Open Source database vendor MySQL AB released its latest incremental release last week (version 4.0.20), but according to some in the community, it still doesn’t address what some say serious licensing concerns. Its license changed from LGPL to GPL (in order to not allow commercial/closed applications to link with its free version, but to instead push to them buy a license). Unfortunately, this license change comes to conflict with other OSS licenses, like PHP’s (and maybe Perl’s too).
Latest MySQL Fails to Quiet Licensing Critics
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
If youre creating commercially licensed solutions, you should expect to have to engage in commercial licensing.
Plus, anybody is free to fork and maintain a LGPL version of mysql from the point at which the license changed.
I use MySQL in a commercial app, and would obviously rather not have to pay $495 per server. However, I find it hilarious that people rag on MySQL for (get this) being FOSS to further its own commercial purposes. Whoever is complaining, I doubt it’s Richard Stallman.
Agreed, I have a PSQL box here thats been running for a couple years with several DBs in it and it works flawlessly. Also Postgres can do a lot of things that MySQL can’t do.
Agreed, I have a PSQL box here thats been running for a couple years with several DBs in it and it works flawlessly. Also Postgres can do a lot of things that MySQL can’t do.
Why don’t you just name those items that you are sure PSQL can do that MySQL cannot? Please be serious. I have my own problems with mySQL that are well documented at: –
http://sql-info.de/mysql/gotchas.html
Please, my humble request is to make discussions here more meaningful and not just academic. What mySQL and PSQL need now is the community to “slap” a programmable GUI onto them. With it, business logic can be programmed into the forms just like those of the MS-Access/Paradox/FoxPro crowd have done for many many years.
Cb..
Wait for it…
Postgres. I have been using Postgres for a couple of years as a back end for banking software that doesn’t have the volume or uptime requirements that would require a tier-one supplier and we have never had a problem with it. I even use it for my personal site.
”
Why don’t you just name those items that you are sure PSQL can do that MySQL cannot? Please be serious. I have my own problems with mySQL that are well documented at: –
http://sql-info.de/mysql/gotchas.html ”
this stuff comes up every time there is a mysql news.however many of these are now obsolete. when was this page last updated anyway?
At the bottom of the link:
Last updated: 2004-05-01 08:56:17
Definitely out of date!
”
Definitely out of date!”
it still contains wrong information. i meant a proper update. not just making some silly changes
It’s not as easy as it could be to switch to Postgres, despite these licencing problems. MySQL runs on Windows, whereas Postgres doesn’t without an ugly hack, and that’s still a big deal for developers.
Still, after a long time of thinking MySQL was good enough (really, I don’t have big database requirements, all I want is something reliable with fulltext search) Postgres is where I’ll be trying in the future.
“Still, after a long time of thinking MySQL was good enough (really, I don’t have big database requirements, all I want is something reliable with fulltext search) Postgres is where I’ll be trying in the future.”
1) sqlite is included with php4 and integrated with php5
2)firebird is pretty good
3) ingres has just been open sourced
Mysql is still a good choice for many applications
That’s the end of MySQL for us, looks like wil be be using PostgreSQL from now on.
what attracted me to mysql in the first place was that development was as fast as its performance.
but now…development seems to have slowed down.
after coding an extensive application that uses mysql, i now long for the features that postgresql and firebird have: stored procs, subqueries, views, etc. and we’ve been hearing those promises from mysql for a few years already…
I know that every time there is a MySQL article someone has to do comparisons with Postgres and we hear all about how wonderful Postgres is and how it does everything and works flawlessly.
I however, have experienced the opposite. I have never managed to stress MySQL to the point where bugs were breaking my applications. My applications have broken Postgres twice, one of which caused me enormous stress.
Both of the bugs I found in Postgres were fixed promptly, not long after I discovered them. How they made it into a “stable” release has worried me ever since.
The problem for me with a lot of a software these days is trust. Up to now my trust in MySQL has not been misplaced (touching wood). My bad experiences with Postgres have scarred me. The gotchas page has valid pints, but many of the things attempted on the page are certainly not things I would do in a real application with any database.
That said even I am enormously impatient for MySQL to get into gear and finally release some of these now old features.
My original grip with Postgres was its text handling abilities, but that was a few year ago – I’ve had a look again and to my surprise it does most of what MySQL.
