Startup time

[Charles Day]

On Thu, Sep 3, 2009 at 11:10 AM, Markus Melms <public1 at divara.de> wrote:

> On Thu, 20 Aug 2009 20:43:15 +0100
> Tarlika Elisabeth Schmitz <gnucash at numerixtechnology.de> wrote:
> > On Mon, 17 Aug 2009 22:25:59 -0400
> > Derek Atkins <warlord at MIT.EDU> wrote:
> >
> > >Seriously, GnuCash is just a large application.  It takes time to
> > >load. [...]
> > >Seriously, how long is it taking to load?  15 seconds?  20?
> >
> > v. 2.2.6 on Linux:
> >
> > 20s without account opening (--nofile)
>
> similar over here. Just stating that it's a large App does in my
> opinion not help much. Would be more interesting to think about how it
> could start faster.
>
> Currently it's perceived that gnucash is just
> inefficient regarding startup time. It's not taking advantage of
> multiple CPU cores since it's not parallelizing anything and it's slow
> on calculating reports, just to mention 2 points.
>
>
Making GnuCash thread-safe (to "parallelize") is non-trivial, and frankly
the front end needs more work than the back end in my opinion. However, Mike
Alexander (if I recall correctly) contributed a patch to improve the speed
of some reports by an order of magnitude. I believe those changes are only
included in the unstable series (2.3.x) though.

Of course there are more important things than gnucash startup time,
> but I would not say "lets not improve something bad because other
> applications are slow, too". And with changes being made it would
> start-up significantly faster, we are not talking about changing from
> 98% efficiency to 100%.
>
> Regards,
> Markus
>
>
Cheers,
Charles