Can any one help me how the sed delete only first line command (D) works.

sed -f sedcmds File.txt

/^$/{
       N
       /^\n$/D
       }

This is line1

This is line2


this is line3




This is line4.

The sed man page is pretty clear about the function of the D command which says that the D command will cause the next cycle to start, but will prevent the reading of the next input line if the pattern space is not empty.

Thus, the way your programme works can be explained with a bit of pseudo code:

while not at end of file 
do
   if pattern space is an empty line  (/^$/)
   then
       read next line and append to pattern space (N)
       if pattern space is two empty lines   (/^\n$)
       then
           delete to the first newline  (D)
           goto loop (leaves a single empty line in the pattern space)
       end
   end

   print pattern space
   read next line

:loop
done

Try:

sed ':a;N;s/\n$//;ta' infile

sed -e :a -e 'N;s/\n$//;ta' infile