2 Replies Latest reply on Dec 13, 2007 4:59 PM by GeorgeWS

    arraycollection limitation

      I am creating a data-centric application where you profile your clients and create notes of each encounter with a customer. My application makes use of a php web service that gets all the information in the database and stores it in an arraycollection. what if the table that it is returning 25 columns of 1,000,000 records. how would this affect the overall application? is this even the correct practice. I load all the information in the client machine to make full use of the sorting and filtering capabilities of flex.

      any insight would be of great help.
        • 1. Re: arraycollection limitation
          m_hartnett Level 3
          I believe that would be a mistake.
          Transfering that much data would take time.
          Changes to the data by one client would not be reflected in all
          clients unless the data is pushed to them again after the save.

          You would be much better off using your DB for filtering/sorting the data to some degree and use flex to filter/sort the result set if needed.

          Just use Flex to query your db as needed bases on what the user is doing.

          just my 2 cents.
          • 2. Re: arraycollection limitation
            GeorgeWS Level 1
            I am doing the same thing 25 columns and I summarize the data One for the start and then when a user makes selections the data filters it works great some tmes it loads kinda slow but very usable. You need all the data in the the array to to fancy filtering with fading animated graphs and stuff. But I suggest Summarizing as much as you can. I recently build a tracker that tracks ever account in California (33,000) with 16 categories, more than 50 items for 26 months (its a lot of data!!!) it takes a few seconds on a broadband to get the data but when its all in barchars interpolate and pie charts move and stuff. I could not be happier!

            Just telling you what in doing im sure you could run more stored prcedures and stuff against the data to get it back even faster...