> there was no actual timestamping so we cannot understand precisely how much time it > took, or average timing of each one, and it is a data we are absolutely not > interested in, quality is the priority. But: > > Each image consisted of 256 bits > there was a total of 256 images > > So, we have a romsize of 65.536 bits, that is 8kb > > All images were typed three times > So we have a grand total of 196.608 bits typed > > Annnounce was doone at 9.37PM of 18MAR and completed was announced at 1.23PM of > 19MAR. That mans that we took less than 16 hours to do the job
I totally agree. Quality and not speed is what is important. Every bit-error will increase the workload on the decoders. An idea just poppep to mind. In an image two out the three might say a bit is a 1 and the third said it is a 0. The third person may be right and the other two have got it wrong. How about throwing open all "disputed" cases for others to type in? Two out of three is good, four out of six is better when it disagrees with the two out of three.
The ROM-size and time-taken speaks for itself. Just from those can a few amazing stats be exxtracted / extrapolated. On average more than 16 Images (3 times) typed in every hour although only 101110 ( ) people were working on it.
Regards Mark
|