Slashdot Mirror
Is MySQL's Community Eating the Company?
mjasay writes "Craigslist's Jeremy Zawodny reviews the progress of MySQL as a project, and discovers that through third-party forks and enhancements like Drizzle and OurDelta 'you can get a "better" MySQL than the one Sun/MySQL gives you today. For free.' Is this a good thing? On one hand it demonstrates the strong community around MySQL, but on the other, it could make it harder for Sun to fund core development on MySQL by diverting potential revenue from the core database project. Is this the fate of successful open-source companies? To become so successful as a community that they can't eke out a return as a company? If so, could anyone blame MySQL/Sun for creating its own proprietary fork in order to afford further core development?"
Monty has been blogging some about the need to be a more inclusive project. Its one thing to be open source, but to be an open source community project thats still owned by a company takes real effort on the part of the company. Perhaps this would encourage some of these enhancements to be rolled into the main branch.
For paying that much money for a company that gives its core product away for free!? MySQL made a bit of money through support contracts, but now they have a lot more zeros to account for when they pay the bills.
This isn't MySQL's fault. If someone wanted to pay me 3million for my piece of crap car, I would sell it for half that, so they thought they were getting a bargain, but how could Sun justify paying that much?
What are we going to do tonight Brain?
Maybe Sun should rename their fork of MySQL to SunSQL Solaris Edition JDK
I'm ready to use PostgreSQL now
This is always the case when you release open source. Someone else can offer support cheaper than you. Someone else can make modifications that people want. Someone else can even fork it and choke you out if they're doing whatever gets more attention than what you are doing. The good news (for them) is that you provided them a getting-started point with all your work so they didn't have to put all that time (and money, since time is money) into getting it off the ground. This is the way GPL/OSS is *supposed* to work. You have to keep investing more time and money while pushing and driving your costs to zero or you'll get snaked by just about anyone else who has the motivation to do so.
If it's all OSS, then why isn't MySQL picking up the best 3rd party pieces and rolling them back into the official distribution?
[Fuck Beta]
o0t!
Sun should welcome such improvements into their dev cycle. If such forks are superior, then they should eventually find their way back into the parent model. The successful business models around OSS rely on the services/consulting/support that sit around and on top of the actual OSS code. Red Hat, IBM, HP, and others all understand this. Sun, unfortunately, still has the MySQL model wrong IMHO.
davejenkins.com |
"If so, could anyone blame MySQL/Sun for creating its own proprietary fork in order to afford further core development?"
Wait - what good would it do for MySQL/Sun to create it's own fork if, by the poster's own declaration, community supported forks are *already* better?
I think, maybe, part of the problem is companies (not just Sun/MySQL, but other companies I've seen this with too) not really treating open source projects *as* open source. They release the software under GPL, or whatever free license, but because they want to maintain 'copyright purity' (that is, the code they distribute is 100% owned by them, because that is the only thing that will allow them to potentially make the codebase proprietary for selling 'enhanced' versions; if they accepted other contributors' code under the GPL, they would then have to accept the code to be GPL forever, for all versions), so they won't/can't integrate other contributors' code into the main distribution (unless they can work out some seperate licensing agreement with the third-party developer).
Whenever you have a situation like that, as a company, you are giving other developers the benefit of Free Software while *denying* it from your own customers (well, sorta, until they stop being your customers and start using the other forks), and yourselves.
I don't know what the 'best' business model is for open source companies, but if you really want to leverage open source/free software, you have to give up on directly charging for 'enhanced' versions of the software, because the only way to play that game is to force this situation where you cannot benefit from the enhancements of the community. If you are successful, like MySQL, then eventually the community grows to the point where the community's developer resources are greater than your own as the company, and you find yourself in a situation where you can't really keep up with the community.
MySQL seems to be a project with alot of mindshare that doesn't execute well.
With commercial software, you're screwed when the vendor decides to do stupid things. With OSS, you have options besides moving to a new platform or living with the vendor's stupid decisions.
At the end of the day, this is good for everyone, and is an example of why OSS is good for society.
