[Babase] Conference call summary
Stephanie Combes
babase@www.eco.princeton.edu
Wed, 12 May 2004 14:50:15 -0400
Hi everyone,
The following is a summary of our discussion today, 12 May 2004.
1. Duke ACCESS to Princeton (Daphne's) machine: Currently, only Susan's
computer at Duke has FTP access to the Princeton computer. Susan believes
that this was done for security and financial reasons. IF it is true that
only ONE computer can have access, we will replace Susan's machine with the
login.biology.duke.edu machine. If more than one machine can be listed as
having access, Susan's machine will remain on. Daphne will check with Sal
regarding this set-up. She needs to tell Sal that the
login.biology.duke.edu (good enough, shouldn't need numeric ip) needs
access. Daphne should also inform Sal that many people have access to the
"login" machine - Daphne will report back to us with any of Sal's
concerns/questions.
1a) As a security aside, apparently the security issue arises from our use
of FTP. We could, in theory, switch to something like SSH which encrypts
passwords. The need for and difficulty of this switch is unclear right
now.
2. INTEGRITY checks/rules: First of all, all of the rules that Karl wrote
in and are currently in use for Babase are documented in the Babase manual.
We should, perhaps, all read it over again once in awhile. For the current
system, we (Daphne or Steph) could make our own "cheat sheet" summarizing
the rules, by reading through the manual. For our new system, Karl is
keeping track of changes and rules. Karl will create our new system to not
allow us to enter data that violate the rules (i.e. you will get an
immediate error). For things that may need to be changed at some point (or
errors we temporarily ignore), we will have daily or weekly (need to figure
out interval) warnings of things that the system doesn't like. [Karl -
chime in here if I'm wrong].
3. DOCUMENTATION of rules/integrity checks: new system - Karl is keeping
track. We need to eventually update/change all of the documentation. Karl
suggested docbook (see www.docbook.org) as a system for doing this. We
can have the new manual set up so that it generates an email/message
telling us that a change has occurred (it may not tell us what, exactly,
occurred, only that a change has taken place. so go read the manual
again). In addition, Daphne/Axel will set up another mailing list for
announcements ONLY. When you make a significant change to the database
(i.e. changed the rules for dcauses) send an email to the announcement
list. This second list will have limited traffic. It will provide us with
a way of zeroing in on emails that create significant changes, and prevent
this information from getting buried under a million emails). The manual
and database will be stored in a shared repository so that changes can be
made from Duke or Princeton.
4. DCAUSE: The plan is as follows: 1. Steph is in charge of fixing up the
dcauses. 2. Steph will create a file for Daphne that has 2 columns - sname
and dcause. 3. The dcause column will have the updated dcauses that were
done by Jeanne and Jessica. 4. The dcause column will also contain "8"
for all other dcauses (blanks and old dcauses). "8" will now be "under
review". These "under review" dcauses will be reviewed and re-assigned by
Steph (with input from Jeanne and Daphne) at a later date. When this later
date arrives, Steph, Jeanne and Daphne will need to go through and clarify
some of the dcause rules (example: the abort definition due to mom dying).
She will, obviously, update all documentation appropriately as listed in
point 3 of this email! Moreover, Karl will relax the data integrity rules
for dcause. This entails no longer having a rule for aborts/fetal loss.
The only dcause rule that will stay in place is that a dead animal must
have a dcause (even if the dcause is "unknown").
5. The CONVERSION: Karl has not been able to work on babase since leaving
Duke. He will likely be able to return to babase at the end of this week.
(Go Karl go!).
I think that's it. Please add any clarifications. I'm not sure if my
email is as long as Susan's, but maybe I can top it next week!
'Til next Wednesday (same bat time, same bat place).
-steph