StarOffice: Slide sorter slow

[Roger Oberholtzer]

On Sat, 2004-03-27 at 15:41, Joel Hammer wrote:
> On Wed, Mar 24, 2004 at 08:46:02AM +0100, Roger Oberholtzer wrote:
> > Don't tell me that the images and things are IN the presentation. I did this
> > once with staroffice 5 and had a slow system as well. And my presentation
> > was only 32 MB or so. I found that if you linked to the images (by
> > reference) and other includes, you got much better time. That is because
> > only the visible images are needed at any given time. Loading the
> > presentation is very very much faster. When doing the actual presentation,
> > SO prefetches the next slide while showing the previous one. So the slides
> > progress at the same rate.
> 
> 
> OK. More action.
> 
> I prepared a powerpoint presentation with my macros,
> but edited them to link to the images in the directory,
> not include them into the powerpoint file. This worked
> fine in powerpoint. The powerpoint file size now is tiny.
> 
> I then transferred the whole thang, images and powerpoint
> file, to my linux box. Everything worked just like before
> StarOffice was WAY too slow on my 1 Gig duron to be
> usable. Not even close.
> 
> So, I must conclude that StarOffice badly needs some
> optimization. I noticed this incredible slowness before
> when I made a bar graph containing several hundred data
> points. I had to use GNUplot finally to do it.

This I can agree with. I had a small plot (less than 500 points) take
so long in SO to plot that at first I had thought SO had crapped out.
It has always been this way in the OO/SO series. The thing that I saw
that took all the time was plotting the axes. And, the crappier thing is
that you cannot modify / turn off plotting the axes until after the plot
is created. So at least the first time you do a plot you get coffee into
the bargain. I am sure deep in the muddy bowels of SO you can set the
default behavior for axes in all plots. But is it not anywhere obvious.
It starts each plot wanting to plot some thingie for each point on the
axis. Brilliant.

For this reason we are starting to do data viewing in 'R'. Damn but I
miss SAS. I wish we did enough statistics to warrant a license. Like in
the good old days.