Conformity is the jailer of freedom and enemy of growth. -JFK
One of the keys to a successful open-source project is to take the improvements being made in forks and bring them back into the main project. One of the reasons forks are created is that users have a need that's not being met by the project. If you bring their solutions back into the mainline project, the fork will tend to die because it's no longer needed.
This is, BTW, one of the reasons to use a GPL-like license. If you do, you're guaranteed that you can bring improvements from forks back into your mainline codebase. If you go with a license that allows you to create a fork with things that aren't available to others, it simultaneously allows others to create forks that aren't available to you. Then you end up in Sun's situation with no way to resolve it except by creating the same improvements yourself. And there's more of your competitors than there are of you, which means they will win this particular race to create improvements. If you go with a license that forces improvements to be available to you but not anybody else, many people who might have created an improvement you could use will simply not contribute to your project. It's a perception issue: GPL-like terms lead contributors to think in terms of their contributions helping everybody and you just happen to be one of that "everybody", while "owner gets everything, everybody else gets what the owner gives them" terms tend to lead contributors to think you want them to work solely for your benefit without you giving them anything in return. That turns a lot of people off.
Lots of wrong things in this article.
OurDelta isn't a fork of MySQL. It's builds for the regular MySQL with optionally some third party patches.
Drizzle isn't a fork of MySQL. It's a complete restart and reengeneering of the database core of MySQL and will likely become a base for the future releases of Sun's MySQL and other database products. Drizzle is to MySQL like MinWin is to Windows, though maybe bad analogy, MinWin is just as porly understood by most people.
Sun doesn't have a propriatary fork of MySQL. Former MySQL AB wanted to put some proprietary services and applications on top of the existing open-source product, but the community reacted and since Sun never approved of this direction, those plans were immediately dropped.
Sun/MySQL can and should be blamed if they are failing the community that made MySQL so popular and strong.
Sun has a bad reputation for having very closed open source projects such as OpenOffice. The project is managed much more like a proprietary venture than an open source project and community input is minimized or ignored altogether.
I can't feel sorry for Sun when they drop buku bucks on MySQL and then complain that others are taking their revenue away from them doing what the OSS community does best - improve the software on their own.
...and that's the way the cookie crumbles.
If Sun actually cared for the "long term unix community" you speak of, you wouldn't have that "unix like" OS in hand.
Thank God both them and various companies are wise to ignore such long term communities so they keep doing favours.
I know a long term Unix Hack. He says. "Fuck Sun. Fuck IBM. Fuck SCO. Fuck Microsoft..... come to think of it Sun aint too bad"
People need incentives to buy products, and open source software doesn't give people that incentive outside of the enterprise realm where paid support is a big money maker. Let's get closer to regular users here. What incentive do home users have to buy StarOffice when OpenOffice is more than good enough? How about pay anything for WordPress when it's free and easily installed by CPanel? I can't see any, aside from altruism, and only economically-clueless nerds tend to put much stock in altruism as a source of revenue (this also explains why so many think that bands will still sell large amounts of recorded music, even though most of it can be downloaded for free on a P2P network).
In the case of MySQL, a big part of their problem is that Sun isn't dumping a lot of R&D money into them to make MySQL 6 really competitive on an enterprise level with PostgreSQL. A pure open source approach isn't going to allow Sun to make good money on their R&D investment, but if they were to dramatically improve MySQL and provide high quality tools at reasonable prices, that sort of hybrid approach would work. Companies that want to make their core software open source are going to have to make compelling products that interact with them if they want to be able to sell more than consulting and support services.
Open source advocates need to be realistic. If you do work outside of the enterprise realm, chances are you will end up doing it for free and never seeing a dime from it unless your users are feeling overly generous. That's just because most users outside of the enterprise realm have no incentive to buy anything you might be selling related to your open source product.
Sun is the worlds largest open source company both in terms of size and contribution.
MySQL
OpenOffice
Java
VirtualBOx
Open Solaris
are all wholly Sun projects but they also contribute to numerous other open source projects.
Sun may not be perfect but, there are none better at the moment.
