well, thank you for your replies. I got disappointed when I saw every one telling me its too bad to use.
and Again, I should say that I have completed main ZZ-D's idea. which you should build 1 2x2x1 block on each side and then permute corners while creating left side's 3x2x1 block. the algorithms I have inserted are the most optimal one, all of them are 5-8 moves.
but I have changed that idea to permute all the corners during insertion of last f2l pair. It will help you to reduce cases to 6 cases, then its better to recognize, and easier to learn. but there is also 6 parity cases to help you go throw your cube If you have done some thing wrong. well, complete ZZ-Orbit has only 12 algorithms which 6 of them are only mirrored of other 6 ones.
Well, after every thing, I know that I have made some thing interesting as my post's topic, and it was all because that Zborowski has appreciated my method. Cause I am in contact with him using EMail.
And isn't solving a method that no one has done it before, a worthy thing?? No one had a documentation on how to do ZZ-D on fly. but Today after learning full cases of ZZ-D (obviously my own algorithms) I made an average of 20 with my method and I got 21 seconds. its notable that my normal average of 20 was 32 seconds 2 weeks before using a method between Fredrich and CLL+ELL. ( here i say this for those who was talking to me about COLL, My main cube was 2x2x2 for 2 years and I have mastered all of those cases both in 2x2x2 and 3x3x3. well Be sure I have thought about that too. )
I dont say my method is GOOD and I dont say my method worth learning. I only say that my method is completed form of ZZ-D which was an unsolved method. those who said "and the closest thing that exists so far is the kind of ZZD in stachu's site... which is 6-7 moves..." how this method is solved and Zborowski himself doesn't know that his method is solved?? then dont fake. thank some thing like ZZ?? I didn't say there were nothing. just try this algorith: <L U' R' U L' U' R> which is the first algorithm that I have learned in world of Rubik's cube and permute your corners with this. then my first solves was with all edges flipped correctly and all corners permuted correctly, I can give you sample solves, but does it means that my method was ZZ-D??? there is lots of algorithms to permute your corners during each step you want, but This is all I know about ZZ-D "It was done before, but no one could solve it virtually on fly" then it was not fast. but actually the method of recognition that I have released is a method that I use in 2x2x2. and I have added looking ahead tips. then you can do it on a fly same as me. cause now I can do it my self.
------------------------------------------------------------------------------------------
this text was written by Zborowski and I have just copied it from my mail box:
When it comes to zz-d there is a small problem - there is a missing link which
nobody has solved yet 
The idea of zz-d is as follows:
1. EOLine
2. 1x2x2 on the left
3. finish 1x2x3 on the left and at the same time remaining six corners permutation (only two additional moves is needed on average -
nobody knows how to do that in flight; this is the missing link...); after this step you have two-move-generation situation
4. 1x2x2 on the right
5. finish 1x2x3 on the right (at this point you have got f2l with oriented edges (cross) and permuted corners
6. finish last layer (permute edges and orient corners - these situations are really nice to remember and make)
short characteristic:
1. small number of moves on average -
47 HTM
2. small number of algorithms - 88 for the last step; steps 1-5 intuitively
3. ergonomics is perfect - since step 2 we use RUL moves, since step 4 we use RU moves
When it comes to permute all corners during insertion of last f2l pair, I call this variant zz-e, but I regard this as introduction to zz-d.
I have never prepared algorithms for zz-e 
------------------------------------------------------------------------------------------
well, every thing is visible for me, I dont know this about you. I can't understand how ZZ-D has been done before me while Zborowski (father of this method) doesn't know about this. (
then please dont be faker). and move counts, I have done my example in 56 moves (52 with cancellation) which is only 9 moves more than what Zborowski expect for
AVERAGE move counts. when that comes to average it will goes up and comes down. then doing only 5 moves more than what Zborowski expect is great while you are tired of not sleeping for 2 days.
and you are true, Im new to this forum. But I am addicted to forums. I take place in lots of forums on internet. and there is always one rule: "be polite, and think before talking" isn't that rule available in this forum?? I have came here and I have name my post "At last ZZ-D method has been completed". is it fake?? i think nope. it was not complete and I have completed it myself. then this title was 100% real. wasn't?? tell me and I will remove my post and I will remove my document and I will stop speedcubing. just prove your Ideas.
And for those who made me disappointed because of new ideas: "Wait until next year, I have started my hard practices to perform a sub-10 record with my method. after that I think there will be a shame on all of you."