Time to move I think.
Perl has two license choices: artistic and GPL. This should preclude any conflicts with a GPL’d MySQL.
You can also use the OTAL (one true agile language)
http://www.ruby-lang.org/en/LICENSE.txt
Their software is GPL’ed, but people are bitching about it? Help me out here….
what attracted me to mysql in the first place was that development was as fast as its performance.
but now…development seems to have slowed down.
Well then, you’ll hate to hear that it’s performance has always been far, far less than myth indicates. Simple fact is, the performance window which MySQL can truely be considered fast is rather narrow.
That’s not saying that PostgreSQL is therefore, the champion. Heck, even the PostgreSQL developers will tell you to pick the best DB for the job. Which, oddly enough, is never something you’ll hear from the MySQL camp. Simple fact is, MySQL is rarely a good pick and much more often, a better solution already exists which provides better performance and more features. Why settle for MySQL when you have so many serious solutions, which are free to boot, to pick from?
Some RDBMS fun. MySQL vs PostgreSQL is like:
Pretty girl vs Hot chick with a nice car and her family is loaded.
Pinto (hope you don’t get rear-ended) vs SUV (can go offroad and carry more useful load).
When it comes to MySQL propaganda, almost every performance benchmark you’ve read is either a lie, biased or the tests were conducted by people that have no idea how to measure DB performance. As a result they draw the wrong comclusions yet are happy to share those results and conclusions with the masses. Heck, even PCMag did some tests and they completely screwed up the results. Heck, I would have a hard time purposely creating results more messed up than what they did. My point? MySQL performance, when correctly benchmarked, is usually measured to be about 2/3 of what PostgreSQL is when under load. So, if you picked MySQL because of it’s performance, you’ve not only been lied to, but you picked wrong.
Long story short, why settle for MySQL when there are so many real solutions to pick from? Among those are, PostgreSQL, Firebird, Ingres (newly open sourced), and I’m certain several others, which I just can’t remember the names of off the top of my head. Beyond that, there are many situations where using a SQL database really isn’t the best option, yet MySQL seems to find use there. Evaluate your needs and pick the best solution/tool for it. Generally speaking, MySQL is rarely, if ever, going to be that tool.
I would only release GPL libraries. If people can’t handle it, that’s their problem.
What in the GPL don’t you like? That you have to give away source code? Awwww. Boohoo. So sad.
Y’know, you only HAVE to give source code to the people who buy your products.
The best licence in my opinion is BSD. As I understand it with this licence you can do whatever you want with the code. Their are no restrictions at all.
“Their software is GPL’ed, but people are bitching about it? Help me out here….”
Easy case, brother: before, they could create closed source applications that used MySQL without having to buy a new license.
Now, they either have to give away their source, or stay closed but pay for a license – kinda like Micro… erm.
Its a PITA for people who were doing the former, and can’t upgrade to the latest and greatest without doing the latter, y’know?
@ urddd:
> The best licence in my opinion is BSD. As I
> understand it with this licence you can do
> whatever you want with the code. Their are no
> restrictions at all.
Bull. If there were no restictions, what is the BSD license text about, then?
The only “unrestricted license” is Public Domain.
@ all others:
Nice to see how “being Open Source” was mutilated in your heads to mean “being GPL”. Some people don’t like the GPL, and rather release their software under other Open Source licenses, like BSD, Artistic, PD or whatever.
They no longer can link with MySQL without being forced to switch to a license they don’t like.
Freedom of choice. Hell, yeah.
> Bull. If there were no restictions, what is the BSD
> license text about, then?
You can do whatever you want with the code provided that you give credits to original author(s). There are no restrictions about what you can do with the code.
Can I rip out a class / module / function and use it in a different project, without having to copy the whole shebang of copyright / credits / disclaimers?
The answer is a crystal-clear “it depends”.
With PD, the answer is “yes”. And most people are courteous enough to still give credit to the PD authors they copied from.
The rest will rip off BSD code without giving credit just as well.
The following quote is at best misleading
“Fedora Project community spokesperson Jack Aboutboul won’t be switching to PostgresSQL. “At this time, migration to Postgres is unnecessary,” he said. “Users can continue to use MySQL 3.x, and hopefully by FC3, this issue will be resolved. In a sentence, it’s not that we don’t want to include MySQL, it’s just that nothing was released using the license exception that’s usable to us.””
