Mono required by ubuntu-desktop

Scott James Remnant scott at ubuntu.com
Fri Aug 4 00:54:00 BST 2006


On Thu, 2006-08-03 at 11:57 -0700, Dan Kegel wrote:

> On 8/3/06, Scott James Remnant <scott at ubuntu.com> wrote:
> > > Low-end PCs typically steal 64MB of that 256MB for video memory.
> > > So the budget is really 192MB, not 256MB.
> > > And ultra-low-end PCs have 128MB with 32MB stolen for video memory,
> > > so their budget is really 96MB!
> > >
> > Not many PCs (at least not many I've ever encountered) steal the RAM for
> > the video card, this tends to be a laptop trick (I'm willing to be
> > corrected here, I've just not seen it in even the cheapest desktops).
> 
> Then you haven't looked carefully.  It is very common.
> Dell's low end unit uses it:
> http://compreviews.about.com/od/budgetdesk/gr/DimensionB110.htm
> Heck, even Dell's low-end media computer uses it!  It's based on this chipset:
> http://www.intel.com/products/chipsets/gma950/
> 
Then I stand corrected;  I don't tend to deal with hardware that "high
up" and hadn't realised how prevalent it had become.

> > > > On an i386, the base OS uses around 32MB of memory and the
> > > > desktop uses a further 80MB of memory ... so we're using about 112MB
> > > > of memory before we've got any user applications loaded.
> > >
> > > On my dapper x86 system with 192MB or 512MB of RAM (booting
> > > with either size gives similar values), all root processes total
> > > 30MB of RSS (about what you quoted), but all user processes (not including
> > > firefox; all I have open is bash) total 163MB.
> >
> > Ah, you've made the first mistake everyone does when talking about RAM
> > usage.
> 
> Indeed, RSS looks unreliable.  (Do you happen to have a script that
> displays a more reliable measure of memory size for each app?)
> 
http://mail.gnome.org/archives/performance-list/2006-April/msg00000.html

Note that it's not possible to determine the memory size of an
individual application simply because there's no such thing.

Also the potential memory usage of an application is far higher than the
amount it will ever actually use at any one time... much of the memory
can be swapped to disk in favour of other applications, and much of it
may never need to be mapped in under normal circumstances.

Added to the fact that Linux tries very, very hard to keep the memory
usage at 100% at all times; it just becomes a nightmare to determine
true "minimum requirements".

> May as well measure what's really important: performance.
> 
That's the best way of doing it!  By adjusting the RAM available, and
measuring the performance, you can determine what the true "requirements
are", your figures below are very revealing!  Thanks for doing them!

> I took some measurements on a 1800MHz Athlon 64 machine running
> 32 bit Dapper, restricted to various amounts of system ram:
> 
> RAM   boot login abi ooo firefox
> 64MB  62 147 18 360? -
> 96MB  47 50  9  67 -
> 128MB 49 27 6 52 13
> 192MB 42 20 5 28 12
> 512MB 41 17 - - -
> 
Excellent, these show that the boot process uses something under 96MB of
RAM (any more doesn't make it significantly faster) and the desktop uses
something under 128MB of RAM (again, any more doesn't make it
significantly faster[0]).  This fits with my own estimates.

OpenOffice needs more, 192MB of RAM to actually get it loaded reasonable
fast.

Why are some of the figures missing here, btw?

> A reasonable criterion for acceptable performance is
> "boot and log in in under 60 seconds". 
> 
Boot time performance is tricky to measure at the moment, simply because
we spend a lot of it waiting for hardware to catch up.  45s is about the
minimum right now (the 40s was probably a warm boot, no?)

That's slow for reasons other than available memory, all those "sleep"
statements add up ;)

Scott

[0] 512MB is interesting simply because that's the point at which it's
    possible to keep the entire desktop in the page cache -- the more
    memory you have to store files, the less time you have to deal with
    the slow disks -- I don't think this matters for a "minimum"
    requirement though.
-- 
Scott James Remnant
scott at ubuntu.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 191 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/ubuntu-devel/attachments/20060804/98a336c1/attachment-0001.pgp 


More information about the ubuntu-devel mailing list