[scribus] Scribus on 'modern' machines

[Craig Ringer]

John Beardmore wrote:
> Craig Ringer wrote:
>> It's often informative to run the command "vmstat 5" in a terminal. 
> 
> Unfortunately I don't have such a tool for XP.

I use Vista much of the time myself (I can't even come close to getting 
Linux to match the 7 hour battery life I get under Vista), yet I still 
think in UNIX.

I think on XP you can get some of the same information using 
"perfmon.msc" (start->run, perfmon.msc).

You can also watch the free/cached RAM and page file use in task 
manager. Some page file use is normal on Windows, but if you see it 
going up significantly when performing tasks and the amount of RAM shown 
as used for cache is fairly low you're probably seeing significant 
memory pressure.

>> "watch lpstat"
> 
> Again not on XP.  I was wondering if GDI calls might be taking a lot of 
> resource, but I don't know.

You might find Process Monitor and/or Process Explorer from sysinternals 
  useful for getting an idea of what processes are doing what on the CPU 
and with other system resources.

http://www.microsoft.com/technet/sysinternals/Security/ProcessExplorer.mspx
http://technet.microsoft.com/en-us/sysinternals/bb896645.aspx?PHPSESSID=d926bdd849b5aab10f7263dd7f5904f2

On XP Process Explorer is probably the better choice for what you're 
trying to do.

In any case, if there's lots of time being taken up by GDI calls etc I 
wouldn't be surprised if it was happening within the Scribus process.

>> By the way, many applications (but not Scribus most of of the time) 
>> benefit quite significantly from a fast hard disk, as does general 
>> system responsiveness.

> OK -  an interesting point.
> 
> I get the impression that this laptop doesn't spend much time accessing 
> the hard disk, so hadn't thought of going for a faster one, though I 
> guess that at least the first time a program is launched will be disk 
> limited.

It depends a great deal on what your workload is. There's always going 
to be an improvement in system responsiveness but it does vary based on 
how much RAM is available for disk cache and other things.

> I suppose the other point is that a faster disk would probably use more 
> power.

On my machine it doesn't seem to make much difference - I get six to 
seven hours of battery life, much the same as the others with the 
5400rpm disks. It's an XPS M1330 that's reasonably power efficient (core 
2 duo with aggressive power management settings, display with LED 
backlight, etc) so it's not a case of the machine being such a power hog 
that the disk doesn't matter. I upgraded my old laptop to a 7200rpm disk 
from a 5400rpm and didn't notice any detectable drop in battery life on 
that one either. It *does* make a difference if I'm doing something that 
lets the disk shut down entirely for long periods, so it appears that a 
7200rpm disk just doesn't use much more power than a 5400rpm.

--
Craig Ringer