[Babase] Re: next conversion
Karl O. Pinc
kop at meme.com
Tue May 22 16:29:46 EDT 2007
On 05/22/2007 03:06:54 PM, kfenn wrote:
> I'm guessing I should run "buildmembers" to update the Babase 1.0
> members table and make that update available in the babase database
> in Babase 2.0.... but I wonder if I should run the repstats and
> cycstats programs? The most accurate repstats/cycstats tables are
> probably in babase_test, correct? Aren't the babase cycstats and
> repstats tables still going to be lame, even if I update them in
> Babase 1.0?
It never hurts to run buildmembers, and it's probably
a good idea if you've gone back and changed older
CENSUS rows. However, niether the MEMBERS
or the CYCSTATS or REPSTATS tables are actually moved from
foxpro, they are all re-computed. So are all the tables
in the docs "Analyzed" section, excepting RANKS which is
analyzed by hand.
Strictly speaking there's not a clear division between
"Analyzed tables" and not analyzed tables. Consider
CYCPOINTS which has automatic Mdates.
All the analyzed data is rebuilt when I copy foxpro
into the babase database. But when I run the conversion
into babase_test everything is built automatically,
and that means that CYCSTATS and REPSTATS are not
generated because they still have to be manually
rebuilt in the new system. Otherwise, conversion
error excepted, babase and babase_test are identical.
See:
rebuild_all_cycstats
http://papio.biology.duke.edu/babase_system_html/re13.html
rebuild_all_repstats
http://papio.biology.duke.edu/babase_system_html/re17.html
Karl <kop at meme.com>
Free Software: "You don't pay back, you pay forward."
-- Robert A. Heinlein
More information about the Babase
mailing list