I guess I am not making myself very clear. There are no
leader/follower relationships, only units of named types.
Clan types may not be mixed, but some unit types may be in more than
one clan. example:
ClanA contains Units 1, 2
ClanB conatins Units 1, 3
ClanC contains Units 4
In Army builder I simply added a type for each clan that the unit
was included in so Unit 2 has the "type:ClanA" and "type:ClanB"
attributes assigned locally. What I want to do is not allow the
selection of units from multiple clans. Keep in mind there are
other units in the race with different types assigned are not part
of this equation and must be ignored.
So if I were to select the units 2&3 I would get an validation
message stating I had chosen units from more than 1 clan, whereas if
I selected 1&2 or 1&3 or 4 alone I would not have.
Thanks for your patience!
Woody
--- In armybuilder@y..., "Daniel Rothman" <drothman.dmth94@g...>
wrote:
> OH! you're looking for transitive support relationships. never
mind
> all what we said before.
>
> you're looking for the unit external attribute called reqd:type=
> (probably)
>
> some less abstract examples would probably help - f'rinstance,
you've
> probably got some leader-type units, and some follower-type units.
> questions like: can follower-type units "permit" other units in
(e.g.
> may a type-C leader, type C/D follower unit then buy a type-D
> follower unit?) are there other conflicts (e.g. type A hates type
D,
> outside of any other relationships...), etc. etc.
>
> it seems like you're trying to solve a general problem, when
you've
> got a specific task to accomplish. General mechanisms are good
(and
> suggesting good ones with possible implementations to Rob is a
Good
> Thing(TM)), but sometimes twisting the specific problem around
helps
> avoid difficulties with a general mechanism.
>
> daniel
>
>
> --- In armybuilder@y..., woodywalton@h... wrote:
> > Pardon me for stating the types were exclusive. They are not.
A
> > unit could potentially have all types. I neglected to mention
that
> > there are many other ancillary types as well.
> >
> > Example is as follows:
> >
> > An army may contain various units of types A & B, in addition to
> > units of one of the following types C, D, E, and F
> >
> > The types represent clan troop choices. Some clans can choose
from
> > some of the same types, while some have completely different
> > choices. Units will fall into multiple types so for example:
> >
> > Unit 1 is of type C and type D
> > Unit 2 is of type D and type E
> > Unit 3 is of type C and type F
> > Unit 4 is of type E and Type F
> > Unit 5 is of type C only
> >
> > I want a validation when someone tries to select units that have
no
> > type similarilties between C, D, E & F while at the same time
> > ignoring A and B altogether.
> >
> > In the above examples a validation would be triggered if one
> decided
> > to select Unit 1, and then tried to select Unit 4(no type in
> > common); It would not trigger if I selected 1, 3, & 5(type C in
> > common)
> >
> > Thanks again!
> >
> > Woody
> >
> >
> >
> > --- In armybuilder@y..., "Daniel Rothman" <drothman.dmth94@g...>
> > wrote:
> > > rob -
> > >
> > > i've been mulling this, and the problem really arises when
you've
> > got
> > > eight X types, and some other number of non-X types. the
> > validation
> > > you suggest will catch all of the non-X types as well as the X
> > types.
> > >
> > > It really looks like he's looking for a token mechanism -
perhaps
> > a
> > > unique (or limited quantity) item that can only be grabbed
> > > and "imprinted" by one (or two, or whatever) of the exclusive
> > groups.
> > >
> > > it'd be very nice if we could have roster types and/or stats -
> > it'd
> > > make this a much simpler proposition.
> > >
> > > daniel
> > >
> > > --- In armybuilder@y..., Rob Bowes <rob@w...> wrote:
> > > > It's a lot easier than that, fortunately.

You can use
> > the "!"
> > > syntax
> > > > and have merely 8 different "trat" attributes. Define each
one
> > by
> > > comparing
> > > > the presence of type1 to the presence of !type1. If each
type
> is
> > > exclusive,
> > > > then the presence of anything NOT of that type is illegal.
And
> > you
> > > really
> > > > don't even need the "-istype" filter on this, since the
> > validation
> > > message
> > > > will only trigger if you have something of typeX to begin
> > with.

> > > >
> > > > Hope this helps,
> > > > Rob
> > > >
> > > >
> > > > At 11:53 AM 8/22/2001 -0500, you wrote:
> > > > >All,
> > > > >
> > > > >Working out some of the last details for validation and was
> > > wondering if you
> > > > >could help with the following:
> > > > >
> > > > >If you had 8 exclusive unit types
(type:type1,type:type2,...)
> > And
> > > you wanted
> > > > >to trigger a validation message if any two different types
> were
> > > selected...
> > > > >
> > > > >would you have to write (8+7+6+5+4+3+2+1) 29
separate "trat"
> > > entries under
> > > > >the race special attributes, or could they all belong to
> > another
> > > mutual
> > > > >type, and then use a "-legal" or "-istype" expression to
make
> > the
> > > one
> > > > >exception and have only 8 "trat" entries.
> > > > >
> > > > >If so what would the "trat" and "-legal" or "-istype"
> > expression
> > > look like?
> > > > >
> > > > >My thoughts would be something similar to this although
with
> > > correct syntax:
> > > > >
> > > > >[trat:type1@1u:alltypes@0u-legal=(if alltypes AND NOT
type1)]
> > > > >
> > > > >
> > > > >Thanks,
> > > > >
> > > > >Woody
> > > >
> > > >
> > > > -------------------------------------------------------------
---
> -
> > ---
> > > -------
> > > > Rob Bowes (rob@w...) (650)
726-
> > 9689
> > > > Lone Wolf Development
> > >
www.wolflair.com
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Get your FREE credit report with a FREE CreditCheck
Monitoring Service trial
http://us.click.yahoo.com/M8mxkD/bQ8CAA/ySSFAA/IMSolB/TM
---------------------------------------------------------------------~->