[Babase] RE: Ositeti groups vs subgroups

Karl O. Pinc kop at meme.com
Fri Feb 5 12:43:24 EST 2010


On 02/05/2010 11:03:51 AM, Niki Learn wrote:
> On 02/05/2010 08:49:29 AM, Niki Learn wrote:
> There are 3 dates, start, end, and permanent. The latter is the
> date the group stopped being a subgroup.  The only rules are
> you can't be in a group before the start or after the end date and
> you can only be in one group on any given date.  Note however
> that if you're in a subgroup, on a date before the 'permanent
> date', you're also automatically considered to be in the
> supergroup. (This can be queried using the supergroup() function,
> which returns the supergroup if there is one on that date and
> the group itself otherwise.)
> 
> Niki:
> Hmm...  How does this supergroup thing work in members?  I guess if I
> just
> look up the individual it will show me which group we told babase the
> individual was in but if I looked up Ositeti as a group it would tell
> me
> everyone who was in that group including those specifically 
> identified
> as
> being in Jill's or Latin's group (6.1 and 6.2) as long as Ositeti
> (6.0) had
> not officially ended as a group?

If you:

select members.sname, members.date, members.grp from members;

you will see the group the individual's were entered into
in CENSUS.

If you:

select members.sname, members.date
       , supergroup(members.grp, members.date)
 from members;

You will see the supergroup, if it exists, or the
group entered in CENSUS if not entered into a subgroup.

We can always make a view that includes all the columns
in members and another column for supergroup(members.grp,
members.date).


  The advantage to keeping 6.0 open 
> as
> a
> group would be that we wouldn't need the unidentified group code
> (6.9), we
> could just say 6.0 when we don't know which group it is.

I think you lose information that way.  You won't be able
to tell if everybody comes back together for a day into
the single 6.0 group.  It may not "matter" late in the game
because the individuals never re-constitute the original group,
but the data would look the same as early in the fission process
when the individuals do come back together into one large group.

  That would
> save
> some messiness in the members table as the baboons wouldn't be
> bouncing
> between 6.1 or 6.2 and this fictitious 6.9 (of course some of them,
> like
> Lawyer, will bounce between 6.1 and 6.2 anyway), they'd just always 
> be
> in 6
> and sometimes specifically in 6.1 or 6.2.  On the other hand, it does
> seem
> that they are permanent groups now so in that respect it doesn't make
> sense
> to say that they are still in 6.0 - a clean break would be nice - and
> that's
> what we have done in the past so far (though perhaps that will change
> when
> we fix the census data for the Hook's and Alto's split periods?).  Of
> course
> the data on these "other groups" have a lot of holes so we will never
> have a
> real handle on when exactly they became two groups but 1 Feb seems
> like a
> good clean spot to use as our best guess.  

You'd make Feb 1 the stop date of 6.0 if nobody would ever be
censused in 6.0 after that date.


> This should also moot the subgroup notes too, right?  The
> individuals are put into CENSUS in their subgroup so there's no
> longer a need for a note that says who's in what subgroup.
> Right?
> 
> 
> Niki:  Yes, those particular subgroup notes go away.  Since the Other
> Group
> Notes that make up a lot of our Ositeti sightings aren't in the
> subgroups
> notes file anyway, I only got to delete three subgroup notes, and 
> then
> I
> altered the two with Kernel's group to make it clear that it is part
> of
> Jill's group.  And the subgroup file as it now exists does not tell
> you who
> is in which subgroup anyway.  It just gives information about when 
> and
> where
> there was a subgroup.  The individuals are listed on paper though.

There's no place for "where" in Babase now so I guess we are
looking at some sort of extension to be able to record this.


Karl <kop at meme.com>
Free Software:  "You don't pay back, you pay forward."
                 -- Robert A. Heinlein

P.S.  Can you configure your email program to work like every
other email program on the planet and have it indent or otherwise
mark what you're replying to so that you don't have
to write "Niki:" and it's easier for us to tell who wrote
what?





More information about the Babase mailing list