hi, am trying to compare ror with any java web-based framework. i was thinking to use j2ee for comparison... any comments on that? also i am not looking for any complex web application since this is a course project...so i would probably test on the sample shopping cart application that we see in ROR papers. so i guess i hav to implement in this j2ee or something like that..or is it already done by someone? how to compare two frameworks? thats the big question...i was googling around..n found that there is a "mercury loadrunner" which gives performance stats. is it a good one? or is there any open source s/w which can do that job. any pointers or suggestions would be helpful thanx, v
on 2007-02-25 05:59
on 2007-02-25 06:52
Hi Doniv, this isn't a comparison site for Ruby on Rails vs X. This is a site for Ruby on Rails development. There are many web frameworks that can be used for web development. Your decision in some cases will largely be dictated by the people that you work for or the goal that you have established for your project. The people within this group have a certain level of passion for Ruby on Rails and using it for web development. If this is a course project, it seems that your goal is to learn by doing your own research. I'm guessing that your goal for this project is compare both performance and productivity. If this is the case, you can start by defining a spec for the web application. You should be able to extract a specification from depot app in AWDwRv2. Then use this spec and implement using both frameworks. The information that you gain here should produce alot of statistics (i.e. LOC, development time, and so on). Next, you can check performance using a light, medium, and heavy load using N number of users doing diferent things on the system. This should get you started and I wish you the best of luck. -Conrad
on 2007-02-25 07:40
doniv wrote: > also i am not looking for any complex web application since this is a > course project...so i would probably test on the sample shopping cart > application that we see in ROR papers. so i guess i hav to implement > in this j2ee or something like that..or is it already done by someone? Try line count. A Ruby project will typically have 10% of the lines of a Java project. Then try development speed. > how to compare two frameworks? thats the big question...i was googling > around..n found that there is a "mercury loadrunner" which gives > performance stats. is it a good one? or is there any open source s/w > which can do that job. That's apples-and-oranges. Mercury is designed for people who write a slow site and then want to tune it to make it faster. Using it to compare two different sites would "laminate" the signals together. That means mixing statistics from diverse sources. One site would be faster on one feature while another is faster on another feature, and the average would not reveal which web platform was "better". Development speed is much more important than execution speed (and Ruby is hardly the zippiest language!), because programmers who finish their features first will have time to profile their app and find the truly slow parts. They are usually easily fixed, once identified. So if one language only requires 10% of another, which do you think is faster to develop features? -- Phlip http://www.greencheese.us/ZeekLand <-- NOT a blog!!!