This may fit under theory rather than no topic. 

 My language, Siye,  has a writing system that is in effect a syllabary with some phonetic and (by virtue of its templatic morphology) semantic  components.  My current reference sheet to the system is a grid with the onset consonant ( or absence thereof)  on one axis and the eval on another.  This scheme however, was designed purely for my convenience rather than reflecting the native order of the language.  I have details of the system on my LiveJournal account but for obvious reasons you may not want to go there.  So I'll provide a brief summary here  before asking my question.

 The native word for a character in the syllabary is literally 'family', but that would be a little confusing in the way that Terrestrial linguists use the word family, so I world war for to them as characters. The most basic character consists of the largest component alone, which is called a mother. A mother can have a smaller component modifying it – this, unsurprisingly, is called a daughter. There can only be one mother in a character, but there can be multiple daughters.  Sometimes there is an even smaller component and Companying the daughter.  This component is called a granddaughter.  Some characters only have daughters;  in these cases the daughters in question are referred to as orphans, but any smaller component is still called a granddaughter. A mother can be reduced in size and become a daughter, but in that case any daughters of the former mother disappear. The most common mothers to become daughters of a different mother are the vowels it disappear as characters during vowel contraction - The vowel of the new character is determined by vowel dominance.  But I feel it is more important to sort out the basic characters first, so I'll put off that discussion until another day.  The primary importance of secondary characters in this discussion is that they are often homophonous wWith primary characters but theoir restricted usage prevents confusion.

So my next task was to establish an  indigenous order of the characters. I examined Chinese characters stroke order  and organization by number of strokes.  In Keno Siye ( The name for the writing system),  which runs from left to right  on the level of completed characters,  a vertical downward stroke is usually the first stroke.  This is followed by the right-hand stroke and then the left-hand stroke and finally the daughter, especially if the daughter is just the single line. In other cases, however,  especially those in which  The mother can be seen as a combination of two right angles,  The right angle which resembles a V is written  but for the right angle which resembles a backward L.  In unusually complex characters,  especially those which have all their daughters on the right side,  The left component of the mother is written first, even though the right-hand component of the mother may be a straight line.  In the case of orphans, and when the daughter is significantly large, The daughter or leftmost orphan  is written first. So there are definite sequences in which the strokes of the mother are written, although it is not always as rigid as Chinese.

Stroke count, however,  is a bit trickier.  The majority of  The mothers and daughters have between three and five strokes.  Due to the genesis of the writing system in our world, there are no one stroke characters or mothers. So stroke number is not sufficient and itself to distinguish many of the characters for what I'm going to call alphabetization,  although I know that is not the correct term.  Currently I am thinking that the construction of a native dictionary would require a combination of stroke number and stroke order  and also the left or right component of the mother. 

  Those of you who have designed a writing system similar to what I have described,  how did you handle this issue?  Is there something I have neglected consider?  
If this makes a difference, this writing system is the formal, careful writing system designed for clarity rather than speed.  There is a cursive system, but I want to fully understand the more linear system  first, since the cursive system is a derivative of the formal linear system.

 One final observation: in some characters the rotation of the geometrical shape seems to affect the stroke order,  although rotation is not a fanatic or semantic indicator.