The efficient method to create trigger

Hi,

I am new to PostgreSQL. I learnt most of things from postgresql.org site. I
want to write a trigger for on my table tech_projects such that whenever
INSERT, UPDATE or DELETE happen on this table, the trigger should get called
and the updated/deleted row must be added to tech_projects_backup table.
Moreover, if there are more than one row updated then I don't want to
reiterate the client-server communication for each row but instead it should
process all rows at once.

Could anyone please tell me which is the best way (PL/pgSQL, PL/Tcl, PL/Perl
or PL/Python). I know the Perl but not aware of Tcl and Python. It would be
great if some one could give example on processing multiple rows at once. I
am not sure how could I use spi_prepare for this.

Thanks, Jignesh