Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
08/13/07 13:02
Read: times


 
#143161 - ... and a marginal remark ...
Responding to: ???'s previous message
... for those who don't have to accommodate to an existing CRC-clone, but their task is to include one to their design: it is possible to "design" the "parameters" so, that if you feed it with a continuous stream of zeros, the resulting checksum is zero regardless of the length of the stream (in fact it's easy - just set everything to zero, in most of algos this will work).

Although a strict matematician will see no problem in this - zero is just a number as good as any other, so it's good for any checksum too - the engineer sees the gotcha - it's quite easy to have such an error in data storage or transmission, which results in such continuous stream of zeros - and that would then pass undetected.

JW


List of 32 messages in thread
TopicAuthorDate
CRC16 CCITT - What a challenge!            01/01/70 00:00      
   Have you seen this?            01/01/70 00:00      
      Of course!            01/01/70 00:00      
         Balance            01/01/70 00:00      
            I know, Russ,            01/01/70 00:00      
         Code? What code?            01/01/70 00:00      
            in the code library - look to the left column            01/01/70 00:00      
               Thanks! I missed that note in the first post.            01/01/70 00:00      
   I need a hint            01/01/70 00:00      
      write down the bits            01/01/70 00:00      
         Thanks for the hint            01/01/70 00:00      
            Your observations plus a comparison in C            01/01/70 00:00      
               The rest of the story ...            01/01/70 00:00      
                  Compilers, but Keil C51 8.09 for the stated 28~            01/01/70 00:00      
                  and this processes only a nibble, isn't it...            01/01/70 00:00      
                     No, it process the entire byte.            01/01/70 00:00      
                        now it\'s my pencil time...            01/01/70 00:00      
                           FYFI            01/01/70 00:00      
                              Clarification re: 8-bit optimized term            01/01/70 00:00      
                              on optimisation            01/01/70 00:00      
                                 Jan's optimization            01/01/70 00:00      
                                    Re: on optimization            01/01/70 00:00      
                                       Can you please be more specific?            01/01/70 00:00      
                                          More specifically ...            01/01/70 00:00      
            and what            01/01/70 00:00      
               More on checksums            01/01/70 00:00      
   No reason to hide the excreted code            01/01/70 00:00      
      Excreted code needs context like any other code            01/01/70 00:00      
   One More Comment            01/01/70 00:00      
      ... and a marginal remark ...            01/01/70 00:00      
         Init all ones frequently specified            01/01/70 00:00      
   18 cycles            01/01/70 00:00      

Back to Subject List