Hi there,
should be first a discussion about what possibilities are, like approach
and then as already existing tools. Before getting to writing code or
scripts, maybe a fair analysis will be good.
From development effort point of view, I still think that first we need
a DB cleanup (as naming and installation procedures) before getting into
more complex tools.
I agree that a step at a time is good, but also a bottom up approach is
even better - it will same a lot of changes and re-changes....
regards,
Bogdan
Henning Westerholt wrote:
On Montag, 14. Mai 2007, Bogdan-Andrei Iancu wrote:
It is about DB re-organizing, I think it was a good point Klaus made
some time ago - all the tables should have a prefix identifying the
module they belong to. Right now there are tens of tables and for some
of them you do not have any clue what module is using them.
So, there are 3 directions here:
1) renaming the tables (adding a prefix of fix len)
2) to be able to configure what tables should be installed (we have
the core, presence, extra - right now, but something for details would
be good - maybe based on modules ?)
3) what Henning suggested - have a general description of DB and
some tools to generate particular DB definitions for mysql, postgres,
dbtext, etc..
Ok, later we can easily add some prefixes. For now i would like to stay with
the current layout, no need to do more then a few task at once. :-)
Are there any objections using parts of the SER scripts and schema files for
this task? This would me allow to save many hours, otherwise i need to
recreate all these stuff from the scratch. I mean, this is GPL code to, and i
would add of course proper copyright headers (giving credit) to their files.
I do some basic tests with the db schema system they created, and it works and
fits fairly well.
Henning
_______________________________________________
Devel mailing list
Devel@openser.org
http://openser.org/cgi-bin/mailman/listinfo/devel
