On Fri, Mar 7, 2008 at 1:59 PM, User Scrappy <scrappy@hub.org> wrote:
>
>
> On Fri, 7 Mar 2008, Dave Page wrote:
>
> > We do tend to do that for all the stuff we deploy now. The main
> > exception to the rule is Marc's stuff that's been in place for many
> > years which is run from his own, largely undocumented scripts (hint
> > hint :-) ).
>
> Tell me what you want to see documented in ~pgsql/bin/mk-snapshot ...
> there isn't anything special in it ... its very self-explanatory (cvs
> export this, make that, move the result here) ... there is no voodoo
> involved here, or secret incantations ... and what runs, and paths to
> them, are plainly available in cron ...
>
I'm not specifically referring to that, but the various scripts that
evolved in the early days. A good example is the archives system which
hasn't really been worked on by anyone except you and very recently
Alvaro.
What we should have though is a brief description on PMT of each
service that exists, what scripts it involves, what user id it runs
under etc.
--
Dave Page
EnterpriseDB UK Ltd: http://www.enterprisedb.com
PostgreSQL UK 2008 Conference: http://www.postgresql.org.uk
