Common Interests
Michael Pasterik
We share common interests in genealogy, programming, and database
design. I developed an 'old style" genealogy program, "Your Family
Tree" for the Amiga back in 1988. Development on that ended
somewhere back in the early 90's as a growing family and other
interest took over. Every so often my mind wanders to the database
structure I have been kicking around for several years, which is very
similar to Exodus. As winter approaches, I hope to have more time to
improve my beginner php/sql skills. I plan to read up on your latest
designs and give the latest code a run.
--Mike
NW Pennsylvania USA
Leif Biberg Kristensen
thanks for your input.
With your background, I believe that you will find that both PHP and
SQL are rather simple languages. PHP in its procedural aspect is
syntactically very similar to C, but a lot more relaxed. The Yggdrasil
code uses a rather small subset of PHP.
SQL takes "a moment to learn, but a lifetime to master". If you want a
good introduction to SQL, I'll highly recommend Joe Celko's "Data &
Databases: Concepts In Practice".
A lot of the Yggdrasil code is written in the procedural language PL/
PGSQL. Anyone who have coded in Algol or Pascal will immediately
recognise the syntax.
It seems like this project appeals mostly to people who started to
program back in the Eighties. I suspect that the lack of objects may
appear "retro" to the kids, who don't realize that there are other
approaches to modular design than OOP. I have found that PHP as a
database frontend lends itself well to a procedural, task-oriented
design. Also, a lot of the business logic resides within the database
as constraints, views, and functions. This "fat db/thin frontend"
paradigm is very alien to the ORM camp, who rather want a massive
object framework to interact with a stunted database backend.
I hope you realize that the Yggdrasil project is not a "Family Tree"
program, although it may certainly contain your own family tree - mine
does :) The scope of the project is to provide a research tool for
exhaustive "one-place" studies, where eg. a parish register is entered
into the source apparatus in its entirety. The hierarchical source
model will probably appear cumbersome to a researcher who follows
families over large areas.
That being said, I certainly wish everybody who wants to mess around
with the code welcome to the project.
regards, Leif.