It is a fantastic problem, the best I have seen using sentinels and probably one of the top problems in Juliafairies this year! The play of Nao and Rose-KA one after the other is complex and striking! A great tribute to the inventor of Sentinels. The master shows that the sentinel has so many secrets yet to be discovered, especially handled by such talented composers.

Thanks for your comments! Yes, of course on d2 stands a black RAO – this is my technical error admitted in writing of the text. I will send to Julia correction of the solution, also with some additions. By the by the problem was tested with the wonderful program of Christian Poisson Win Chloe – with version 3,32!

About to No.900 (PAP)
Regarding my №900 Nichola Predrag publish the following version:
“…….However, your “delusive idea” was probably created by your good intuition, for instance ( on the occasion of Kjell Widlerts comment):
W: Se8,LEf6,Pb5,Qg4,Sd3,Kc2; B: Kd5,RPb3,Sc3,RAd2,Ba1
h#2; Condition Sentinelles (RP is Rose-hopper)
1.LEe5[+wPf6] RAf7[+bPd2] 2.LEg5[+wPe5]+ RPxg5[+bPb3]#
1.LEc6[+wPf6] RAa5[+bPd2] 2.LEb7[+wPc6]+ RPxb7[+bPb3]#
1.LEd6[+wPf6] RAb7[+bPd2] 2.LEd8[+wPd6]+ RPxd8[+bPb3]#
1.LEe6[+wPf6] RAd8[+bPd2] 2.LEf7[+wPe6]+ RPxf7[+bPb3]#
Both bRA and bRP guard flights in the mates”.

I once again thank Nicola`s and other commentators for the interest that they exhibit towards my problem!
But in my opinion, this version is unacceptable and I will point out many reasons for this.
1. Of course, this problem (here is used exactly the LEO-location&antibattery-creation from my version with 4 solutions) is with only 4 solutions but not by five, as is the original! It is a very essential minus of the proposed version because in four solutions the square “f3” remains unvisited – therefore – the paradoxical geometry of the original here is not preserved. (see the great scheme of Shankar!). In N900 in the process of play on each field by the large Rose-circle b3 – d2 – f3 – g5 – f7 – d8 – b7 – a5 there is white or black figure! By the by, in the course of my work I have presented many versions of this task with 4 solutions – including and those by using of Rose-hopper, by using of model-mates etc.
According to already mentioned reason – very significant in quantitative and qualitative terms, I chose the the version with 5 solutions!
2. For me, the main quality of a problem is the complex of ideas – This complex must be sufficiently interesting, complex and even paradoxical. It is not mandatory in every case and at all costs the play to ends with model mats – especially when the attainment of such finals obviously is not possible or leads to big compromises – for example – reducing the number of solutions -as in the version of Predrag, the use of the most powerful white figure – queen, just as technical and stationary element, etc. Incidentally,
I think that there is no matter the fact that in my №900 the square b3 is attacked twice. It is obvious that with 5 thematic solutions moldel mates are not possible here. The thematic black figures play with maximum activity, but will be fetched and unrealistic if we expect of them more “feats”!
Do not assume that I hate the model mats – the opposite is true : Nearly 1,000 of my composed problems (almost 16%) are created with model mats! I think that as a judge of competitions also I appreciate properly the tasks with model matеs
But I’ve never prepared to sacrifice an interesting idea, due to model mate – in particular – to shorten contents, to maim the wealth of ideas, only in the name of the “sacred model mate”! My principle is the following – model mate is an interesting and nice finish. It is good to achieve a model mate, if it is possible without compromises! But the model matе should be not a fanatical purpose!
3. I already mentioned that (on the occasion of Juraj Lorinc`s comment) that I appreciated as an aesthetical weakness the “dual routes” of the piece(s) Rose&Family , for exmple – such as in the Predrag`s version: RAg5 but via e4-g5 or f3-g5; RAf7 – but via c4 – d6-f7 or f3-g5-f7 etc.
A similar weakness (as an improvement) was proposed by our good friend Kenneth about my problem № 895 –Julias Fairies, 2015-09-16:
the proposal was: to remove black Sc2 and to add blacl Pc5 – without change of the solutions! BUT!!!! In such position is obviously that the initiall setting is not quite normal in aesthetical attitude: for example, here the square “f5” is already controlled from RAa3 (via Pe3) but the other thematic fields: “e4” and “f6” arw not controllwd in the beginning! This aesthetic disbalance – the unequal status of the three black themed pieces – disturbs from the outset! The same I have to repeat for mate-finals in solutions a) and b) – here the square f5 is controlled twice – with help of the black Pd6 and from white Pe3! So the the black mating Pawn on “d6” includes in a) and b) solutions only one black thematic piece but in solution c) – two thematic pieces!
In other words – the double control of RAa3 over “f5” violates also the thematic identity in the play of Pd6! This is a good example how a “small compromise” can lead to another, more unpleasant disadvantage!
Therefore in my №895 the black Sc2 is a very usefull piece!