• Visitors can check out the Forum FAQ by clicking this link. You have to register before you can post: click the REGISTER link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. View our Forum Privacy Policy.
  • Want to receive the latest contracting news and advice straight to your inbox? Sign up to the ContractorUK newsletter here. Every sign up will also be entered into a draw to WIN £100 Amazon vouchers!

Ruby on Rails

Collapse
X
  •  
  • Filter
  • Time
  • Show
Clear All
new posts

    #21
    I've spent a long time with Ruby on Rails. It's pretty damn good but can take a while to get to grips with. Ruby itself is a lot more complex than C++ which can you a headache when you're trying to unravel some of the 'magic' in Rails and 3rd party gems.

    Overall I would say it's perfect for database driven intranet apps (I believe the Beep use it a lot internally).

    For large sites it's also fine providing the developers are well clued up on the various caching schemes.

    Comment


      #22
      Originally posted by jkoder View Post
      Good note about the -server flag. You can also change this in the JVM config file. I'd bet you're right that there is a lot of people not using this.
      In fact the first question I ask when people tell me they have a java performance problem is "have you enabled -server"? I have seen some systems run like a real dog until it was set properly.
      This default font is sooooooooooooo boring and so are short usernames

      Comment


        #23
        Because I am a boring geek I decided to do some performance testing of my Plan B server with and without -server.

        Running a tomcat server (6.0.18) under JDK 1.7 a batch job kicks off when the server starts, the time to complete job is...

        -client (default) - 12 minutes
        -server - 5 minutes 20 seconds

        so it doubles performance on my setup.

        I hope that information enriches your life as much as it did mine

        Comment


          #24
          Originally posted by minestrone View Post
          Because I am a boring geek I decided to do some performance testing of my Plan B server with and without -server.

          Running a tomcat server (6.0.18) under JDK 1.7 a batch job kicks off when the server starts, the time to complete job is...

          -client (default) - 12 minutes
          -server - 5 minutes 20 seconds

          so it doubles performance on my setup.

          I hope that information enriches your life as much as it did mine
          So does this mean now that you earn less as you don't have to wait as long for the job to finish so you can get more work done in a day?
          This default font is sooooooooooooo boring and so are short usernames

          Comment


            #25
            Originally posted by MPwannadecentincome View Post
            So does this mean now that you earn less as you don't have to wait as long for the job to finish so you can get more work done in a day?
            Or does it mean you earn the same but get more posts in on CUK

            Comment


              #26
              Originally posted by minestrone View Post
              A common mistake is to not charge for 4 weeks performance tuning work while only adding -server to the JVM.
              An old boss once offered the sage advice that no app should be delivered with maximum possible performance. His logic was that no matter how fast it was, there would always be pressure to speed it up, and if you had already done everything you knew, you'd have nowhere to go - a bit like having volume that goes up to 11.

              Comment


                #27
                I think I am getting to the stage in my career where I care so little about the average Client Co's system I would use a speed up loop.

                Comment


                  #28
                  Originally posted by minestrone View Post
                  I think I am getting to the stage in my career where I care so little about the average Client Co's system I would use a speed up loop.
                  Nice article. In Java though that would more than likely get inlined.

                  Comment

                  Working...
                  X