Fedora/Red Hat never switched to MySQL in the first place – it added MySQL as a (non-default) option, Postgresql has always been the main DB in a RH/Fedora Install
To me this is a very good example of a commercial failure of the GPL. MySQL is trying to sell a software product that is GPL’ed and has to jump trough all kinds of hoops to accomplish that. Under the GPL the software itself has zero value. You could try to sell support for GPL’ed software like Red Hat does or you could also sell proprietary software like Novell. But what MySQL tries to do won’t really work. Besides, this whole mess is based on the false assumption of the FSF that linking with a GPL’ed program makes the linking program also GPL. That will most likely not hold up if brought to court.
If you choose the GPL for your software because you believe in the goals of the FSF that’s fine. But if you want to commercially exploit your code the proprietary way is imo the only good way.
“Freedom of choice. Hell, yeah.”
You are not forced to use MySQL. If the shoe doesn’t fit, get another, cheapskate. Full freedom of choice.
I’m more than happy using my Object Prevalence system, It may not scale ad infinitu,,but it’s thousand of times faster than mysql, and much more easy than any persistence layer or hand crafted DAO startegy I’ve seen.
Perhaps you should tell that to the Trolltech people, hmm?
“Besides, this whole mess is based on the false assumption of the FSF that linking with a GPL’ed program makes the linking program also GPL. That will most likely not hold up if brought to court.”
If you’re talking about dynamic linking then i would agree, but not static linking.
Your little rant about PD has a couple of minor problems.
1. There is no such things as a PD license. There are many licenses which attempt to define something as being PD. But no single PD license exists.
2. PD does not exist, legally, everywhere. For example there is no legal precedent for the existence of PD. If you try to relese something as PD here in Germany you are releasing it into a nether-world of anonymous property. Even in the US PD is a very nebulous affair. Something becomes properly PD once the license it was released under expires. Our legal system has little in the way of precedents for dealing with things which were never released under a license.
3. Software without a license, in the US defacto PD, is an absolute no-no for serious developers. Although it should not work this way-if you can not document where the code comes from you will be forced to prove that you did not steal it in the event you are ever faced with litigation-and we live in a very, very litigous world.
4. PD software belongs not to everyone-but to noone. It is a noman’s land of anonymous IP. No accountability, no responsibility. “freedom” defined as the absence of regulations or limitations only capture one aspect of fredoom-the freedom to. This interpetation of freedom lends itself to abusive behavior, behavior which violates the propiety of others. A society can gleefully believe itself to be free, with such a one-sided interpretation of freedom, all the while living under an omnipresent tyranny for they failed to understand that without freedom-from, “freedom” is a hollow claim which only is best suited for empty rheteorical assertions of self-identity.
Moreover the BSD license and the GPL license are reflective of the legal atmosphere and experiences of the respective authors. The GPL has somewhat of a bite to it. This bite is borne of negative experiences. The first great opensource legal battle came into being because the AT&T employess thought that the BSD copyright notices in their code was meaningless-so they sed’ed their way through the code simply removing it, and attributing themselves as the authors.
Of course the BSD copyrights were not meaningless and the judge had to make AT&T acknowledge that much of their supposed code was in fact BSD. The philosophy behind the GPL is really rather simple, yet sublime and easy to miss. The mandate behind the GPL is that the tools with which software is built, which itself is software, must belong to the developers themselves. Most software licenses stipulate ownership as a precondition for developmental usage.
The GPL turns this around- if you use GPL source you yourself are a co-owner of that source, as a co-owner you as a developer are subject to certain restrictions(ie. you must provide the source unmodified upon distribution) and given certain entitlements(ie. the freedom to use other GPL code in your GPL software). For developers the price that GPL requires of one to pay is miniscule in contrast to the benefits it brings. It is precisely that “freedom from” aspect which is so strongly coded in the GPL. Developers are free from a software industry in which propietary toolchains dominate and dictate software development.
Now why is this freedom important ? If one traces the history of software development one quickly finds that most software authorship in the early days was quite “free”, liberal and open-most development occurred in a tiny handful of corporations and the major universities around the world-open and free code exchange was the norm, people collaborated with each other and shared there experiences without the encumberance of the whole IP issue. This is what the internet was prior to HTML. Starting in the late 70’s and really comming into full swing in the early 80’s there began something of a IP land grab.
