if KVM is host-Linux only, and I need to run VMs inside OS X.

(I know Mac OS has all sorts of proprietary VM systems like parallels and so on, but I'm a poor student to just needs to run basic Linux boxes for development purposes, not windows for games.)

Unfortunately, I can't run KVM on my Windows workstation... :(

Windows 8 comes with Hyper-V.

The nice thing about VirtualBox and VMWare is they're both cross-platform, meaning images for them will boot on different types of systems. That's what makes Vagrant (http://vagrantup.com/) so flexible and useful on teams with mixed environments.

Openoffice was transferred to the Apache foundation in 2011.


You could consider switching to a Paas - https://www.openshift.com/open-source

http://www.xenproject.org/ is nice too

You should give Apache OpenOffice a try. They've currently have an exclusive right on IBM's Lotus Symphony code, which would make a great addition to any open source office suite.

"exclusive right" and "open source" don't usually go together. What license is the Lotus Symphony code under?

It hasn't been released yet, but once it's tightly integrated with AOO it'll be put under an Apache license.

I wouldn't grab the torches and pitchforks unless/until it's clear they're relicensing to a non-free license. The GPL is written to apply to code; it doesn't make much sense to apply it to documentation.

The license they moved to is right there in the blog post, with a relevant snippet from the text-

>This software and related documentation are provided under a license agreement containing restrictions on use and disclosure and are protected by intellectual property laws. Except as expressly permitted in your license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify, license, transmit, distribute, exhibit, perform, publish, or display any part, in any form, or by any means. Reverse engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is prohibited.

I'm not sure how much clearer they could be that it's not a free license.

> ...are provided under a license agreement...

How do you know that specific license isn't the GPL?

That blurb restates common law. It does not interfere with the specific license that the documentation was covered under, it's just there to remind people.

Indeed. But there's a further blurb below that which specifically states that the manpage is not under the GPL and that restricts redistribution and modification.

Has Oracle has eroded credibility for being a good steward for any open source technology at this point? My question is how long before @Mitchellh leads an effort to fork VirtualBox to allow Vagrant to have a reliable future?

It's kind of sad seeing something digging it's own grave like that.

I've long moved to Postgres and never want to go back, but I am glad to see just how thoroughly the MySQL community up and moved on to MariaDB. Linux distros are now shipping with it as the default "MySQL".

That represents a very good effort to take an open project and keep it open. Two thumbs up for the community. Now, if we can just talk about silently munging data instead of giving proper errors...

It was a bug. Corrected. http://bugs.mysql.com/bug.php?id=69512

If you're going to switch away from MySQL anyway, I would suggest taking the hit and moving to PostgreSQL.

Switch to MariaDB but don't even think you will remove MySQL experience from your CV. As most of the CV is processed automatically you could lost few point simple because lazy HR people don't care about what's happening in the IT world.

On the other hand, taking things like MySQL and PHP off your resume will lead to working with the kind of people who don't like to use MySQL or PHP. Not a bad thing. Avoiding working for a company where an HR department screens your resume instead of a hiring manager is another plus.

Hmm, so there's a bunch of hosted MySQL providers. How long before they become MariaDB hosted providers?

Oracle, your engineers know exactly what will happen. Why don't you?

> Oracle, your engineers know exactly what will happen. Why don't you?

I think Oracle knows what they're doing. Oracle likes paying customers and I suspect that they don't make much, if anything, on MySQL users. I've been using MySQL for years and I've never knowingly paid Oracle a dime.

Why do you think they acquired MySQL in the first place?

A nice side effect of acquiring Sun, nothing more

Why did they buy MySQL engine providers Innobase and Sleepycat a few years earlier?

AFAIK Oracle didn't really have an offering in the embedded DB market before that. So at least that sort of makes sense. MySQL doesn't compete with Oracle's database, so buying it to kill it doesn't make sense.

That could explain Sleepycat, but not Innobase which was platform-specific to MySQL

This is not a well informed comment. Oracle already bought InnoBASE Oy which built InnoDB. With the acquisition of MySQL now they have the whole thing. MySQL may take a bite into Oracle's bottom line at the low end, but it really sticks it to Microsoft and SQLServer, which is why Oracle sunk so much effort into better Windows support in 5.5.

Viva Maria!

I am more interested in using MySQL for my next project now.

I know that this is not a popular opinion, but I disagree with the philosophy behind the GPL. While I think that sharing is fine (and for that there are the BSD and MIT licenses, among many others), I believe that it is very often better for both developers (and users) if users pay for the software they use. (And yes, I have heard the pro-GPL argument about how you can charge for support, but I know that's bogus in practice.)

I am not saying the GPL should be outlawed or anything like that. I'm just saying that it's not a moral ideal (the actual moral ideal is you making a sustainable living off of the code you write), and it's practically speaking a bad idea in many cases.

> I know that this is not a popular opinion

Nonsensical, inconsistent opinions aren't very popular, you're right. There is no reason to like BSD/MIT because they allow "sharing" but dislike the GPL because people should pay for software, except that one allows you to personally profit from the work of others without giving back, and the other does not.

Meanwhile, IBM makes billions in "bogus" revenue based on Linux solutions, and Red Hat had $1.3 billion in "bogus" revenue last year.

The GPL and paying for a product are completely different things and they can exist together (i.e. paying for a product that is itself GPL)

What do you find preferable about BSD and MIT licenses in this respect?

My employer's legal department loves BSD and MIT licenses because it's easier to sell software that way. When we need to update or add a new library or application, BSD/MIT licensed software get approved much more quickly. For example, we need to update the snmp package but we can't get it through legal because the license for many of the modules changed to GPLv3. A new python module would really make some things easier, but we can't get it through legal (GPL again).

I must ask, have your employer ever purchased a proprietary application/module to use in any product? In a logical world, those kind of deals should have a ever worse time going through the legal department than GPL.

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