TMG-L Archives

Archiver > TMG > 2003-06 > 1054498419


From: Richard Brogger <>
Subject: Re: [TMG] Automatic "relation" tag - Opinions wanted
Date: Sun, 01 Jun 2003 15:13:39 -0500
References: <5.1.1.6.0.20030601140128.02d4a030@mail.fscv.net>


wrote:
>
> Yes, I would very much like to have multiple auto relationship tags.
>
> If I could see at a glance the relationship of the focus person to:
>
> 1. Joseph, the immigrant father
> 2. John who migrated to Rockingham
> 3. me
>
> it would tell me at a glance where this person fits in the tree and how
> much effort I want to put into him.
>
> Walt

Hi Walt,

It appears that you and I are seeing the same advantage of multiple
relationship tags. I too have a couple of persons that are keys to
determining how important other individuals are. Being able to see how
each person fits into the three way relationship would be nice nice
information to have.

I do think that we need to clarify at least one point. Do you want the
relationship tags to be generated one time and semi permanantly added
to the list of tags or do you want them generated each time you open
Person View? I use very small data sets and I can live with either
method. However, I can not see someone with a large data set wanting
to slow things down even more.

IMO, it is one thing to start a process one time and go to bed while
the computer makes hundreds of relationship tags. It is something
different to expect the computer to do those calculations as I wait.
To add three relationship tags in the Sample Project and do it on the
fly would not be a problem for an average computer. However, even a
top of the line, current computer would see a speed hit if the data
set had 200,000 people.

It was in a private exchange that the idea of adding the relationship
tags to a table came up. Doing so should make TMG faster but like any
idea, the results are not all positive. As I envision the process, the
relationship tags would have P1 and P2 like the marriage tag. For your
first example, Joseph would be P1. Generating the tags would not be a
big deal. John Cardinal might be able to do that with Utility 5.
However, if Joseph is related to 2,000 people in your data set,
opening Person View for Joseph would be something to behold because he
would have 2,002 relationship tags. (assuming that you used 3 focus
persons) That fact means that something has to be added to TMG so that
2,000 of those relationship tags do not appear in Joseph's Details
unless you want them to appear.

Another advantage to semi permanent relationship tags is that they
would have sentences that could be used in reports. Let's say that a
user has George Washington in their family and everyone in the family
wants to know how they are related to George Washington. That
relationship, for everyone who is related, could appear in the report.
Again there are always exceptions. For example, I can't imagine
including the relationship tag sentence in a report when P2 is related
1,234 ways to P1. However, if the object of the report is those
relationships, that might be useful. Like any feature in TMG, I am
sure that users will find uses that we have not thought of.

Oops! Maybe I have taken the wrong approach. In writing the comment
about other uses my brain went off on a tangent and I will blame it on
Teresa. She can find ways to use features that were never intended and
in thinking about that I came up with a different approach.

Let's forget the Automatic "relation" tag for a moment. Instead, let's
think about generation of Custom Tags as secondary output of a report.

Using the Automatic "relation" tag discussion as a basis, what if:
1) the user designs a custom "Related" tag
2) the List of Persons report could generate a "Related" tag for
everyone in the data set, who meets the filter criteria, and include
the results that would normally be produced by the Automatic
"relation" tag.

The results would be about like those of the proposed additions to the
Automatic "relation" tag. The ability to toggle those tags to Off for
P1's Person View would still be needed. However, the ability to add
custom tags would have uses far beyond showing just the results
currently produced by the Automatic "relation" tag. For example,
Kinship report results might be used instead of those produced by the
Automatic "relation" tag. I have not thought about the ramifications
of this idea but I believe it has potential.

Let's say that one makes a non person called "Rutheford Cemetery".
Using the secondary output of a report, a "Cemetery" tag is generated
for everyone buried in Rutheford Cemetery and they are made P2. When
the user went to the Rutheford Cemetery person, P1, that Detail would
show everyone buried in that cemetery. Now that sort of flips the
coin. I would not want the Cemetery tag to appear for P2 but I would
want it to appear for P1.

Since this whole idea is new to me, did I just lay an egg or does this
secondary output idea have possibilities?

Richard Brogger

This thread: