Yes, im searching for the warp point in the file!. Since the file will only be added max 10 records/seg it will be eassy to trak it's position in real time and read the last values beeing feed in to the CSV file. I already did the binary search thing (it's kind of a binary search): You get the value of the first line, and the middle line of the file, then you go in to the half where the value for the index key is higer and repeat till only one line is left, since past the warp point all the values of the key row are lower than the values before the warp point, this converges to the warp point.
Im attaching the latest verision of the files. If some one wants to try it out and improve it / find errors I would apreciate it very much!
