Page 1 of 1

Preferred Name Database Conservation

Posted: Fri Mar 30, 2007 1:25 pm
by atticusewig
I was recently updating the Preferred Names
in our Stake, to de-captilize the names that
were left over from that update we had
several months ago, and came upon
something.

Some people have no middle names, and
their preferred name was identical (with
exception of case) to their given name.
An example would be:

GivenName
[INDENT] David[/INDENT]
Preferred Name
[INDENT] David[/INDENT]

I guess that when then member signed
up on the LUWS, they filled out their
preferred name, although leaving it
blank would have accomplished the
same thing. Basically wasting space
in the database.

Now, while Ward Website Admins
could correct identical matchs of
single-name given names to the
enter preferred name, I imagine
it would be more productive for
the Church to do it on the database
side with a bit of SQL code.

Provided that the Preferred Name
database field is not always
fixed-width regardless of the
data in it, this should conserve
quite a bit of space and help
speed up the listing of the
Directories as well.

Another naming occurance I noticed
is the whole "My middle name is
one letter and only one letter".

Surname
[INDENT] Schuster[/INDENT]
Given Name
[INDENT] Simon L (with the L on its own)[/INDENT]

Sometimes, a website admin
makes the mistake of entering
in the preferred name for the
member and adding a period
after that letter. Trapping
instances of this practice
with a bit SQL code, might
catch a lot of unnecessary
punctuation, and save a
small amount of database space.

- Atticus Ewig

Posted: Sat Mar 31, 2007 7:18 am
by WelchTC
Interesting observation. However I will note that if there were 10 million records in the database and each one had on average a 10 character name that was doubled, that would only be 100 million bytes. In 1990 that was a lot of storage space. In today's world that is fairly inconsequential. I do like the way you are thinking, however. Waste not, want not. ;)

Tom

Posted: Sat Mar 31, 2007 1:44 pm
by russellhltn
tomw wrote:However I will note that if there were 10 million records in the database and each one had on average a 10 character name that was doubled, that would only be 100 million bytes.
On the flip side, for every record that didn't have a Preferred Name, the computer would have to calculate one. So one might argue that it's better to give everyone a Preferred Name to save CPU cycles. Classic trade-off between storage efficiency and CPU efficiency.