Re: [GT] trunk/GT/Tools.pm 2008-03-30 04:07:51 UTC (rev 594) or later

["Robert A. Schmied" <ras AT acm.org>]

Weigert, Thomas wrote:
> I had hoped to hide the data base access in the find_calculator routine,
> so that the scripts need not understand and deal with setting up the
> data base access. 

ok, other than not burdening already working scripts with two
statements how is than an improvement?

i can understand the change if things were in active development, but
why the change when it was fixing something that wasn't broke?

but assuming that would be a good thing (to hide the data base access...)
is there a way to go forward instead of changing everything back?

deleting the sole $db->disconnect() method seemed to do it, but i don't
know the downside to not disconnecting at script end. it may be somewhat
unimportant (and might could be handled via the perl equivalent of a
shell trap) considering that the scripts and many gt modules do a lot
of dies which may or may not do a proper db disconnect.

just one of the many many many details that have yet to be dealt with.

ras

> > -----Original Message-----
> > From: Robert A. Schmied [mailto:ras
AT
acm.org]
> > Sent: Saturday, May 03, 2008 7:32 PM
> > To: devel
AT
geniustrader.org
> > Subject: Re: [GT] trunk/GT/Tools.pm 2008-03-30 04:07:51 UTC (rev 594)
> > or later
>
>  
>
> > i'd still be interested understanding in your rationale to move it
> > outside of the application.