| Re: 8 Reasons Not To Use MySQL |
| Written by ircmaxell | |
| Saturday, 26 May 2007 | |
|
While browsing SlashDot tonight, I came across this article on "8 Sound Reasons Not To Use MySQL." While this article is written with enterprise databases in mind, I find that it has some inherant flaws that need to be discussed. Don't get me wrong, the article does have a significant amount of good information in it, but a lot of what is said needs to be taken with a grain of salt. While I do agree with the author, that MySQL shouldn't really be used in enterprise applications, I feel that the reasons he cites are somewhat weak.
MySQL Uses the GPLThe author states here that a reason not to choose MySQL is because it uses the GNU/GPL software license. If anything, I would feel that this is MORE of a reason TO use MySQL. The fact that its development and testing is open to all users of the product will, in the end, return a more stable platform. The author mentions "For some environments GPL'd software is a non starter." I would like to know why. Unless you are planning on modifying the database system itself (modifying the code), the GPL does not effect the usage of the software at all! Even if a company is distributing a product that includes MySQL, the product only falls under the copy-left clause if it uses MySQL apart from its standard interface. Under 99.9% of usages, use of MySQL would fall under the library clause of the GNU/GPL, and hence be immune to the copy-left protection.
MySQL Doesn't Use the GPLI find that this argument is absurd. The author describes GPL'd code as being a "non-starter" in the section above, yet says that it may not be a fit because it doesn't use the GPL. I also find this argument flawed, because there aren't any other database systems that use a "Free" license, with the exception of PostgreSQL. The classes of databases that he describes in the articles are almost never released under any form of "Free" license. Integration With an Existing EnvironmentThis is actually one argument that I agree 100% with the author on. If you have always used Oracle, and want to continue to use Oracle, then why add another DBMS to the mix? Product MaturityI find this to be one of the biggest cop-outs of all his arguments. He is arguing that product maturity plays a major part in terms of product stability and performance. I actually believe that it is the exact opposite. Once a product passes the one year mark of stable releases, I feel the maturity issues disappear. The reason for this is simple, new technology. Take a look at the Microsoft Windows platform. Despite being developed for over 25 years, the latest version is still unusable (Vista). The maturity aspect is an easy way to dismiss the promise of continued development. There is a reason Google uses MySQL... Feature Set MaturityWhile I tend to agree with this argument more than the Product maturity, but I don't agree for the same reasons as the last section... Availability of CertificationWOW.... To list this as a reason against a DBMS simply shows a major error in management style. What does a certification mean? Absolutely nothing. Some of the best developers in the world don't have a college degree, yet alone certification. Certification allows management to find a mediocre level of competency, rather than spending the time to find an employee that is actually capable and good. Corporate ConsiderationsThis argument is old. It is past its date. Linux is not maintained by any "publicly traded company", yet more than 50% of the Fortune 500 companies use it. "You can't go wrong by choosing IBM," but you also can't save $$$. Companies that don't take chances don't get ahead. To say that it shouldn't be enterprise because it isn't supported by a public company is another cheap shot against MySQL. There's a saying... What would you rather have, a piece of cr@p in a box with a warranty, or a product that actually works? Perception of Scalability This is another issue I find with this document. Other enterprise DBMS are complete operating systems, and run natively on a system. This is why vertical scalability is an issue with the DBMS. What the author fails to realize, is that MySQL runs inside a host operating system. So if you can scale the host vertically, MySQL will automatically scale (without anything on MySQL's part). With Linux virtualization reaching a peak of development, and with Linux having unparalleled hardware compatibility, I believe that MySQL would actually scale better than other DBMS. This is because you can add ANY machine that will run Linux to the database, not just one that's supported by the DBMS. Conclusions I'm not disagreeing as to whether or not MySQL is a good enterprise DBMS. What I am saying, is that the reasons given in this article are flawed at best. I don't claim to be an expert in DBMS or programming for that matter. What I do claim to have is common sense. I feel that when you take common sense into account when reading this article, its flaws are bold and apparent. This article proves that the mindset that corporate America has been using is in itself flawed. Why this mentality continues to filter through the corporate world, I will never understand...
Trackback(0)
Comments (1)
  written by esperado, June 13, 2007
120% ok with everything you says. On my point of view, GNU/GPL's based softwares are often 10 time better than commercial softwares. Several reason for that: Devs does what they like, the way they like and does not work with a gun on their head with a yersterday time limit and a stupid commercial boss to order stupid mickeys. The community is hudge and no commercial compagnies can afford such a "test team" on such various environnements. The source is awailable, improvements and debugs are fast to come from a friendly and generous family of devs from all over the planet. Last, love gives more results than money i can't buy my child with ;-)
Write comment
|
|
| Last Updated ( Saturday, 26 May 2007 ) |
TrackBack URI for this entry
| < Prev | Next > |
|---|
