• 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!

Hash collisions

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

    Hash collisions

    Pissed off due to discovering that 64-bit hash function that I expected to have nil collisions in my lifetime actually seems to have a FECKLOAD in sample of just 40 mln URLs

    Its MD5 time, which sucks as its slowers to calculate and takes double the space

    #2
    Exactly what I was just thinking
    I've seen much of the rest of the world. It is brutal and cruel and dark, Rome is the light.

    Comment


      #3
      Looks like its some other bug than hash's collision, phew, what a fecking relief!

      Comment


        #4
        Originally posted by AtW
        Looks like its some other bug than hash's collision, phew, what a fecking relief!
        Aye, ye cannae alter the laws of mathematics, captain!

        Comment


          #5
          Originally posted by wendigo100
          Aye, ye cannae alter the laws of mathematics, captain!
          Dude, fast low-bit hash collisions are a reality, for example 32-bit CRC is really bad in this respect (for a reasonable number of different entries of course), MD5 - 128 bit is good in this respect but its way slower and takes whole 128 bit, so I settled on 64-bit hash function, naturally I did not do any fundamental theoretical research into how good the function is -- just tested on a bunch of data and it appeared OK, this incident really got me scared but luckily traced error to piece of code always returning same URL for some of URLs, which explained same hashes. Phew that was close

          Comment


            #6
            Originally posted by AtW
            Pissed off due to discovering that 64-bit hash function that I expected to have nil collisions in my lifetime actually seems to have a FECKLOAD in sample of just 40 mln URLs

            Its MD5 time, which sucks as its slowers to calculate and takes double the space

            AtW trying hard to sound clever ...
            Zzzzzzzzzzzzzzzzzzz

            Comment


              #7
              Originally posted by SandyDown
              AtW trying hard to sound clever ...
              Zzzzzzzzzzzzzzzzzzz
              Have you washed all the dishes and prepared dinner for your partner, love? Better stick to things you know best!

              Comment


                #8
                Originally posted by AtW
                Have you washed all the dishes and prepared dinner for your partner, love? Better stick to things you know best!
                Yawn... when out of arguments you resort to this stupid statements ...

                Yawn again .. and Zzzzzzzzzzzzzzzzzz

                Comment


                  #9
                  Oi, take this and SKA updates to technical where they belong.
                  I am not qualified to give the above advice!

                  The original point and click interface by
                  Smith and Wesson.

                  Step back, have a think and adjust my own own attitude from time to time

                  Comment


                    #10
                    Originally posted by SandyDown
                    Yawn... when out of arguments you resort to this stupid statements ...
                    When you have nothing to say on topic you say "Zzzzzzzz", if you ain't got anything to say on the topic then don't say anything.

                    I don't even think you know how hashes and hashtables work in the first place.

                    Comment

                    Working...
                    X