An important question is not whether the Open Source community is eating some of SUN's cake, but whether the cake itself (and thus SUN's total amount of cake) is larger because of the community. I don't have any figures but this is at least a considerable possibility. After all you have something technically superior like PostgreSQL *ahem*
Half of a big cake or all of a small one. SUN bet on the former, I think.
Aide-toi, le Ciel t'aidera - Jeanne D'Arc.
There's no need to put the blame on the corporate body. Developers have mortgages too.
"So, Developer Danny, I notice that 8 out of your last 10 commits have had someone else's name on them. Can you explain to us what value you bring to SUN, and why we shouldn't just hire or reward the 3rd party contributors directly?"
If you were blocking sigs, you wouldn't have to read this.
I've been in contact with a dozen companies that all use the community version of MySQL. Without paying any support because none is needed when you have a semi-competent DBA around. If MySQL wasn't there, then it would have been PostgreSQL instead. If there was no MySQL, PostgreSQL nor any other high quality free RDBMS, they would have to use a commercial system instead. There are thousands of companies out there in the same situation and I don't think that MySQL has gained as much money as the commercial vendors has lost thanks to MySQL:s freeness.
Football Odds
s/KDE/Qt/
KDE is purely Free Software and there is no proprietary version.
KDE is build on top of Qt, which is dual-licensed and available either via the GPL or via a commercial license. The code is identical, but the commercial license allows you to build non-GPL products with the tools, and provides some support.
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
thats what ibm and others are doing, and its working well. sell your customers COMPLETE solutions. do not put stuff out separately.
think it like this - imagine you are going to offer a webserver solution. hey, the database server development is already handled by the open source community, cutting many of your development, bugfix, testing costs.
AND you will sell support. no, really. no business can go to an open source forum, post their problem and wait for a useful answer in a busy workday. they will want to have someone to call and get support fast. AND that will be the company who sold the solution to them. charge reasonably for support.
do NOT try to go into the ancient 'hey we did something, we are gonna sell it and make money'. in our days and times, support, service are constant revenue streams. whilst you buy a server every few years. which you would want to bank on ?
Read radical news here
Much of the article & threads here seem to be supposition, and niche arguments. MySQL has the mindshare because, back when RedHat was all the rage on production servers, MySQL + Apache was just an RPM away, and LAMP started to really kick in (Linux, Apache, MySQL and PHP). PHP has big mindshare too, and the MySQL functions *are* the DB functions for a lot of coders out there.
So even if you fork, add third party patches, or whatever... the fact is that the basic MySQL dominates the low to mid range server DB market in Open Source, and that's that. Of course there are better alternatives available, but hiring staff that know those alternatives isn't as easy.
So I reckon Sun won't be affected too much, their product does what most people need already. Those who need something else can pay Oracle, MS or work with PostgreSQL, which kinda got to the party late. Yes, it is more powerful. But it's LAMP and not LAPP, and the tutorials for PHP/MySQL outnumber PHP/Postgres by a large factor.
Conversion Rate Optimisation French / English consultant
Hi!
We typically call Drizzle a fork, since we do have a common ancestor at this point (though it is doubtful you could apply a patch between the two). We are pretty up front about this though. Drizzle is supported by Sun which the article does not mention, though we are different in that we have patches that have to date come in from 30+ companies.
OurSQL is more of a distribution then anything else. Their tree is a collection of patches they apply at each release.
Cheers,
-Brian
You can't grep a dead tree.
THE CAKE IS A LIE!
Exactly. I have very few problems with Sun, and I don't HATE Microsoft like so many people do. Making money is not evil. I wouldn't want to see the MySQL project thrash Sun. Although I doubt it will: Most companies I've heard speak about their software support DEMAND professional support they can call whenever they have a problem. Sun provides that.
I don't HATE Microsoft like so many people do. Making money is not evil.
Note: I strongly dislike Microsoft, but not because they're profitable.
Dewey, what part of this looks like authorities should be involved?
Seriously. What is the business case where one may justify using MySQL and paying for it?
