ask one question about e - generation solver mechanism

hi,guys

Always ,i am not very clear about the e language generation solver mechanism,i read the reference, and some relevant pdf files in doc directory,all i acquire is the special rule for generation,you know, that kind of special rules are focused on special constraints.sometimes even yo u change just one word ,the constraints will be syntax invalid ,even worse ,some constraints are in vain , but you really do not know that. The worst case is :specman does not print warnings ,or even give you some clue about it ......... 

the quantity of special constraints are huge,hard to remember if you do not use them all the times.At first ,i think if i read all the special rules around the constraint of generation solver, i can touch the core of that mechanism, but ,still i failed.............................i always believe that the constraint is one of many arts of verification.could you give me some advice,suggestions. Am i wrong from the very beginning?Am i in the wrong direction to learn such one language tool.

my teammate could always say :"do not constrain a lot in struct or uint,put them into functions!",i still can't think so,that is one way to avoid make yourself in trouble ,you choose the different ways,but you do not be confront with the facts which specman could do that perfect if the users totally understand the generation solver mechanism..........

oh,God,help me out! thank you ,and also to the human beings who give advice. 

there is something i think i should talk about,the constriant GUI debugger ....i do labs or debug constraint errors on every step the pdf says,but sometimes  it's still hard to locate the wrong codes,all i see in my eyes is one talbe with many  grids in it,the generation flow is clear,but it is not  i thought what it's supposed to be.so ,a very low debug effeciency i work with.