In the course of a couple of years the benign land of defacto PD software became an ultra-contentious IP landscape with lawyers going for each others jugulars. He who wielded over the largest amount of IP could determine and dictate the software landscape. Only in the past years have we started to witness the absurd dimensions that this pervision has created-ie. companies which only exist to hold patents-threating and suing everyone and everybody because they @own@ some IP, which they neither developed, nor created.
What has made this situation so intractable is the individual commercial software developers see themselves as being in the same plight that large multinational software conglomerates are. In the name of “enterprise” and “entrepreneurs” under the rubric of “freedom” individual commercial software developers have unwittignly played the role of pawns in the grand schemes of major multinational software firms. Yet individual commercial software developers aren’t even playing the same game, let alone in the same league, as these major firms.
Individual commercial software developers cannot afford to purchase a license to the source of the libraries upon which they are dependent. The “spielraum” they are afforded is determined by the companies which they are beholden to. Major companies toss out little crumbs of code and say, “let them eat cake”. Individual commercial software developers lend the major corporations the legitimacy which the major corporations themselves could not afford. And this social legitimacy expresses itself in public law allowing for propietary software firms to artificially monopolize the software infrastructure of our publicly owned institutions, underwritten by our tax dollars. And of course this is all done in the name of “free markets” and “competition”.
PD is accurately used to describe things in the past tense, ie. what things have become. But there is no corrolate to the production of PD, to developing PD software. Perhaps we could lament the fate of PD, perhaps we could reflect on the reflection which the demise of PD means to our culture. Perhaps then we could revisit ancient concepts like the “common weal” from whence “commonwealth” is derived, but not now, and not here (U.S.).
Easy case, brother: before, they could create closed source applications that used MySQL without having to buy a new license.
Now, they either have to give away their source, or stay closed but pay for a license – kinda like Micro… erm.
But if they’re selling it to profit, why shouldn’t they pay?
Several things: first off, Perl is dual-licensed, and one of the choices is the GPL. So, Perl is GPL-compatible, and Eugenia, you should fix your article. People arguing about who’s at fault for license incompatibility should point to the PHP people; there’s a huge amount of GPL software in the world and license incompatibility with the GPL is a big drawback for any open source software. It is PHP that made this mistake; there’s no reason that MySQL AB should torpedo their business model to allow PHP to continue with this mistake.
Second, someone said that this shows that the GPL is a “commercial failure”. Quite the contrary, it’s showing that combining the GPL with commercial licensing is quite successful. TrollTech is doing it, very successfully, for QT; Aladdin has been doing it with Ghostscript for years, and now MySQL AB is doing it. For it to work, people who don’t like the GPL have to have an incentive to pay.
The simple fix is for the PHP folks to accept dual licensing, like TrollTech did with QT. Then PHP can be interlinked with any GPL software. Resistance is useless; open source licenses that try to be more restrictive than the GPL will need to be adjusted.
“will you please give away your hard work, so i can _sell_ mine”
maggots. the lot of you with that alignment.
I assume by small commercial vendors you who mentioned them mean the “shareware” people. Well I’m beginning to think that the best thing for them is to put their stuff under the GPL and ask donations for it. DJ Delorie (The developer of the DOS/Windows console version of the GCC developemnt environment , DJGPP and several linux versions of the default Windows games under the name Ace of Penguins.) seems to have made a pretty good little business for himself out of doing this. furhthermore this WOULD allow them to use MySQL and QT for free in money making ventures under Linux, at least the way I read the GPL.
@ Never Mind:
> You are not forced to use MySQL. If the shoe doesn’t
> fit, get another, cheapskate. Full freedom of choice.
Fine. So my application, let’s say it’s BSD, which has worked beautifully with MySQL for ages, has now to be ripped apart, plugged on top of a different database and going through all kinds of regressions, forcing long-time users to screw up their server setups, because I have the “freedom” to chose another database?
Nice concept of what software should be like. Oh, “Never Mind”…
@ karl:
Stunning how you could put together so much legalese on PD.