Whenever the issue of MySQL vs other RDBMS is raised on Slashdot, its niche is typically defined as cheap hosting and other low-budget solutions where reliability and data consistency are simply not as important. Nothing wrong with that, but isn't it precisely the segment that's not going to pay so long as they can get the same for free? Support - I doubt many people actually care much about that, especially for MySQL. And any Web solution, no matter how big, doesn't have to care about the license of the DB backend.
This leaves commercial apps who don't like GPL - but they can avoid it much easier by using one of the many viable alternatives, from SQlite or Firebird on the embedded side, to Postgres and again Firebird on the full-fledged database side; and for $$$, if one is willing to spare some, why not just pay a little bit more to get Oracle or even MSSQL?
So it seems to me that the target audience to which MySQL caters is precisely the one that's less likely to pay for any "enterprisey" features, and stick to the base F/OSS version, with "good enough" community patches. Or am I missing something?
First thing,
I am a hardware guy by inclination and training. I had some programming experience, but never enjoyed it.
I also tend to look more at the business side of things sometimes. I am the Chairman of my Company's ESOP committee. We are 100% employee owned. We are not an IT company. I am the only IT person.
I have always had a hard time believing in the business model of the Open Source community. As an IT person and a software customer, I can appreciate the ability to view and modify the source of software that isn't doing everything I want it to do, but as a business person, I have a hard time picturing a long term model where open source is a product I can make money off of.
In the short term, customers would purchase the software and support from me, but as the community grew around the software, it would fork in new unsupported directions, the community and customers would become more savy and need less support from me.
Long term, I think open source will work real well for drivers, routers, switches items that the hardware is secondary to the appliance.
But for standalone software products like Databases? I just can't see it, no matter how many Open Source advocates try to convince me of it.
I don't hate MS either, but I don't think the reason people hate MS is because they make money. It's because they've often been assholes about it. People generally love a good success story as long as the recipient of the success seems to have done it in a fair way. Witness Warren Buffet. Or Google... as long as they shy away from evil. Apple is loved for now, but they're evil enough to get themselves thoroughly hated down the road.
But the point is: making money is not evil, but if you make money while being evil, people will hate you. And that's as it should be, really. Nobody likes to see assholes get ahead.
Cheers.
Making money is great. Making money through rampant corporate skulduggery isn't.
Y'know, I'm with you in general, but pointing at the Xbox division as the epitome of capitalism is, I think, misguided. They've lost billions upon billions of dollars and have no chance of being profitable (as a whole) this generation, and certainly weren't profitable last generation. Xbox is an example of a company trying to leverage a monopoly to fund expansion into a new market at a huge loss, with only vague plans on *ever* making a profit. That isn't exactly the ideal entrepreneurial spirit.
Google is an advertising company. People don't hate them simply because they're a *good* advertising company, carefully controlling their corporate image. Google makes money by putting ads in front of you - hardly somehting to love them for.
Socialism: a lie told by totalitarians and believed by fools.
The X-Box 360 was put out before their production line was ready to make it. More like capitalism at its worst.
"So, Developer Danny, I notice that 8 out of your last 10 commits have had someone else's name on them. Can you explain to us what value you bring to SUN, and why we shouldn't just hire or reward the 3rd party contributors directly?"
Which of the six developers who's work I committed do you propose to hire? If they're busy vetting other people's code, when will they have time to keep writing this great stuff?
This is known as the "developer as interior decorator" model, where the developer is hired on their ability to mix and match code that is in good taste that blends attractively with the existing architecture.
"Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
Remember Sparcworks, the official, surprisingly expensive compiler for Solaris with really annoying license requirements that your management made you buy, that immediately became shelfware in favor of the free and far superior GCC if you hoped to do anything approaching ANSI C development?
You mean the Sun compiler that is ANSI compliant and produces better (smaller, faster) code than GCC? And what's "annoying" about the license? At my last firm we developed trading software for Solaris, and none of our customers gave a fuck about the Sun compiler license (although one insisted on our code being compiled with it, thanks to bad experiences running GCC compiled code on Solaris thanks to previous issues with a shared libgcc).