in

The ORM Foundation

Get the facts!

Feature Request - add role names to verbalization of fact types with ring constraints

Last post 05-13-2008 15:06 by Terry Halpin. 3 replies.
Page 1 of 1 (4 items)
Sort Posts: Previous Next
  • 05-13-2008 2:55

    Feature Request - add role names to verbalization of fact types with ring constraints

    Hi,

    Related to previous post: adding explicitly named roles to the verbalization would be useful in fact types with ring constraints - may make a good bridge to full verbalization of fact types with ring constraints.

    Thanks,

    BRN..

  • 05-13-2008 11:15 In reply to

    Re: Feature Request - add role names to verbalization of fact types with ring constraints

    Answer

    Hi Brian

    Currently we do not verbalize role names, but it would be a fairly simple exercise to add this for all fact types, not just ring fact types. Would that help?

    Cheers

    Terry

  • 05-13-2008 14:42 In reply to

    Re: Feature Request - add role names to verbalization of fact types with ring constraints

    Hi Terry,

    The only fact type, where verbalizing role names would help (that I can think of), is the recursive form (is there a better name for types where one Object plays both sides of a pedicate?) - so far, those are the only role names I (actually either your or Matt's suggestion), found it helps to to explicitly name roles.

    What might help is a warning issued when recursive fact type roles haven't been explictly named. Along the lines of the need to set a datatype; as that help avoid problems closer to physical implementation.

    Verbalization of all roles, only as a toggled option.

    Thanks again,

    BRN..

  • 05-13-2008 15:06 In reply to

    Re: Feature Request - add role names to verbalization of fact types with ring constraints

    Thanks for your reply, Brian.I prefer "ring fact type" for such cases, reserving "recursive" for those cases that involve recursion.e.g. "Person is an ancestor of Person" is a ring fact type that may be derived from the asserted fact type "Person is a parent of Person" using a recursive derivation rule such asPerson1 is an ancestor of Person2 iff Person1 is a parent of Person2  or Person1 is an ancestor of some Person3 who is an ancestor of Person2 .The use of subscripts to distinguish object variables avoids the need to explicitly supply role names (e.g. “parent”, “child”).That said, role names can be handy in lots of fact types, not just ring fact types, in order to express textual rules in attribute style.e.g. Person was born on Date [birthdate]  Person was hired on date [hiredate]. Constraint: For each person, hiredate > birthdate. CheersTerry

     

Page 1 of 1 (4 items)
© 2008-2014 The ORM Foundation: A UK not-for-profit organisation -------------- Terms of Service