Well, it's at the same time very detailed question and very broad. It's like why Hyundai created a new car, there is Rolls Roys already here (200K). Or like why CUBRID, NoSQL, etc. when there is Oracle Enterprise for 10K (Oracle was one of the first DBs in this nice world).
It's all about the diversity which gives users more freedom in deciding which tool is best for its particular project or event part of that project. The more alternatives, the cheaper (TCO in our case), the better for consumers. Typical rule of the market. More detailed answers: CUBRID has been developed with Web optimization in mind, where constant availability and speed are the most important factors. What relates to the question is: will I need to use additional third-party tools to have my work done if I use this tool? In case of MySQL, for instance, how do you implement High-Availability? Noway, unless you use MySQL Cluster or other third-party tools like DRBD and together with Linux Heartbeat. This is probably something simple web sites never bother about, but for real world mission critical application this is the number one headache. Besides, the simplest replication cannot be done properly, I mean with 0% loss, on MySQL. As you know MySQL does not provide 100% guarantee for data replication. You can refer to their manual for more info. What about CUBRID? As we have thought this through when we had created our architecture, we considered high-availability from the very beginning. CUBRID HA is a very powerful and distinguishable feature with its native CUBRID Heartbeat. This is something almost all our big users like NHN use in their web services (NHN has 17+ million UV per month). Moreover, in CUBRID 8.4.0 we have reworked the HA feature in the terms that it is now much easier to configure and more convenient to monitor. The data consistency during HA is always 100% guaranteed. I think you should think about it if you really value what you have on your site. I would talk for a long about the CUBRID License Policy (http:// cubrid.org/license), but I think you know yourself what MySQL's policy is. CUBRID engine is entirely GPL, while everything else users develop upon/for CUBRID can be BSD, thus we do not require to open your code. You can't do this with MySQL. Either you buy the Enteprise License for 2K at least, or open your code. Everything derived from/related to MySQL, must be open. You feel free with CUBRID. At CUBRID we develop faster, we react to user requests faster. I suppose this is in the nature of the developers (Koreans work really a lot, ~60 hours a week in general, I was surprised at first, then got used to it). At CUBRID we want to provide ease of development for users. For this we have created video tutorials, everyday post new tutorials, improve the APIs constantly. We are more close to the community. You probably do not feel like at home at MySQL community. But you do at CUBRID's. At least I will guarantee your every response is addressed. We provide compatible APIs, at least do every endeavor to, so that our users do not have to change much of their existing code to comply with CUBRID style. See CUBRID PHP API Comparison list with MySQL (http:// www.cubrid.org/cubrid_mysql_php_function_comparison). Almost all functions are identical thanks to the contributions made by our Chinese users. The same we try to do with Ruby, Python, etc. Here we encourage the users like you to jump in and help us to make our drivers easier for our peers. In terms of performance, you should see our comparison results. Like http://www.cubrid.org/ssd_performance_test or http://www.cubrid.org/benchmark_result. I suppose in the coming week I will try to upload the latest performance test results we have conducted recently. Due to the index restructurization in the new CUBRID 8.4.0 the SQL statements (we tested complex multi joined queries) run 30-40% faster on CUBRID than on MySQL, considering the same environment and settings. This is the second part of our mission: to fast database engine in addition to constant availability. But like I said, the questions like "why should I use Cubrid instead of Mysql or Postgres ?" are quite broad. So, feel free to ask specific ones, and I will be happy to continue our discussion. Btw, did I provide enough reasons? On May 17, 12:13 pm, Bruno Rocha <rochacbr...@gmail.com> wrote: > Esen, > > I would like to know whats the grand key features of Cubrid over Mysql or > Postgres, I am interested in this project but I have a simple question "why > should I use Cubrid instead of Mysql or Postgres ?" > > -- > Bruno Rocha > [ About me:http://zerp.ly/rochacbruno]