home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Monster Media 1993 #2
/
Image.iso
/
gene
/
pafrv934.zip
/
INSTALL.EXE
/
FAMCHECK.REV
< prev
next >
Wrap
Text File
|
1993-05-01
|
4KB
|
63 lines
╔══════════════════════════════════════════════════════════╗
║ Review of PAFCHECK - From File PAFCHECK.REV ║
╟──────────────────────────────────────────────────────────╢
║ PAF Review - Edition 93.3 - May 1993 ║
╚══════════════════════════════════════════════════════════╝
FAMCHECK analyzes your PAF data and points out possible logical
problems, particularly with dates. A correspondent pointed out
that FAMCHECK was written before release 2.2 of PAF and questioned
whether it was still of value. In response I decided to run
FAMCHECK through the mill at St. Clair Laboratories. Having made
a copy of my PAF data, (please be faithful in doing this!) I went
on to configure and run FAMCHECK to see: first - could I get it to
run; and second - would it find anything in my data.
The user interface is not pretty. The user must enter all
parameters in a configuration file with an editor. Each has to be
on the correct line; it does not support a "keyword=" format. You
specify the location of the three PAF files (INDIV2, MARR2, and
NAME2.DAT), the location and name of a file in which to record the
results, and the starting and ending marriage number (or zeros in
to do the whole file).
When I asked FAMCHECK to check the first 100 families, it
completed in a few seconds, showing a couple of violations of the
rules. Not having any difficulties with that, I turn it loose on
my entire database of about two thousand families. By the time I
walked to the microwave and back, it was done. Note that it may
take longer on a slower computer.
The result of this run was a file that was five pages long when
printed. Each error message included an MRIN, one or two RIN's,
one or two names and a descriptive message. It was ordered by
MRIN. Every item on the report was indeed a potential problem,
according to the rules defined in FAMCHECK's documentation. I
didn't know whether to be impressed or embarrassed!
Since PAF now has a similar report available on its menu, I
considered the possibility that perhaps FAMCHECK should be
retired. After a little thought, though, I decided there were
some valid reasons not to do so:
- FAMCHECK's report has a predictable order (by MRIN) while the
PAF report is completely random, as near as I can tell.
FAMCHECK also includes the MRIN on the listing.
- FAMCHECK includes complete source code. It could be customized
by anyone with C programming skills to meet their particular
needs. The checking criteria in the PAF version cannot be
changed, but simple changes in the criteria used by FAMCHECK
would not be difficult.
- FAMCHECK is an excellent, working example for someone wanting
to use PAF data from the C programming language.
Perhaps someone will decide to add a good user interface, nicer
looking reports, and validation criteria that the user can
configure. This would make FAMCHECK a data consistency checking
tool that is both easy to use and extremely valuable. Perhaps
this will be done at some point by FAMCHECK's author, or someone
else with his permission. I hope all you programmers out there
are paying attention to this one.