Yes, I know it is not a “license”, strictly speaking. Practically, it does not matter though.
> If you try to relese something as PD here in Germany
> you are releasing it into a nether-world of anonymous
> property.
If I set up a SourceForge project as “Public Domain”, with every source file carrying a notice “This is released into the Public Domain. Use, modify, and redistribute at will”, things are pretty clear-cut:
Either I wrote that code myself, or I copied it from other PD sources. I *could* have stolen it, which would mean I cannot “license” it as PD and any such license is void. Same goes for any GPL or BSD code, though – both the copying from other sources and the voidness of license if I stole the code.
No matter whether I am “allowed” to release that code as “Public Domain”, I am *very* unlikely to sue you for using, modifying, and/or redistributing the code, now am I? About as likely as some GPL author releasing v2 of his code as commercial software and sueing people for still using v1 under GPL.
It doesn’t really matter whether e.g. German law “knows” about PD as a “license” or not – by stating “This is Public Domain”, I am making my intend clear, and you can hold me to it. (“Billigkeit” is the German word for that.)
As for proving the code is mine, and not stolen – again, that’s just the same with GPL or BSD code.
I agree with your notions about the “IP landgrab”. The thing is, in my eyes the FSF and their GPL is just another party in that landgrab: IP is taken, and you must use it only if you play by their rules.
The thing is, the FSF has been *very* successful in brainwashing people into thinking their rules are “better” than the rules of the corp’s. Funny that those who are most vocal about it are usually very surprised to find what the fine print in the GPL actually *means*. (Not looking at you, here.)
You also have the choice of staying with a version of MySQL that has a acceptable licensing. I never said you’d have to like your options, but hey, that’s life, and you’re totaly free to make your choices.
My choice is to strongly dislike the FSF, the GPL, and the Linux / *BSD approach of eternal beta.
> The thing is, the FSF has been *very* successful
> in brainwashing people into thinking
> their rules are “better” than the rules of the corp’s.
> Funny that those who are most vocal about it
> are usually very surprised to find what the fine print
> in the GPL actually *means*. (Not looking at you, here.)
Are you kidding? The GPL is way less restrictive than usual commercial EULA.
GPL allows unlimited use of the executable binary at no cost / usual commercial EULA allows limited use at a cost.
GPL allows restricted use of source code / usual commercial EULA doesn’t allow use of source code at all.
Anyone care to suggest what might happen if I used some of Microsoft’s source code in one of my programmes?
Y’know, you only HAVE to give source code to the people who buy your products.
The problem is, of course, that they can then proceed to give it to whoever they want.
It’s kind of hard to make money selling software when your first customer can give your product away to the world…
Don’t ask me, brother! If your license says somebody can do that, and they do that, you really shouldn’t be suprised. By having such a license you gave them the option to try and make a profit, or to not try and make a profit; its a moral issue. There are peeps, you know, code-jockies out there that are actually suprised that if you give someone the chance to take your work and make a profit off it, no strings attached, that some will do it instead of paying you for support, or donate, or give you a handjob behind the quickie mart.
When is free software not really free software? When software is free only to be politically correct!
Anyone remember the e-mail on a Debian mailing list where someone mentions that he was contacted by a “concerned” member of the WineX team who would be “disappointed” at having to change the license of WineX so people would not be able to apt-get the latest and greatest WineX? Yeah, ok. At least he was civil about it.
“You have the freedom to do everything but…“
Despite the above post, I really don’t care much, if at all, about the issue. I’ve never used MySQL before, I don’t have to deal from migrating because of the new licensing issues, and I don’t even like databases in general!
Still, I make it sound like I stuck it to the man, though, huh? Yeah, stuck it to him right round.
MySQL AB wrote the code so they get to choose the license. Don’t like it? Use something else.
Personally I would have preffered to see it stay LGPL, but it is their code and they have every right to use whatever license they choose. And everyone is perfectly free to fork the latest LGPL licensed release and continue from there.
Now, they either have to give away their source, or
[i]stay closed but pay for a license – kinda like Micro…
[i]erm.
<BR>
<BR>
[i]But if they’re selling it to profit, why shouldn’t they
[i]pay?
<BR>
<BR>
Perhaps they’re not selling it at a profit, but distributing internally, and don’t want to write something for the benefit of their competitors.